Perfectionism, Failure and Self-conscious Emotions: A role for

Self-compassion?

Natalie Rebecca Almond

Submitted for the degree of

Doctor of Psychology (Clinical Psychology)

School of PsychologyFaculty of Arts and Human Sciences

University of SurreyGuildford, SurreyUnited KingdomSeptember 2015

Abstract

Objectives

This study investigated: (1) whether maladaptive perfectionism predicted the

experience of self-conscious emotions such as shame, guilt and pride following an

imagined failure and (2) whether self-compassionate writing could reduce shame and

guilt and increase pride relative to a control and self-esteem writing task.

Design

The study used a correlational design to assess the relationship between

maladaptive perfectionism and self-conscious emotions. The study also used an

experimental between-subjects design to investigate the effect of writing task on

self-conscious emotions controlling for initial levels of self-conscious emotion using

ANCOVA.

Methods

Ninety-five University of Surrey students completed an online study that

manipulated imagined failure on an academic assignment, and measured maladaptive

perfectionism and shame, guilt and pride. Participants were then randomly allocated

to either a self-compassionate, self-esteem or a control writing task. Self-conscious

emotions were then measured again.

Results

Following imagining failure maladaptive perfectionism was positively

associated with state shame and guilt and negatively associated with state pride.

When measured after the writing tasks, the means for shame and guilt were lowered

and the mean for pride was increased. However, contrary to predictions, shame was

not predicted by writing condition, guilt remained highest following completion of a

self-compassionate writing task and pride was highest following the completion of

the control-writing task.

Conclusions

Maladaptive perfectionism is correlated with self-conscious emotions

following imagined failure. Self-compassionate writing tasks do not appear to be

more effective at improving self-conscious emotion than other writing tasks.

2

Acknowledgements

I would like to thank all of the service-users I have worked with over the

course of the last three years. I have learnt something valuable from each person and

I sincerely hope that the work was valuable for the service-users. I would also like to

thank all of the clinical psychologists that welcomed me into their teams and have

supervised my clinical practice during training. I have learnt vast amounts from each

supervisor in terms of your therapeutic skills, working in teams, self-care, and what

makes good supervision!

I would like to thank all of the research supervisors that have been involved

in contributing to this research and my personal and professional development as a

clinical psychologist, in particular Dr Laura Simonds, and Dr Linda Morison. In

addition, Dr Clara Strauss, Dr Jason Spendalow, and Dr Kate Gleeson all played

important roles in the process of developing this project for which I am grateful. I

would also like to thank Andrew Barnes for his technical help and support in setting

up the online questionnaire. I am indebted to the kindness of the University of Surrey

students who completed my online questionnaire: without them there would not be

any research data.

I would like to recognise the numerous clinical and academic psychologists

who contributed to the taught programme. I have learnt a great deal about a great

many topics and I am sure that this comprehensive programme of teaching will

provide a useful resource to inform my thinking for many years to come. I would

also like to thank Elly Tomlinson who acted as my mentor throughout training.

Elly’s experience, wisdom, kindness and support throughout the rollercoaster of

training has been invaluable.

Finally, I would like to thank my wonderful husband Paul without whom I

could not have completed this training course. I’m not sure either of us knew at the

beginning of this course what we had let ourselves in for! Over the last three years I

have developed as a psychologist and as a person and you have made space for this

change and supported it. You have provided emotional support, practical support,

friendship and love, unwaveringly. Thank you for celebrating my successes and

supporting me in the face of difficulties. Bowlby, (1973) describes the benefits of an

attachment relationship that provides a secure base as follows: “Human beings of all

ages are at their happiest and are able to deploy their talents to their best advantage

when they are confident that standing behind them, there is a trusted person who will

come to their aid should difficulties arise.” Thank you for being my secure base and

supporting me to flourish.

2

Table of contents

Abstract.........................................................................................................................1

Objectives...................................................................................................................1

Design........................................................................................................................1

Methods......................................................................................................................1

Results........................................................................................................................1

Conclusions................................................................................................................2

Acknowledgements.......................................................................................................1

Table of contents..........................................................................................................3

MRP paper....................................................................................................................8

Perfectionism, Failure and Self-conscious Emotions: A role for Self-compassion?

........................................................................................................................................8

Abstract.........................................................................................................................9

Statement of journal choice.......................................................................................11

Introduction................................................................................................................12

What is perfectionism?.............................................................................................12

Perfectionism and failing to meet high standards...................................................12

Failing to meet standards and shame......................................................................13

Failing to meet standards, perfectionism and shame..............................................13

Perfectionism and self-conscious emotions.............................................................13

Maladaptive perfectionism and self-conscious emotions.........................................14

Shame and self-compassion.....................................................................................17

Overview of study.......................................................................................................19

3

Method........................................................................................................................19

Ethical Approval......................................................................................................19

Study Design............................................................................................................19

Procedure.................................................................................................................19

Dependant measures................................................................................................24

Maladaptive Perfectionism..................................................................................24

Shame, guilt, and pride........................................................................................25

Participants..............................................................................................................25

Results.........................................................................................................................27

Data handling..........................................................................................................27

Analytic approach....................................................................................................28

Correlational analysis..............................................................................................28

Effect of writing condition on Time 1 variables.......................................................30

Analysis of Covariance (ANCOVA) Analyses..........................................................31

Content Analysis......................................................................................................35

Discussion....................................................................................................................41

Maladaptive perfectionism, and self-conscious emotions........................................41

Self-compassion and self-conscious emotions.........................................................43

Limitations and suggestions for future research......................................................46

References...................................................................................................................48

List of Appendices......................................................................................................56

Appendix A.................................................................................................................58

Appendix B.................................................................................................................62

Appendix C.................................................................................................................63

4

Appendix D.................................................................................................................64

Appendix E.................................................................................................................78

Appendix F..................................................................................................................79

Appendix G.................................................................................................................84

Appendix H.................................................................................................................85

Appendix I..................................................................................................................88

Appendix J..................................................................................................................89

Appendix K.................................................................................................................90

Appendix L.................................................................................................................92

Appendix M................................................................................................................93

Appendix N.................................................................................................................99

Appendix O...............................................................................................................100

MRP Research Proposal..........................................................................................101

Can compassionate-imagery reduce perfectionist’s experience of shame?........101

Introduction..............................................................................................................102

Background and Theoretical Rationale.................................................................102

Research Question.................................................................................................105

Main Hypotheses....................................................................................................105

Method......................................................................................................................106

Participants............................................................................................................106

Design....................................................................................................................107

Measures/Interviews/Stimuli/Apparatus................................................................107

5

Perfectionism measure...........................................................................................107

Manipulation of failure..........................................................................................108

Measures of shame and guilt.................................................................................108

Manipulation of compassionate-imagery...............................................................108

Measure of mood....................................................................................................109

Procedure...............................................................................................................109

Ethical considerations............................................................................................110

R&D Considerations..............................................................................................111

Proposed Data Analysis.........................................................................................111

Service User and Carer Consultation / Involvement.............................................112

Feasibility Issues....................................................................................................112

Dissemination strategy...........................................................................................113

Study Timeline........................................................................................................113

References.................................................................................................................114

MRP Literature review...........................................................................................120

Perfectionism, parental bonding and attachment.................................................120

Journal choice and justification..............................................................................121

Abstract.....................................................................................................................122

Introduction..............................................................................................................123

Perfectionism.........................................................................................................123

Measures of perfectionism.....................................................................................124

Defining perfectionism...........................................................................................125

Development of perfectionism................................................................................126

Attachment and perfectionism................................................................................127

6

Measuring attachment............................................................................................128

Method of the literature search..............................................................................130

Inclusion/ Exclusion criteria..................................................................................130

Results.......................................................................................................................131

Perfectionism and Parental Bonding Inventory.....................................................133

Overall summary of perfectionism and parental bonding inventory.....................138

Adult attachment....................................................................................................139

Overall summary of perfectionism and adult attachment......................................146

Discussion..................................................................................................................149

Theoretical implications........................................................................................149

Limitations and implications for future research..................................................151

Conclusion.............................................................................................................153

References.................................................................................................................154

Clinical Experience..................................................................................................160

An overview of the clinical experiences gained on placement.............................160

Year 1.....................................................................................................................161

Year 2.....................................................................................................................161

Year 3.....................................................................................................................162

Assessments...............................................................................................................164

Year I Assessments.................................................................................................165

Year II Assessments................................................................................................165

Year III Assessments..............................................................................................166

7

MRP paper

Perfectionism, Failure and Self-conscious Emotions: A role for Self-

compassion?

8

Abstract

Objectives

This study investigated: (1) whether maladaptive perfectionism was

associated with the experience of self-conscious emotions such as shame, guilt and

pride following an imagined failure and (2) whether self-compassionate writing

could reduce shame and guilt and increase pride relative to a control and self-esteem

writing task.

Design

The study was a randomised control study. Following an imagined failure the

effect of three different writing tasks (control, self-esteem and self-compassion) on

shame, guilt and pride was assessed.

Methods

Ninety-five University of Surrey students completed an online study that

manipulated imagined failure on an academic assignment, and measured maladaptive

perfectionism and shame, guilt and pride. Participants were then randomly allocated

to either a self-compassionate, self-esteem or a control writing task. Shame, guilt and

pride were then measured again.

Results

Following imagining failure maladaptive perfectionism was positively

associated with state shame and guilt and negatively associated with state pride.

When measured after the writing tasks, the means for shame and guilt were lowered

and the mean for pride was increased. However, contrary to predictions, there was no

significant effect of writing condition on shame. Writing condition did significantly

effect guilt and pride but not as hypothesised. Guilt was least reduced by the self-

compassion writing, and pride was most increased following control writing.

9

Conclusions

Maladaptive perfectionism is correlated with shame, guilt and pride

following imagined failure. Self-compassionate writing tasks do not appear to be

more effective at improving self-conscious emotion than other writing tasks.

10

Statement of journal choice

I plan to submit this paper to the British Journal of Clinical Psychology. The

guidelines for authors submitting to this journal can be seen in Appendix A. This

journal is appropriate because the content of the current paper falls under the overall

aims and scope of the journal. The British Journal of Clinical Psychology publishes

original research that is empirical in nature and covers all aspects of clinical

psychology. The journal considers psychological treatment interventions of

individuals, which is important, because this is what one part of my study considers.

Also the journal consider papers considering all psychological disorders regardless

of age groups or setting, this is important given my use of a non-clinical sample. The

journal has previously published empirical investigations of perfectionism in non-

clinical samples.

I have chosen this particular clinical psychology journal because it has a good

impact factor of 2.377. Much research on self-compassion has been published in

social and personality psychology journals and whilst some of these may have higher

impact factors, I believe it is important to publish these findings in a clinical

psychology journal because this journal is more likely to be read by more clinical

psychologists.

11

Introduction

In this study I consider whether perfectionism is associated with the extent to

which an individual experiences self-conscious emotions (such as shame, guilt and

pride) following the experience of failing to meet their self-defined standards. I also

consider whether a brief self-compassionate writing task can reduce shame and guilt

and increase pride relative to a self-esteem and control writing task.

What is perfectionism?

In this study perfectionism is defined as “… the setting of, and striving to

meet, very demanding standards that are self-imposed and relentlessly pursued

despite causing problems. It involves basing one’s self-worth almost exclusively on

how well these high standards are pursued and achieved.” (p. 9, Shafran, Egan, &

Wade, 2010). Perfectionism is associated with predictors of maladjustment such as

negative affect (see Stöber & Otto, 2006) and predicts and maintains a range of

mental health difficulties and distress, (see Egan, Wade, & Shafran, 2011 for a

review).

Perfectionism and failing to meet high standards

A perfectionist’s sense of self-worth is overly dependent on their ability to

strive for and meet their own self-defined (unrealistically high and inflexible)

standards. Thus perfectionists direct much effort self-evaluating their progress in

meeting these standards. Shafran et al (2010) proposed that perfectionists often fail

to meet their self-defined standards for performance and this failure leads to

counterproductive behaviours (such as procrastination or avoidance) and self-

criticism. These outcomes maintain the perfectionistic striving. The perfectionist will

continue to strive to improve their self-worth by re-attempting to meet their high

standards. Ironically, then perfectionists commonly experience failure, and this

12

failure perpetuates further perfectionistic striving (and further failure). The emotional

consequences of failing to meet these standards for maladaptive perfectionists are

currently unclear.

Failing to meet standards and shame

Failing to meet ideals, standards and live up to expectations has been

consistently linked to the experience of shame (Miller, 1996; Reimer, 1996;

Sorotzkin, 1985, see Gilbert, 1998 for a review). Tangney (2002) describes shame as

a painful negative scrutiny of the entire self and a feeling that “I am an unworthy,

incompetent, or bad person.” Gilbert (1998; 2003) suggests that shame is a perceived

failure to create a desirable, attractive self and suggested that there may be two types

of shame, internal shame and external shame. Internal shame originates inside the

self and involves self-generated criticism, and negative self-evaluation. External

shame, which originates outside the self, involves a distressing awareness that others

view the self negatively.

Failing to meet standards, perfectionism and shame

Mascolo and Fischer (1995) suggested that shame is generated by the

appraisal of having failed to live up to personal (and other peoples) standards of

worth and that shame is functional because it highlights behaviours that threaten self-

worth. Perfectionists are continually failing to live up to their own personal standards

and therefore never attain the self-worth that they are seeking. Furthermore

perfectionists monitor and evaluate whether they are meeting these standards. When

perfectionists fail to meet their standards they engage in negative self-evaluation and

self-criticism and thus are likely to experience an internal sense of shame.

Perfectionism and self-conscious emotions

Shame is one of a number of self-conscious emotions, along with guilt and

13

pride. People experience self-conscious emotions when they evaluate some aspect of

the self (personality characteristics, abilities, or behaviour) against a set of standards

(Tangney, 2002). Perfectionists are continually self-evaluating and often conclude

that they have underachieved, thus there has been repeated suggestions in the

literature then that perfectionists may be prone to experiencing shame and guilt and

may be unable to experience pride (e.g. Hamacheck, 1978; Hewitt & Flett, 1991;

Sorotzkin, 1985).

Shame and guilt are often perceived to be similar and overlapping so it is

plausible that perfectionists may also be more likely to experience guilt. Both shame

and guilt arise from negative self-evaluations and may be elicited by similar events

and scenarios. However, different types of appraisal may trigger guilt and shame: a

negative evaluation of the self will result in shame; whereas a negative evaluation of

the action or behaviour will result in guilt. To clarify, shame is associated with

appraisals that “I am bad”, whereas guilt is associated with appraisals that “I did a

bad thing” (Lewis, 1971). Pride on the other hand is a positive self-conscious

emotion arising from positive self-evaluations that contribute to a person’s sense of

self-worth pride is associated with feelings of accomplishment and satisfaction.

Given perfectionists tendency to chronically self-evaluate and conclude that they

have underachieved it seems likely that perfectionists may experience less pride.

Maladaptive perfectionism and self-conscious emotions

Stöber, Harris, and Moon, (2007) note that despite the longstanding

hypotheses that perfectionists are likely to experience more shame and guilt and less

(if any) pride, empirical support for this hypothesis is sketchy. There are a number of

reasons why this is the case. Firstly, perfectionism has been conceptualized in a

number of ways and secondly there is variation in whether proneness to self-

14

conscious emotions, or state or trait self-conscious emotions is considered.

Perfectionism has often been considered as a multi-factorial concept and thus

the most commonly used measures of perfectionism measure different aspects of

perfectionism. The Multifactorial Perfectionism Scale (F-MPS: Frost, Marten,

Lahart, & Rosenblate, 1990) measures concern over mistakes, doubts about actions,

personal standards, parental expectations, parental criticism, and organization and

the Multifactorial Perfectionism Scale (H-MPS: Hewitt & Flett, 1991) measures self-

oriented perfectionism, other-oriented perfectionism and socially-prescribed

perfectionism; and the Almost Perfect Scale- Revised (APS-R: Slaney, Rice,

Mobley, Trippi & Ashby, 2001) measures discrepancy, high standards, and order.

Research using these multifactorial measures of perfectionism to investigate self-

conscious emotions has demonstrated that self-oriented perfectionism is associated

with proneness to guilt (Hewitt & Flett, 1991), socially-prescribed perfectionism is

associated with proneness to guilt and shame (Hewitt & Flett, 1991; Tangney, 2002)

and that negative perfectionism is associated with both state and trait shame and guilt

(Fedewa, Burns & Gomez, 2005).

More recently numerous factor analytic investigations have attempted to

identify the key aspects of perfectionism (Bieling, Isralei, & Antony, 2004; Dunkley,

Blankstein, & Berg, 2012; Enns, Cox, & Clara, 2002; Rice, Ashby, & Slaney, 1998;

Suddarth & Slaney, 2001). Reviewing these factor-analyses Stöber & Otto (2006)

concluded that perfectionism has two factors; an adaptive factor that captures having

high personal standards and striving for excellence and a maladaptive factor that

seems to capture self-evaluation/ criticism and monitoring of whether an individual

is meeting their own standards. This has lead Stöber, et al. (2007) to argue that only

maladaptive perfectionists should experience increased shame and guilt and reduced

15

(or absent) pride.

Only one study has examined how maladaptive perfectionists state shame,

guilt and pride were affected by failing a task (Stöber, et al, 2007) and there were

numerous problems with this study that make the results difficult to interpret.

Success and failure was defined in relation to objective standards (university grades)

rather than in relation to the individual’s self-defined standards and expectations for

their own performance. It is failure to meet self-defined standards that is central to a

maladaptive perfectionist’s experience. Therefore this study will consider whether

the extent to which an individual endorses maladaptive elements of perfectionism is

associated with the extent to which they experience state shame, guilt and pride

following failure to meet their own self-defined standards for performance. Based on

the existing literature it is predicted that following a failure (to meet self-defined

standards of performance) higher levels of maladaptive perfectionism will be

associated with higher levels of state shame, and may be associated with higher

levels of state guilt and lower levels of state pride.

In the second part of this study the aim is to consider whether a brief

intervention at the point of failing to meet self-defined standards can reduce the

emotional impact of this failure. As noted above failing to meet standards may elicit

both shame and guilt. In shame the object of negative evaluation is the whole self

(Lewis, 1971). The experience of shame is associated with stable and global

attributions about the self, an inward attentional focus, the desire to escape/avoid,

and meaning-making that the individual is bad or worthless (Tangney & Dearing,

2002). Guilt is often attributed to the action taken and so may be less damaging for

the individual in the longer term. However, guilt and shame are often both elicited by

similar situations therefore, guilt will still be measured in the second part of the

16

study but reducing shame is the primary focus. Research has suggested that the

development of self-compassion may serve to reduce an individual’s experience of

shame.

Shame and self-compassion

Self-compassion may provide a means to reduce overly self-critical responses

to negative events (Neff, 2003a, 2011). Self-compassion consists of three elements

“being open to and moved by one’s own suffering, experiencing feelings of caring

and kindness toward oneself, taking an understanding, non-judgmental attitude

toward one’s inadequacies and failures, and recognizing that one’s experience is part

of the common human experience” (Neff, 2003a p. 224). These three elements of

self-compassion are referred to as mindful awareness of feelings, self-kindness and

an acceptance of common humanity respectively.

Johnson and O’Brien (2013) argue that self-compassion can reduce the

experience of shame in response to negative events. They propose that each of the

components of self-compassion can reduce specific components of shame.

Specifically, being kind to oneself can soften self-criticism and negative self-

evaluation, acceptance of one’s common humanity can reduce the perceived need for

behavioral withdrawal and social isolation, and mindful awareness of feelings can

decrease the tendency to avoid or suppress emotional reactions to negative events.

Similarly it is plausible that being kind to oneself could soften the negative self-

evaluations present in guilt and acceptance of one’s common humanity may serve to

reduce feelings of guilt.

Consistent with this conceptualization, there is evidence that self-compassion

is negatively correlated with shame (Barnard & Curry, 2012), and therapeutic

interventions that promote self-compassion have been shown to reduce shame

17

(Gilbert & Procter, 2006; Judge, Cleghorn, McEwan & Gilbert, 2012). However,

these interventions are time-consuming and costly and use a combination of tasks

(e.g., compassionate writing, compassionate imagery and mindfulness) to create all

three elements of self-compassion. Consequently research has begun to consider the

effect of briefer self-compassionate interventions.

One promising intervention is the use of self-compassionate writing. Self-

compassionate writing has been shown to reduce negative affect (compared to a self-

esteem or control writing task) following a negative event (Leary, Tate, Adams,

Allen & Hancock, 2007) and three doses of self-compassionate writing have been

shown to reduce state shame (compared to expressive writing) about a shameful

event (Johnson & O’Brien (2013). These studies suggest that self-compassionate

writing could reduce the experience of shame and possibly guilt. However, this

remains open question because Johnson and O’Brien’s study failed to include both a

control writing condition and a self-esteem condition, thus the effects of the self-

compassionate writing task cannot be distinguished from the effect of writing, or

from the effect of raising self-esteem. Whilst self-esteem can be thought of as

positive feelings about the self and feeling valued by others, (Leary & MacDonald,

2003), self-compassion can be thought of as an orientation to care for oneself (Leary

et al., 2007). Therefore the current study focuses on whether a self-compassionate

writing intervention can produce lower levels of shame than a self-esteem

intervention and a control writing condition.

It is predicted that a brief self-compassionate writing exercise will lead to

lower levels of shame than a writing control task, and a self-esteem writing task. It is

possible that self-compassionate writing will have a similar effect on other self-

conscious emotions producing lower levels of guilt and higher levels of pride than a

18

writing control task.

Overview of study

A flowchart summarising the procedure can be seen in Appendix B. The

study was completed in one sitting by the participant. The use of Part 1 and Part 2

when describing the study is to aid clarity by separating out the aspects of the study

that addressed the two research questions: (1) to explore the association between

maladaptive perfectionism, shame, guilt and pride following imagined failure and (2)

to assess whether a writing task manipulation will have a differential effect on

shame, guilt and pride following imagined failure. Part 1 of the study addresses

question 1 and Part 2 of the study addresses question 2.

Method

Ethical Approval

The project was submitted for ethical review to the University of Surrey

Faculty of Arts and Human Sciences ethics committee and this project received a

favourable ethical opinion (see Appendix C for proof of review and approval).

Study Design

The study was a randomised control study. Following exposure to an

imagined failure task the study assessed the effect of three different writing tasks

(control, self-esteem and self-compassion) on shame, guilt and pride. Shame, guilt

and pride were measured twice, once following the imagined failure task and once

following the writing task.

Procedure

All of the measures and materials used in this study can be seen in full in

Appendix D. The study was hosted via an online portal, affiliated to the University

of Surrey, which provided a URL that was accessible from any internet-connected

19

computer. Participants completed all of the study online at this URL.

The online portal had a fixed design template for the screen- this had a white

background with a University of Surrey symbol at the top right corner. All of the text

used was black in colour and was presented in the Arial font at size 14 point. All

items that participants were asked to rate were presented with the item text centred in

the middle of the rating scale. Each individual measure or manipulation was

presented on a new screen and was followed at the bottom by a button to click to

move to the next screen unless otherwise specified.

The first screen provided participants with introductory information that

explained the purpose of the study, the payment for the study, and how data would

be used and stored. Participants were asked to read this information and then provide

consent to complete the study. Participants were asked to click yes to complete the

study, or to close the window if they did not consent to complete the study.

If participants consented to complete the study a new screen appeared that

displayed the following text “The following items are designed to measure attitudes

people have toward themselves, their performance, and toward others. There are no

right or wrong answers. Please try to respond to all of the items. Use your first

impression and do not spend too much time on individual items in responding.

Respond to each of the items using the scale below to describe your degree

of agreement with each item.” Participants were then provided with all of the items

from the measure of maladaptive perfectionism in a random order on one screen.

Participants rated their agreement by clicking the radio button to indicate their

selection.

All participants were then asked to complete the manipulation of failure.

For a perfectionist failure is defined as a failure to meet their own self-defined

20

standards. Therefore an imagery task was used to create a manipulation where each

participant could experience a failure to meet their own self-defined standards. There

is good evidence that imagery can create similar emotional and motivational

responses to those of real experiences (Dadds, Bovbjerg, Redd, & Cuttmore, 1997)

and imagery and re-scripting is an important aspect of clinical assessment and

intervention (Hackmann, Bennett-Levy, & Holmes, 2011).

The imagery task was a modified version of an imagery task used by Gilbert,

Baldwin, Irons, Baccus, and Palmer (2006). Participants were shown the following

instructions on screen “The following task is timed and you will not be able to

progress to the next page until the time allowed has passed.” Participants then read

the instructions that were designed to create the self-defined standards of

performance “We would like you to imagine a piece of coursework you have been

working on and putting a lot of effort into as part of your university studies. Imagine

the grade you want to get and how you would feel if you achieved it . . . Close your

eyes or look down and try to imagine this for 30 seconds." This screen was presented

for 50 seconds, allowing participants 20 seconds to read the instructions and 30

seconds to complete the task. After the 50 seconds had elapsed a button appeared on

screen so that participants could click to progress to the next screen. This screen

provided the manipulation of failing to meet these standards. Participants were

provided with the following written instructions “Now imagine that when your essay

is marked and comes back to you, the grade it is given is two grades lower than you

wanted. Close your eyes or look down and try to imagine this for 1 minute." below

these instructions there were three written prompts of what to think about during this

minute (e.g., What would this lower grade mean to you?). This screen was displayed

for 80 seconds, allowing 20 seconds to read the instructions and one minute to

21

complete them. After this time has elapsed a button appeared on screen to allow

participants to move to the next screen.

All participants were presented with the following text to read “The

following are some statements which may or may not describe how you are feeling

right now, following imagining the scenario. Please rate each statement using the 5-

point scale below. Remember to rate each statement based on how you are feeling

right at this moment.”. Participants were then presented with all items from the

measure of state shame, guilt and pride in a random order, on one screen and again

clicked a radio button to indicate their agreement.

The failure manipulation had been conducted using imagery so it was

difficult to objectively determine the extent to which participants had completed this

task. An adapted version of Gilbert et al. (2006) imagery manipulation check was

used to assess this. Participants saw the following text on screen “Please answer the

items below by clicking your answer on the scale”. Participants were asked to rate

their agreement with three questions to measure imageability of the failure scenario:

for example, “how easy was it to imagine the situation?”, on a ten-point scale

anchored at 1 (not easy to imagine) and at 10 (easy to imagine). The three items had

a Cronbachs alpha = .87. These three items were averaged together and the range of

possible scores was from 1-10. Participants were also asked to complete an item that

assessed the percentage of allotted time spent imagining; “how much of the allotted

time did you spend thinking about this situation?” anchored at 0% (none of the time)

and at 100% (all of the time).

All participants were then randomly assigned by the online software to one of

three writing tasks: writing control, self-esteem, or self-compassion. These writing

tasks were based on those used by Leary et al., (2007). The self-esteem and self-

22

compassion conditions had minimal changes from Leary’s self-esteem and self-

compassion writing tasks. Leary’s writing control condition was based on

Pennebaker, Colder, and Sharp’s (1990) task that encouraged self-disclosure of

emotion but this task had been shown to change how people feel about an event. In

this study the aim was to create an emotionally neutral control task that would

control for the process of writing, without changing the participants’ emotions.

Therefore participants were asked to describe buildings on their university campus,

write about their university, and being a student at their university.

The following text was then displayed to all participants “In the next task we

would like you to continue to think about the assignment that you did not do as well

on as you had anticipated. We would like you to write about this by answering the

questions below by typing in to the text boxes provided.” The remainder of the

information on this screen diverged dependent on writing task.

Each writing task asked participants to respond to three questions. For each

questions there was a specified minimum number of characters required. The

character required minimums were set based on pre-testing of these questions. The

character required minimums were the same across the three conditions, the first

question was set at 150 characters, the second question and third questions were set

at 400 characters. At the bottom of these questions participants were presented with a

button to click to proceed to the next screen. If participants had not typed enough to

meet the minimum threshold they were prompted to provide more detail in their

answer to that question.

The next screen was a repetition of the state shame, guilt and pride measure.

This screen was presented in the same way as the first time the measure was

presented but the items were presented in a random order each time, for each

23

participant.

Participants were presented with the following information on the next screen

“We would be grateful if you would complete a few final questions about this

study.” They were asked “What do you think this study was about?” which they

completed by typing into a text box and “Were you suspicious at any point that the

study was looking at something other than what was stated?” and were asked to click

one of three options “Not at all”, “A little”, or “ A lot”. All participants were then

asked to provide demographic information and to read the debriefing information.

Dependant measures

Maladaptive Perfectionism.

Maladaptive perfectionism was measured using the discrepancy subscale of

the APS-R (Slaney et al, 2001). The 23-item scale has three subscales: discrepancy,

high standards, and order. The APS-R is the only perfectionism scale designed to

capture the adaptive elements of perfectionism in one scale (high standards) and the

maladaptive components of perfectionism in a separate scale (discrepancy). The

APS-R has acceptable construct, concurrent, convergent and discriminant validity

(Rice, Ashby & Slaney, 2007). The three factors have been shown to be reliable with

a confirmatory factor analysis producing Cronbachs alphas that ranged from .85- .92

for the subscales (Slaney et al., 2001) and test-retest reliability is good (Rice &

Aldea, 2006). In this sample Cronbachs alpha= .94.

The discrepancy (maladaptive perfectionism) subscale has 12 items (e.g.,

“Doing my best never seems to be enough”) that participants rated from 1 (strongly

disagree) to 7 (strongly agree). Scores on these items were averaged together (the

range of possible scores was from 1 to 7) and a higher mean indicated higher levels

of discrepancy (maladaptive perfectionism),

24

Shame, guilt, and pride.

Shame, guilt, and pride were measured using the State Shame and Guilt Scale

(SGSS: Marschall, Sanftner, & Tangney, 1994). Robins, Noftle, and Tracey (2007)

note that the subscales have good Cronbachs alphas; shame (α = .89), guilt (α = .82),

pride (α = .87). No other psychometric information is available however, this

measure is often used in the published literature (Fedewa et al., 2005; Stöber et al.,

2007). In this sample Cronbachs alpha for shame, guilt and pride at Time 1

= .87, .88, and .92 respectively and at Time 2= .90, .87, and .92 respectively.

The SGSS has 15 items, (five items measure shame, guilt, and pride) which

participants rated from 1 (not feeling this way at all) to 5 (feeling this way strongly).

Example items for shame, guilt, and pride respectively are as follows “I want to sink

into the floor and disappear”, “I feel remorse, regret”, and “I feel good about

myself”. For each of the subscales the five item scores were averaged together (the

range of possible scores was from 1 to 5) and higher scores represented higher state

levels of that emotion.

Participants

One hundred and twenty-nine student participants were recruited from the

University of Surrey. The study was advertised to undergraduate and postgraduate

psychology students via the School of Psychology research participant recruitment

system. Psychology students were offered either research credit or entry to a prize-

draw for £50 (first prize) or £25 (two runner-up prizes) as payment for their

participation. The study was also advertised to other University of Surrey students,

by emailing the secretaries of other University of Surrey departments and asking

them to circulate an advertisement for the study. Participants recruited via the latter

method were offered entry to the prize draw only because they were not eligible for

25

research credit.

Twenty-eight participants were removed from the final dataset because they

had not completed the study. Eleven of these did not progress as far as completing

the first task (measure of perfectionism), one dropped out after the measure of

perfectionism, and the remaining 16 dropped out before completing the writing task.

The online nature of the study meant that it was difficult to determine whether

people had completed the study in one sitting, therefore, a further six participants

who took longer than two hours to complete the study were excluded, leaving N=95.

A power analysis conducted using G*Power version 3.1.9.2 (Faul, Erdfelder, Lang,

& Buchner, 2007) suggested that a sample size of 158 would be required to detect

the proposed main effect in ANCOVA if it had a medium effect size. Thus it is

possible that this study was underpowered.

It was plausible that individuals who were more perfectionistic may have

been more likely to dropout of the study because they did not want to imagine

failing, or may have been more likely to procrastinate and been removed from the

sample for taking too long to complete the study. In order to test this possibility

independent t-tests were computed. In order for t-tests to be valid the data needed to

be normally distributed on discrepancy (maladaptive perfectionism) for both groups,

and there should be homogeneity of variance across the two groups. Histograms on

discrepancy for the two groups with Z scores of skew, and kurtosis suggested that

both groups had an approximately normal distribution (see Appendix E) and the

Levine’s test suggested that there was a marginally non-significant difference in

variance across the two groups (F (1, 116)= 3.79, p = .05). Given this marginal result

the statistics for the adjusted t-test that does not assume equal variances are reported.

The t-test revealed no significant differences in discrepancy (maladaptive

26

perfectionism) between those who did not complete the study and those in the final

sample (t (47.36) = -1.10, p = .27, Mnon-completers= 4.01 vs. Mcompleters= 4.26). This

suggests that those who completed the study did not differ in how perfectionistic

they were compared to those who did not complete the study.

Results

Data handling

All data were collected by the online survey software Qualtrics. The online

study had been programmed to code the data such that any items were reverse-coded

as necessary. The data was downloaded from the online survey software and checked

for any errors. There were no errors and no missing data that needed to be addressed.

All variables were explored for outliers using boxplots and z-scores as described in

Field, (2009, p.99-103). For variables where significant outliers were identified the

outlying data-points were removed. This resulted in 2 outliers being removed from

the imageability scale.

Means were calculated for each scale/ subscale with items reversed as

necessary. The means, standard deviations, ranges and for all variables can be seen

in Table 1.

Table 1. Means, Standard Deviations, Ranges for all Variables.

Variable N Mean (SD) RangeDiscrepancy

(Maladaptive perfectionism)

95 4.26 (1.27) 5.75

Imageability (3 items) 93 8.06 (1.47) 9.00

Percentage time (1 item) 95 67.4 (21.0) 95.00

Time 1 shame 95 2.62 (1.08) 4.00

Time 2 shame 95 1.97 (0.96) 3.60

Time 1 guilt 95 2.79 (1.06) 4.00

27

Time 2 guilt 95 2.06 (0.90) 3.60

Time 1 pride 95 2.08 (0.85) 3.20

Time 2 pride 95 2.72 (1.00) 4.00

Analytic approach

All analyses were conducted in SPSS Statistics (Version 21, IBM

Corporation, 2012). A correlational analysis was conducted to consider whether

maladaptive perfectionism, was associated with shame, guilt, and pride (following

imagined failure). ANCOVA were conducted to consider whether following

imagined failure there was a significant effect of writing task on Time 2 levels of

shame, guilt and pride. Finally a content analysis was conducted to aid understanding

of the findings.

Correlational analysis

Pearson’s product-moment correlations were calculated to assess the extent

to which discrepancy, shame, guilt and pride and the manipulation check variables

percentage time and imageability were linearly associated. For significance tests of

correlation coefficients to be valid it is necessary that one of the variables has an

underlying Normal distribution (Altman 1991). Examination of histograms of the

variables and Z scores for skew and kurtosis (see Appendix F) suggests that there

was some deviation from Normality for Time 1 shame, Time 1 pride, imageability,

and percentage time, Time 2 guilt and Time 2 shame. Therefore Spearman's

correlation coefficients were also calculated. Comparison of Pearson's correlation

(Table 2) with Spearman's (Appendix G) shows that the pattern of results was similar

for both methods. The pairs of variables involved in the predicted

28

Table 2. Bivariate Pearson’s Correlations for Time 1 Variables.

Imageabili

ty

Percentage

time

Time 1

shame

Time 1

guilt

Time 1

pride

Time 2

shame

Time 2

guilt

Time 2

pride)

Discrepancy

(n=95)

r = .17

p = .10

r = .33**

p = .001

r = .48***

p < .001

r = .40***

p < .001

r = -.33**

p = .001

r = .50***

p < .001

r = .43***

p < .001

r = -.30**

p = .003

Imageability

(n=93)

r = .17

p = .09

r = .21

p = .05

r = .09

p =.39

r = -.17

p =.10

r = .10

p = .34

r = .13

p = .23

r = -.09

p = .41

Percentage time

(n=95)

r = .37***

p < .001

r = .32**

p = .002

r = -.16

p = .12

r = .26*

p = .01

r = .12

p = .25

r = .03

p = .78

Time 1 shame

(n=95)

r = .78***

p < .001

r = -.68***

p < .001

r = .63***

p < .001

r = .51***

p < .001

r = -.33**

p = .001

Time 1 guilt

(n=95)

r = -.49***

p < .001

r = .61***

p < .001

r = .65***

p < .001

r = -.22*

p = .03

Time 1 pride

(n=95)

r = -.44***

p < .001

r = -.30**

p = .003

r = .52***

p < .001

Time 2 shame

(n=95)

r = .83***

p < .001

r = -.58***

p < .001

Time 2 guilt

(n=95)

r = -.51***

p < .001

Note- Asterisks denote the significance level * p < .05, ** p < .01 and *** p <.001

29

correlations between discrepancy and Time 1 emotions met the assumptions for

Pearson’s correlations (Altman, 1991) because discrepancy appeared normally

distributed and so are reported below. There was a positive association between

discrepancy (maladaptive perfectionism) and Time 1 shame and between

discrepancy and Time 1 guilt. There was also a negative association between

discrepancy and Time 1 pride. These findings supported the initial hypotheses.

It was also important to consider whether the manipulation check measures

of imageability and percentage time spent imagining were associated with

maladaptive perfectionism. Discrepancy (maladaptive perfectionism) was not related

to the imageability of the failure scenario, however, there was a significant positive

relationship between discrepancy and estimated percentage of time spent thinking

about the imagined failure scenario. This suggests that those who endorsed

maladaptive perfectionism more strongly also reported spending more of the allotted

time thinking about the imagined failure scenario.

Effect of writing condition on Time 1 variables

Prior to conducting the ANCOVA it was important to consider whether there

was a significant effect of writing task on the Time 1 variables. Participants were

randomly allocated to writing task by the survey software (after completing all of the

Time 1 variables) so any differences would be due to chance. A series of 3 (writing

task: control vs. self-esteem vs. self-compassion) one-way ANOVAs were conducted

to assess this. In this study the group sizes are approximately equal and so ANOVA

is robust and accurate and can cope with deviations from normality and homogeneity

of variance (Field, 2013). However, these assumptions were still considered.

Histograms of the residuals (with Z scores for skew and kurtosis) and Levine’s test

for homogeneity of variance can be seen in Appendices H-I. These results show that

30

there was some lack of normality of distribution for imageability, Time 1 shame, and

Time 1 pride, and there was heterogeneity of variance for discrepancy. Although due

to equal group sizes ANOVA should be robust to these deviations, the analysis was

repeated (for those variables that showed wither a lack of normality or heterogeneity

of variance) using Kruskal-Wallis tests to check the robustness. The Kruskal-Wallis

tests can be seen in Appendix J and they do not materially differ from the one-way

ANOVA presented below.

The one-way ANOVAs revealed that there were no significant effects of

writing conditions on discrepancy (F (2, 92) = .10, p > .05, η2= .0021) imageability

(F (2, 90) = .15, p > .05, η2= .0033), percentage time spent on failure scenario (F (2,

92) = 1.37, p > .05, η2= .029), Time 1 shame (F (2, 92) = .06, p > .05, η2= .0013),

Time 1 guilt (F (2, 92) = .39, p > .05, η2= .0084), nor Time 1 pride (F (2, 92) = 1.57,

p > .05, η2= .033). The randomisation had not resulted in an unequal distribution of

characteristics across the three groups therefore the ANCOVA could be conducted as

planned.

Analysis of Covariance (ANCOVA) Analyses

The Time 1 and Time 2 means for shame, guilt and pride can be seen in

Figure 1, 2 and 3 respectively. These figures suggest that for all conditions the levels

of shame and guilt are lower and the levels of pride are higher at Time 2 (following

completion of writing tasks) than following completion of the imagery failure task.

31

Time 1 Time 2 (model 3)0

0.5

1

1.5

2

2.5

3

Control Self-esteem Self-compassion

Sham

e

Figure 1. Means for Shame as a function of writing condition at Time 1 and Time 2

(model 3) N=95.

Time 1 Time 2 (model 3)0

0.5

1

1.5

2

2.5

3

3.5

Control Self-esteem Self-compassion

Gui

lt

Figure 2. Means for guilt as a function of writing condition at Time 1 and Time 2

(model 3) N=95.

32

Time 1 Time 2 (model 3)0

0.5

1

1.5

2

2.5

3

3.5

Control Self-esteem Self-compassion

Prid

e

Figure 3- Means for pride as a function of writing condition at Time 1 and Time 2

(model 3) N=95.

ANCOVAs were conducted to assess whether there was a significant effect

of writing task on Time 2 emotion, after Time 1 emotion and discrepancy were

controlled for. Three separate ANCOVAs were conducted for shame, guilt and pride

respectively.

ANCOVA was chosen rather than examining change from baseline (by

conducting a 3 (writing task) x 2 (time) repeated measures ANOVA) because change

from baseline methods make more assumptions about the relationship between Time

1 and Time 2, which potentially introduces bias and reduces statistical power (Everitt

& Wessely, 2008 p. 93-96; Senn, 2006).

Due to the equal group sizes meeting the assumptions required for ANOVA

is not critical (Field, 2013). Despite this these assumptions were still considered and

the histograms of the residuals for ANCOVA with Z scores for skew and kurtosis

and the Levine’s tests can be seen in Appendices K-L. These results show that

normality assumptions were met for all variables. There was some heterogeneity of

33

variance for Time 2 guilt but given the equal group sizes the results are robust to this

(Field, 2013).

In each of the three ANCOVAs the covariate Time 1 emotion was

significantly related to Time 2 levels of emotion. Time 1 shame was significantly

related to Time 2 shame (F (2, 90) = 35.00, p < .001, r = .53, Time 1 guilt was

significantly related to Time 2 guilt (F (2, 90) = 52.05, p < .001, r = .61 and Time 1

pride was significantly related to Time 2 pride (F (2, 90) = 20.85, p < .001, r = .43

The covariate discrepancy was significantly related to Time 2 shame F (2,

90) = 8.18, p < .05, r = .29, and Time 2 guilt F (2, 90) = 5.79, p < .05, r = .25, but

not to Time 2 pride F (2, 90) = 3.05, p > .05, r = .18.

It was hypothesised that there would be a significant effect of writing task on

Time 2 emotions. Specifically it was predicted that the self-compassionate writing

would lead to lower levels of shame (and guilt and higher levels of pride) than the

writing control task, and the self-esteem writing task.

Contrary to these predictions there was no significant effect of writing task

on Time 2 shame after controlling for the effect of Time 1 shame and discrepancy (F

(2, 90) = 2.40, p > .05, partial η2= .05, Mcontrol = 1.80 Mself-esteem= 1.93, Mself-compassion=

2.17). There was a significant effect of writing task on Time 2 guilt after controlling

for the effect of Time 1 guilt and discrepancy (F (2, 90) = 4.24, p < .05, partial

η2= .09) and a significant effect of writing task on Time 2 pride after controlling for

the effect of Time 1 pride and discrepancy (F (2, 90) = 4.72, p < .05, partial η2= .09).

An examination of the means for Time 2 guilt and Time 2 pride (see Table 3)

suggests that whilst writing task does predict Time 2 emotions it does not predict

Time 2 emotions in the way that was hypothesised.

Post hoc tests (Bonferroni) were conducted to decompose the significant

34

effects of writing task on Time 2 emotions (see Table 3). There were significant

differences between the control condition and the self-compassion condition on both

pride (p < .01) and guilt (p < .05) such that pride is higher in the writing control

condition than the self-compassion condition and guilt is higher in the self-

compassion condition than the writing control condition. There was also a

marginally non-significant difference (p = .05) suggesting that guilt was also higher

in the self-compassion condition than the self-esteem writing condition. All other

comparisons failed to reach significance (p > .05).

Table 3. Means for Time 2 Emotion Variables where there was a Significant Effect

of Writing Condition for Post-hoc Analysis (N = 95).

Control Self-esteem Self-

compassion

Time 2 guilt 1.91a 1.92ab 2.32b

Time 2 pride 3.07a 2.70ab 2.44b

Note- Superscripts denote the post hoc Bonferrni tests. Within rows means that do

not share a superscript letter differ at p < .05.

Content Analysis

The effect of the writing tasks on pride, shame and guilt were not as

originally predicted. Therefore a content analysis was conducted on participant’s

responses to writing tasks. The answers to all three questions provided in the writing

task were considered as one text for each participant. The text was analysed using

Linguistic Inquiry and Word Count (LIWC). LIWC is a text analysis software

program designed by Pennebaker, Booth, and Francis (2007). LIWC analyses text on

a word-by-word basis and calculates the number of words people have used and the

degree to which people use different categories of words. Each word is compared

against a file of more than 2,000 reference words, divided into 72 categories. After

35

counting the number of words in each category, the output is given as a percentage

of the total words in the text sample. Newman, Pennebaker, Berry, and Richards

(2003) note that although computerized word count approaches are typically blind to

context they show promising and reliable results in personality, social, and clinical

psychology (Mergenthaler, 1996; Pennebaker et al., 2001; Rosenberg & Tucker,

1979; Stone, Dunphy, Smith, & Ogilvy, 1966). The dimensions captured by LIWC

have been used to predict social judgments (Berry, Pennebaker, Mueller, & Hiller,

1997), personality (Pennebaker & King, 1999), psychological adjustment (Rude,

Gortner, & Pennebaker, 2004), and health (Pennebaker, Mayne, & Francis, 1997).

The LIWC considers categories that fall under four superordinate category

areas; these are linguistic processes, psychological processes, personal concerns, and

spoken categories. Herein, 11 categories from across the first three superordinate

categories were selected on the basis of their potential to explain the findings. The

categories selected were (1) overall word count, (2) first-person pronouns singular,

(3) third-person pronouns singular, (4) third-person pronouns plural, (5) past-tense,

(6) present-tense, (7) future-tense, (8) positive emotions, (9) negative emotions, (10)

work and (11) achievement. The rationale for selecting these categories is outlined

below. Word count was chosen to determine whether the different writing tasks had

generated different amounts of written text. It was predicted that there would be no

differences in word count between writing tasks because they had been constructed

in similar ways to try and reduce this possibility.

It is possible that the way the failure imagery task was written about could

increase or decrease the extent to which participants experience self-conscious

emotions. Specifically writing about the imagined failure in the third rather than

first-person may decrease the extent to which the person feels shame or guilt about

36

their performance, indeed this is what the self-compassion condition encourages

(thinking about others who may have had similar experiences). Similarly the tense in

which the imagined failure is written about may also change the effect of the failure;

there may be less shame if the person has consigned the failure to the past, rather

than construing it in the present. The negative and positive emotion words were

considered because the frequency of these emotion words could have affected the

extent to which participants reported feeling shame, guilt or pride. One may have

expected higher levels of positive emotion words in the self-esteem condition.

Finally because the task failed was an academic assignment it is plausible that, in the

different writing conditions, people may have varied in the extent to which they

wrote about work and achievement, for example in the self-esteem condition people

may have talked about achievement more and work less than in other conditions.

LIWC calculated the frequency scores for these variables. Word count is a

total of words typed, however, for the other variables LIWC calculates a frequency

percentage by dividing the frequency of the occurrence of the category words by the

total number of words. The variables calculated from the content analysis were

explored for outliers using boxplots and Z-scores as described in Field, (2009, p.99-

103). The outlier analysis highlighted that a number of data-points from two

participants were outliers on a number of these variables. These participants’ data-

points were outliers because the percentage frequency scores were inflated due to

small total word counts (14 and 16 words). Therefore although these participants

responses were only identified outliers on particular variables they had the potential

to bias all variables calculated as percentage of total text, therefore these participants

data are only included in the analysis of the word count variable and are excluded

from the subsequent analysis of variables calculated as a percentage of total text.

37

Due to the equal group sizes meeting the assumptions for ANOVA is not

critical, but despite this the assumptions were still considered. The histograms of the

residuals with Z scores for skew and kurtosis and Levine’s tests can be seen in

Appendices M-N. Some of the histograms revealed a skew that was consistent with a

floor effect. This was particularly pronounced for the third-person singular category

where 91 out of 95 participants had not written any words of that category in their

text (a score of 0%). Therefore no further analysis was conducted on this variable.

There was evidence of non-norrmality and heterogeneity for most of the

variables so although results should be robust based on equal group sizes  (Field,

2013) the analysis was repeated using Kruskal-Wallis to check the robustness. The

results for the Kruskal-Wallis are shown in Appendix O and do not differ from the

ANOVA in terms of significance or pattern of means and mean ranks. The results of

the one-way ANOVA on writing task are summarised in Table 4.

Reviewing the statistics in Table 4, the number of words written (word

count) did not vary across writing conditions. This suggests that it is content of

participants writing rather than the amount of writing that influenced their guilt and

pride following the writing tasks. There was a significant main effect of writing

condition on the following LIWC variables first-person singular, third-person plural,

past-tense, future-tense, negative emotions, work and achievement words. Post-hoc

tests (Bonferroni) were conducted to decompose the significant main effects of

writing task on the LIWC variables (see Table 4.) Below I discuss only the LIWC

variables that differed significantly by writing condition.

38

Table 4. Means, Standard Deviations and Statistics for One-way ANOVAs investigating the effect of writing condition on LIWC variables

Dependant variable

Control

Means

Self-

esteem

Self-

compassion

SD F (DOF,

DOE)

p η2

Word count 128.83 134.14 118.40 53.95 0.71 (2, 92) > .05 .02

1st person singular (% of text) 2.14a 8.94 b 5.05 c 3.87 23.26 (2, 90) <. 001 .43

3rd person plural (% of text) 0.20 a 0.29 a 1.85 b 1.12 22.93 (2, 90) <. 001 .33

Past-tense (% of text) 0.50 a 4.79 b 5.60 b 2.40 41.45 (2, 90) <. 001 .50

Present-tense (% of text) 10.04 10.98 10.27 4.86 0.31 (2, 90) > .05 .05

Future-tense (% of text) 0.05 a 1.69 b 2.74 c 1.43 29.34 (2, 90) <. 001 .39

Positive emotions (% of text) 5.20 6.61 4.68 4.22 1.73 (2, 90) > .05 .02

Negative emotions (% of text) 0.76 a 2.23 b 5.12 c 1.97 42.09 (2, 90) <. 001 .50

Work (% of text) 8.88 a 6.04 b 5.26 b 3.12 11.93 (2, 90) <. 001 .25

Achievement (% of text) 2.23 a 5.66 b 4.96 b 3.15 10.19 (2, 90) <. 001 .19

Note- Superscripts denote the post-hoc Bonferrni tests. Within rows Means that do not share a superscript letter differ at p < .05.

39

The use of first-person singular words is highest in the self-esteem condition

and lowest in the control condition, with the self-compassion condition falling

between these two means and all of these differences are significant. The use of

third-person plural words is significantly higher in the self-compassion condition

than, in the self-esteem condition or in the control condition. It is notable that all of

the percentage scores for third-person plural words are low indicating that there is

little use of these types of words in the text produced in each condition. The use of

past-tense words was significantly lower in the control condition than in the self-

esteem and self-compassion conditions. The use of future-tense words was lowest in

the control condition and highest in the self-compassion condition, with the self-

esteem condition falling between these two conditions, all of these differences were

significant. It is notable that the use of future-tense words was relatively low across

conditions. The use of negative emotions words was much lower in the control

condition than in the self-compassion condition, with the self-esteem condition

falling between these two conditions, all of these differences were significant. The

use of words about work was significantly higher in the control condition than in the

self-esteem and self-compassion conditions, which did not differ from each other.

It is interesting to consider how the results of the analysis of the written text

fit with the effects of writing condition of self-conscious emotions. The means

suggested that Time 2 guilt and shame were lower than Time 1 guilt and shame and

that Time 2 pride was higher than Time 1 pride. The ANCOVA revealed that there

was a significant effect of writing condition on both Time 2 pride and Time 2 guilt.

Time 2 guilt was significantly higher in the self-compassion writing condition than

in the control condition and there was a trend towards guilt being higher in the self-

compassion than the self-esteem writing condition. Time 2 pride was higher in the

40

control condition than in the self-compassion condition and the self-compassion and

self-esteem conditions did not differ from each other.

Comparing the pattern of means on pride with the patterns of means on the

LIWC variables, the only variable that shows a similar pattern of means is the

“work” words variable. It is possible that the opportunity to describe their University

is related to this increase in pride. Comparing the patterns of means on guilt with the

patterns of means on the LIWC variables, the only variable where self-compassion

was clearly different from both self-esteem and control conditions (which were

broadly similar to each other) is on the third-person plural pronouns variable.

Discussion

This study aimed to address two separate research questions. The first

research question was to identify if there was a relationship between maladaptive

perfectionism and the experience of self-conscious emotions following an imagined

failure to meet self-defined standards for performance. The second research question

was to determine whether a self-compassionate writing task could reduce the

experience of shame and guilt and increase the experience of pride compared to self-

esteem or writing control tasks. Below I discuss the findings of this study in relation

to the research literature considering each of the research questions in turn.

Maladaptive perfectionism, and self-conscious emotions

It was hypothesised that those who endorsed maladaptive perfectionism

more strongly would also report more shame and guilt and less pride following

imagining failing to meet their self-defined standards on a piece of academic

coursework. Maladaptive perfectionism (discrepancy) was found to be positively

associated with Time 1 shame and Time 1 guilt and negatively associated with Time

1 pride. The findings support these hypotheses.

41

A number of studies have investigated the relationship between perfectionism

and self-conscious emotions (Fedewa et al, 2005; Klibert, et al, 2005; Tangney,

2002; Stöber et al, 2007) but as noted above the variation of in the way that

perfectionism and self-conscious emotions were measured makes these findings

difficult to interpret. However, there was some evidence that an individual’s level of

perfectionism was related to their proneness to experience shame and to their

experience of state shame and guilt.

The current findings appear to conceptually replicate the findings of Stöber et

al (2007) who demonstrated that ‘unhealthy’ perfectionists experienced more shame

and guilt and less pride than ‘healthy’ perfectionists. However, surprisingly

Steober’s unhealthy perfectionists did not differ in shame, guilt and pride from those

he defined as non-perfectionists. Comparing Stöber et al.’s findings to the current

findings may explain why. Stöber et al. used median splits to categorise their sample

on the basis of participants’ ratings on both high standards and discrepancy subscales

to create the three groups. “Unhealthy perfectionists” were those with high scores on

both the high standards and the discrepancy subscales of the APS-R, and “non-

perfectionists” were those with below median scores on high standards and any score

on discrepancy. Hence, in Stöber et al.’s study, unhealthy perfectionists and non-

perfectionists could have the same score on discrepancy. The current study argues

that variation in discrepancy alone represents the best measure of maladaptive/

unhealthy perfectionism. By categorising participants using their ratings on the

subscale for high standards, Stöber et al. implicitly assume that what represents high

standards is the same for everyone. However, individuals vary in their level of ability

to complete various tasks and so the same objective “high standards” could be

achievable and realistic for one person and excessively high for another individual.

42

This highlights one of the strengths of the current study, which was to ensure that the

standards that participants failed to meet were self-defined standards for

performance.

Interestingly the findings from the first part of the study suggest that those

participants who endorsed discrepancy more strongly also reported that they spent a

greater percentage of the allocated time thinking about the failure scenario.

Assuming that these self-reports are accurate, this hints at several possibilities. First,

that those participants who report spending longer on the failure task experience

more shame and guilt and less pride (regardless of level of perfectionism), second

that maladaptive perfectionists focus more on evidence that they have failed to meet

their standards, or third, both of the above. Disentangling these explanations

represents an interesting question for future study.

Self-compassion and self-conscious emotions

The second part of the study tested two hypotheses: first, whether a self-

compassionate writing intervention reduced state shame and guilt and increased state

pride relative to a self-esteem and control writing task. The means suggested that all

conditions lead to reduced shame and guilt and increased pride. However, none of

these hypotheses about the differential effects of writing condition were supported.

Levels of shame did not significantly vary across writing condition. Although this

study may have been underpowered there was variability in the means for guilt and

pride, so it appears that this may only have been problematic for shame. Levels of

guilt and pride did vary across writing condition but this variation did not support the

predictions. Higher levels of pride were found in the control condition compared to

the self-compassion condition, and higher levels of guilt were found in the self-

compassion condition compared to the control condition.

43

An exploratory content analysis was conducted using LIWC to consider what

might underlie these differences. These analyses revealed that the length of written

content did not vary across writing conditions. However, individuals who completed

the self-esteem writing task used more first-person singular words than those who

completed other writing tasks. Individuals who completed the self-compassion

writing task used more third-person plural words that those who completed other

writing tasks. These findings suggest that the task instructions were followed

because the self-esteem condition promoted a focus on the self and the self-

compassion condition a focus on common humanity. Similarly there was evidence

that there was a lower frequency of past-tense words and future tense words used in

the control condition than in the other two writing conditions. This most likely

reflects the nature of the control task. Furthermore the frequency of both past- and

future-tense words was highest in the self-compassion condition suggesting that

these participants were locating this experience amongst their own and others’

similar previous experiences and anticipated future experiences. The frequency of

achievement words is lower in the control condition that the other writing conditions

This is likely because in the control condition individuals are describing their

university whereas the other writing tasks ask individuals to write about their

response to the failure scenario.

Comparing the pattern of means and post-hoc tests on guilt and pride with the

LIWC variables suggested that the pattern of means for Time 2 pride was similar to

the pattern of means for frequency of words about work. Participants were asked to

write about their University in the control condition and so higher frequency of

words about work and higher levels of pride makes intuitive sense. Whist

participants were asked to describe the campus, list the buildings, and describe what

44

it was like to be a student, an examination of the written text reveals that some

students used the task to describe their pride at being a University of Surrey student.

It is notable that during the period the study was conducted the University of Surrey

was ranked 6th in the Guardian league table and this was heavily publicised on the

University campus. The Social Identity approach (Tajfel & Turner, 1979; Turner,

Hogg, Oakes, Reicher & Wetherall, 1987) suggests that individuals will consider

themselves in terms of their most salient identity. For participants who had just been

asked to imagine failing a University assignment and who were subsequently asked

to describe their University, their identity as a University student was particularly

salient. Furthermore, identifying with this group identity provides a higher status and

represents an adaptive means of raising self-esteem - whilst they as an individual

have failed, they will succeed as a University of Surrey student.

The findings suggested that guilt scores were higher following self-

compassionate writing than following control writing (although an examination of

the means suggests that both means were lower than the scores at Time 1). Similarly,

there was evidence of a higher frequency of negative emotion words in the self-

compassion condition than in the control condition. This finding could be explained

by the nature of the self-compassionate writing task instructions. Indeed, Neff

(2003b) suggests that self-compassion creates coping by encouraging individuals to

approach their emotions, rather than avoid their emotions. Thus it may be that those

who completed self-compassionate writing tasks are experiencing more negative

emotions. This argument suggests that an increased frequency of negative emotions

words should also be accompanied by higher levels of shame following self-

compassionate writing. Indeed, a consideration of the Time 2 means for shame

suggests a similar albeit non-significant pattern. This may be attributable to a lack of

45

power. Considering the self-compassion task instructions may also provide an

explanation for why the reduction of shame is not impaired to the same extent as

guilt. The instructions focussed on the imagined failure of the task, and so the

negative emotions were about a particular event rather than about the individual

person. Thus the task instructions may have been more likely to impair reduction in

guilt, because guilt is often about an action, whereas shame is about the person

(Lewis, 1971).

Limitations and suggestions for future research

The current study has a number of potential limitations. The hypotheses were

tested in a specific context, which was that of academic performance and imagined

failure. This represents only one domain in which individuals may express their

perfectionism, and one type of failure within this domain. Similarly, self-compassion

was manipulated using only one approach, that of writing. In clinical interventions a

variety of methods are used including imagery, compassionate letter writing, and

compassionate mindfulness. It is possible that the use of different methods may have

led to different results. The measure of state shame, guilt and pride (Marschall et al.,

1997) was used because it is commonly used within the literature and was the only

measure of state shame, guilt and pride that was available. In this study a non-

clinical population (university students) was used. All ratings of self-conscious

emotions were around or below the midpoint of the scale and so perhaps there was

not a great deal of guilt and shame to reduce. It is currently unclear to what extent

self-compassionate writing is a useful strategy for individuals from non-clinical

population. It may be that self-compassionate approaches are more effective than

other approaches such as self-esteem and control writing in clinical populations only.

One strength of the study was that the failure that participants were exposed

46

to was imagined. Whilst this could be considered a weakness compared to placing

participants in a manipulated objective failure scenario, there is good evidence that

imagery can create similar emotional and motivational responses to those of real

experiences (Dadds, et al, 1997) and indeed imagery and re-scripting is an important

aspect of clinical assessment and intervention (Hackmann, et al, 2011). Furthermore

it was the use of imagery that allowed participants to individually self-define the

standards that they expected to meet. It is failure to meet one’s own standards that is

central to the definition of perfectionism and so with this in mind, it seems that using

a task that allowed for failure to meet self-defined standards represented a strength of

this approach.

This study represent a first step in examining the relationship between self-

compassion, shame, guilt and pride. The findings would be strengthened by

increasing their generalizability by considering alternative domains of perfectionism,

different control conditions, different manipulations of failure and self-compassion

and different measures of self-conscious emotions. It would also be important to

clearly define the populations who may benefit from self-compassionate writing

interventions.

Future research should consider whether a self-compassionate writing task

reduces state shame, when compared to self-esteem, expressive writing and writing

control tasks, and whether any reductions in state shame are accompanied by

reductions in shame-proneness. Research should also consider whether repeated

exposures to self-compassionate writing would increase the usefulness of the self-

compassionate writing task in improving self-conscious emotions following failure,

and whether there is an optimal number of exposures required to create this effect.

47

References

Altman, D. G. (1991). Practical statistics for medical research. London: Chapman

and Hall.

Barnard, L. K., & Curry, J. F. (2012). The relationship of clergy burnout to self-

compassion and other personality dimensions. Pastoral Psychology, 61, 149–

163. doi: 10.1007/s11089-011-0377-0

Berry, D. S., Pennebaker, J. W., Mueller, J. S., & Hiller, W. S. (1997). Linguistic

bases of social perception. Personality and Social Psychology Bulletin, 23,

526-537. doi: 10.1177/0146167297235008

Bieling, P. J., Israeli, A. L., & Anthony, M. M. (2004). Is perfectionism good, bad,

or both? Examining models of the perfectionism construct. Personality and

Individual Differences, 36, 1373-1385. doi: 10.1016%2FS0191-

8869%2803%2900235-6

Dadds, M. R., Bovbjerg, D. H., Redd, W. H., & Cutmore, T. R. (1997). Imagery in

human classical conditioning. Psychological Bulletin, 122, 89 –103. doi:

10.1037//0033-2909.122.1.89

Dunkley, D. M., Blankstein, K. R., & Berg, J. (2012). Perfectionism dimensions and

the five-factor model of personality. European Journal of Personality, 26,

233-244. doi: 10.1002/per.829

Egan, S. J., Wade, T. D., & Shafran, R. (2011). Perfectionism as a transdiagnostic

process: A clinical review. Clinical Psychology Review, 31, 203–212. doi:

0.1016/j.cpr.2010.04.009.

Enns, M.W., Cox, B. J. & Clara, (2002). Adaptive and maladaptive perfectionism:

developmental origins and association with depression proneness.

Personality and Individual Differences 33, 921–935. doi:10.1016%2FS0191-

48

8869%2801%2900202-1

Everitt , B. S., & Wessely, S. (2008). Clinical Trials in Psychiatry (2nd edition).

Chicester, West Sussex, UK: John Wiley & Sons.

Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: A flexible

statistical power analysis program for the social, behavioral, and biomedical

sciences. Behavior Research Methods, 39, 175-191.   

Fedewa, B. A., Burns, L. R., & Gomez, A. A. (2005). Positive and negative

perfectionism and the shame/guilt distinction: Adaptive and maladaptive

characteristics. Personality and Individual Differences, 38, 1609-1619. doi:

10.1016/j.paid.2004.09.026.

Field, A. (2005). Discovering statistics using IBM SPSS statistics (2nd Edition).

London, UK: Sage.

Field, A. (2013). Discovering statistics using IBM SPSS statistics (4th Edition).

London, UK: Sage.

Frost, R. O., Marten, P., Lahart, C. M., & Rosenblate, R. (1990). The dimensions of

perfectionism. Cognitive Therapy and Research, 14, 449–468. doi:

10.1007%2FBF01172967

Gilbert, P. (1998). What is shame? Some core issues and controversies. In, P. Gilbert

& B. Andrews, (eds) Shame: Interpersonal behavior, psychopathology and

culture. (pp 3-38). New York: Oxford University Press.

Gilbert, P. (2003). Evolution, social roles, and differences in shame and guilt. Social

Research: An International Quarterly of the Social Sciences 70, 1205-1230.

doi: not available.

Gilbert, P., Baldwin, M., Irons, C., Baccus, J., & Clark, M. (2006). Self-criticism and

self-warmth: An imagery study exploring their relation to depression. Journal

49

of Cognitive Psychotherapy, 20, 183–200. doi: 10.1891%2Fjcop.20.2.183.

Gilbert, P., Baldwin, M. W., Irons, C., Baccus, J. R., & Palmer, M. (2006). Self-

criticism and self-warmth: An imagery study exploring their relation to

depression. Journal of Cognitive Psychotherapy: An International Quarterly,

20 (2),183-200. doi: http://dx.doi.org/10.1891/jcop.20.2.183.

Gilbert, P., & Procter, S. (2006). Compassionate mind training for people with high

shame and self-criticism: A pilot study of a group therapy approach. Clinical

Psychology and Psychotherapy, 13, 353–379. doi: 10.1002/cpp.507

Hackmann, A., Bennett-Levy, J., & Holmes, E. A. (2011). Oxford Guide to Imagery

in Cognitive Therapy. OUP: Oxford.

Hamachek, D. E. (1978). Psychodynamics of normal and neurotic perfectionism.

Psychology, 15(1), 27-33. doi: not available.

Hewitt, P. L., & Flett, G. L. (1991). Perfectionism in the self and social contexts:

Conceptualization, assessment, and association with psychopathology.

Journal of Personality and Social Psychology, 60, 456–470. doi:10.1037%2F

%2F0022-3514.60.3.456.

IBM Corporation. (2012). IBM SPSS Statistics for Windows, Version 21.0.

[Computer software] Armonk, NY: IBM Corp.

Johnson, E. A., & O’Brien, K. A. (2013). Self-compassion soothes the savage ego-

threat system: Effects on negative affect, shame, rumination and depressive

symptoms. Journal of Social and Clinical Psychology, 32(9), 939-963.

doi:10.1521/jscp.2013.32.9.939.

Judge, L., Cleghorn, A., McEwan, K., & Gilbert, P. (2012). An exploration of group-

based compassion focused therapy for a heterogeneous range of clients

presenting to a Community Mental Health Team. International Journal of

50

Cognitive Therapy, 5(4), 420–429, doi: 10.1521%2Fijct.2012.5.4.420

Klibert, J. J., Langhinrichsen-Rohling, J., & Saito, M. (2005). Adaptive and

maladaptive aspects of self-oriented versus socially prescribed perfectionism.

Journal of College Student Development, 46, 141–156. doi:

10.1353%2Fcsd.2005.0017,

Leary, M. R., Tate, E. B., Adams, C. E., Allen, A. B., & Hancock, J. (2007). Self-

compassion and reactions to unpleasant self-relevant events: The implications

of treating oneself kindly. Journal of Personality and Social Psychology, 92,

887–904. doi:10.1037/0022-3514.92.5.887.

Leary, M. R., & MacDonald, G. (2003). Individual differences in self- esteem: A

review and theoretical integration. In M. R. Leary & J. P. Tangney (Eds.),

Handbook of self and identity (pp. 401–418). New York: Guilford Press.

Lewis, H. B. (1971). Shame and guilt in neurosis. New York: International

Universities Press.

Mascolo, M. F., & Fischer, K. W. (1995). Developmental transformations in

appraisals for pride, guilt and shame. In Fischer, K. W. & Tangney, J. P.

(Eds.), Self-Conscious Emotions: Shame, Guilt, Embarrassment and Pride

(pp. 64-113). New York: Guilford.Mas

Marschall, D. E., Saftner, J., & Tangney, J. P. (1994). The State Shame and Guilt

Scale. George Mason University, Fairfax, VA.

Mergenthaler, E. (1996). Emotion-abstraction patterns in verbatim protocols: A new

way of describing psychotherapeutic processes. Journal of Consulting and

Clinical Psychology, 64, 1306-1315. doi:10.1037//0022-006X.64.6.1306

Miller, S. B. (1996). Shame in context. London: The Analytic Press.

Neff, K. D. (2003a). The development and validation of a scale to measure self-

51

compassion. Self & Identity, 2, 223–250. doi:10.1080/15298860390209035.

Neff, K. D. (2003b). Self-compassion: An alternative conceptualization of a healthy

attitude toward oneself. Self and Identity, 2, 85–102.

doi:10.1080/15298860390129863.

Neff, K. D. (2011). Self-Compassion. New York: HarperCollins.

Newman, M. L., Pennebaker, J. W., Berry, D. S., & Richards, J. M. (2003). Lying

words: Predicting deception from linguistic styles. Personality and Social

Psychology Bulletin, 29(5), 665-675. doi: 10.1177/0146167203251529.

Pennebaker, J.W., Booth, R.J., & Francis, M.E. (2007). Linguistic Inquiry and Word

Count: LIWC 2007. Austin, TX: LIWC (www.liwc.net).

Pennebaker, J. W., Colder, M., & Sharp, L. K. (1990). Accelerating the coping

process. Journal of Personality and Social Psychology, 58, 528–537.

doi:10.1037/0022- 3514.58.3.528.

Pennebaker, J. W., & King, L. A. (1999). Linguistic styles: Language use as an

individual difference. Journal of Personality and Social Psychology, 6, 1296-

1312. doi: 0.1037//0022-3514.77.6.1296

Pennebaker, J. W., Mayne, T. J., & Francis, M. E. (1997). Linguistic predictors of

adaptive bereavement. Journal of Personality and Social Psychology, 72,

863-871. doi: 10.1037//0022-3514.72.4.863

Pennebaker, J. W., Francis, M. E., & Booth, R. J. (2001). Linguistic Inquiry and

Word Count: LIWC 2001. Mahwah, NJ: Lawrence Erlbaum.

Reimer, M. S. (1996). “Sinking into the ground”: The developmental consequences

of shame in adolescence. Developmental Review, 16, 321-363.

doi:10.1006/drev.1996.0015

Rice, K. G., & Aldea, M. A. (2006). State dependence and trait stability of

52

perfectionism: A short-term longitudinal study. Journal of Counseling

Psychology, 53, 205-213. doi: 10.1037%2F0022-0167.53.2.205

Rice, K. G., Ashby, J. S., & Slaney, R. B. (1998). Self-esteem as a mediator between

perfectionism and depression: A structural equation analysis. Journal of

Counseling Psychology, 45, 304-314. doi: 10.1037%2F%2F0022-

0167.45.3.304

Rice, K. G., Ashby, J. S., & Slaney, R. B. (2007). Perfectionism and the five-factor

model of personality. Assessment, 14, 385-398. doi:

10.1177/1073191107303217

Robins, R. W., Noftle, E. E., & Tracey, J. L. (2007). Assessing Self-Conscious

Emotions: A Review of Self-Report and Nonverbal Measures (pp. 443-469).

In Tracy, J. L., Robins, R. W., & Tangney, J. M. (Eds.) The Self-Conscious

Emotions: Theory and Research. New York: Guilford.

Rosenberg, S. D., & Tucker, G. J. (1979). Verbal behavior and schizophrenia: The

semantic dimension. Archives of General Psychiatry, 36, 1331-1337.

doi:10.1001/archpsyc.1979.01780120061008

Rude, S.S., Gortner, E.M., & Pennebaker, J.W. (2004). Language use of depressed

and depression- vulnerable college students. Cognition and Emotion, 18,

1121-1133. doi:10.1080/02699930441000030

Senn, S. (2006). Change from baseline and analysis of covariance revisited. Statistics

in Medicine, 25, 4334-4344 . doi:10.1002/sim.2682

Shafran, R., Egan, S., & Wade, T. (2010). Overcoming perfectionism. London:

Constable and Robinson.

Slaney, R. B., Rice, K. G., Mobley, M., Trippi, J., & Ashby, J. S. (2001). The

revised Almost Perfect Scale. Measurement and Evaluation in Counseling

53

and Development, 34, 130–145. doi: not available.

Sorotzkin, B. (1985). The quest for perfection: Avoiding guilt or avoiding shame?

Psychotherapy, 22, 564–571. DOI: 10.1037/h0085541

Stöber, J., Harris, R. A., & Moon, P. S. (2007). Perfectionism and the experience of

pride, shame, and guilt: Comparing healthy perfectionists, unhealthy

perfectionists, and non-perfectionists. Personality and Individual Differences,

43, 131-141. doi: 10.1016/j.paid.2006.11.012.

Stöber, J., & Otto, K. (2006). Positive conceptions of perfectionism: Approaches,

evidence, challenges. Personality and Social Psychology Review, 10, 295–

319. doi: 10.1207%2Fs15327957pspr1004_2

Stone, P. J., Dunphy, D. C., Smith, M. S., & Ogilvy, D. M. (1966). The General

Inquirer: A computer approach to content analysis. Cambridge, MA: MIT

Press.

Suddarth, B. H., & Slaney, R. B. (2001). An investigation of the dimensions of

perfectionism in college students. Measurement and Evaluation in

Counseling and Development, 34, 157-165. doi: not available.

Tajfel, H. & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W.

G. Austin and S. Worchel (eds), The Social Psychology of Intergroup

Relations. Monterey, CA: Brooks/Cole, pp. 33-47.

Tangney, J. P. (2002). Perfectionism and the self-conscious emotions: Shame, guilt,

embarrassment, and pride. In P. L. Hewitt & G. L. Flett (Eds.),

Perfectionism: Theory, research, and treatment (pp. 199-215). Washington,

DC: American Psychological Association.

Tangney, J. P., & Dearing, R. L. (2002). Shame and guilt. New York: Guilford.

Turner, J. C., Hogg, M. A., Oakes P. J., Reicher S. D., & Wetherell, M. S. (1987).

54

Rediscovering the Social Group: A Self-Categorization Theory. Oxford and

New York: Basil Blackwell.

55

List of Appendices

Appendix A...............................................................................................................58

Instructions for authors- British Journal of Clinical Psychology..........................58

Appendix B................................................................................................................62

Summary of procedure...........................................................................................62

Appendix C...............................................................................................................63

Proof of ethical review...........................................................................................63

Appendix D...............................................................................................................64

Information sheet....................................................................................................64

Consent sheet..........................................................................................................66

Measure of Perfectionism.......................................................................................67

Failure imagery task..............................................................................................70

State Shame and Guilt scale...................................................................................71

Manipulation check................................................................................................73

Writing tasks...........................................................................................................74

Feedback and demographics..................................................................................75

Debrief....................................................................................................................77

Appendix E................................................................................................................78

Histograms for discrepancy (maladaptive perfectionism).....................................78

Appendix F................................................................................................................79

Histograms of variables for correlations...............................................................79

Appendix G...............................................................................................................84

56

Appendix H...............................................................................................................85

Histograms of residuals from one- way ANOVA’s.................................................85

Appendix I.................................................................................................................88

Appendix J................................................................................................................89

Appendix K...............................................................................................................90

Histograms of the residuals for all ANCOVAs on shame, guilt and pride.............90

Appendix L................................................................................................................92

Appendix M..............................................................................................................93

Histograms of residuals for ANOVA on LIWC variables......................................93

Appendix N...............................................................................................................99

Appendix O.............................................................................................................100

57

Appendix A

Instructions for authors- British Journal of Clinical Psychology

The British Journal of Clinical Psychology publishes original contributions to scientific knowledge in clinical psychology. This includes descriptive comparisons, as well as studies of the assessment, aetiology and treatment of people with a wide range of psychological problems in all age groups and settings. The level of analysis of studies ranges from biological influences on individual behaviour through to studies of psychological interventions and treatments on individuals, dyads, families and groups, to investigations of the relationships between explicitly social and psychological levels of analysis.The following types of paper are invited:• Papers reporting original empirical investigations• Theoretical papers, provided that these are sufficiently related to the empirical data• Review articles which need not be exhaustive but which should give an interpretation of the state of the research in a given field and, where appropriate, identify its clinical implications• Brief reports and comments

1. CirculationThe circulation of the Journal is worldwide. Papers are invited and encouraged from authors throughout the world.

2. LengthThe word limit for papers submitted for consideration to BJCP is 5000 words and any papers that are over this word limit will be returned to the authors. The word limit does not include the abstract, reference list, figures, or tables. Appendices however are included in the word limit. The Editors retain discretion to publish papers beyond this length in cases where the clear and concise expression of the scientific content requires greater length. In such a case, the authors should contact the Editors before submission of the paper.

3. Submission and reviewingAll manuscripts must be submitted via http://www.editorialmanager.com/bjcp/. The Journal operates a policy of anonymous peer review. Before submitting, please read the terms and conditions of submission and the declaration of competing interests.

4. Manuscript requirements• Contributions must be typed in double spacing with wide margins. All sheets must be numbered.• Manuscripts should be preceded by a title page which includes a full list of authors and their affiliations, as well as the corresponding author's contact details. A template can be downloaded from here.• Tables should be typed in double spacing, each on a separate page with a self-explanatory title. Tables should be comprehensible without reference to the text. They should be placed at the end of the manuscript with their approximate locations indicated in the text.• Figures can be included at the end of the document or attached as separate files,

58

carefully labelled in initial capital/lower case lettering with symbols in a form consistent with text use. Unnecessary background patterns, lines and shading should be avoided. Captions should be listed on a separate sheet. The resolution of digital images must be at least 300 dpi.• All papers must include a structured abstract of up to 250 words under the headings: Objectives, Methods, Results, Conclusions. Articles which report original scientific research should also include a heading 'Design' before 'Methods'. The 'Methods' section for systematic reviews and theoretical papers should include, as a minimum, a description of the methods the author(s) used to access the literature they drew upon. That is, the abstract should summarize the databases that were consulted and the search terms that were used.• All Articles must include Practitioner Points – these are 2–4 bullet points to detail the positive clinical implications of the work, with a further 2–4 bullet points outlining cautions or limitations of the study. They should be placed below the abstract, with the heading ‘Practitioner Points’.• For reference citations, please use APA style. Particular care should be taken to ensure that references are accurate and complete. Give all journal titles in full and provide DOI numbers where possible for journal articles.• SI units must be used for all measurements, rounded off to practical values if appropriate, with the imperial equivalent in parentheses.• In normal circumstances, effect size should be incorporated.• Authors are requested to avoid the use of sexist language.• Authors are responsible for acquiring written permission to publish lengthy quotations, illustrations, etc. for which they do not own copyright. For guidelines on editorial style, please consult the APA Publication Manual published by the American Psychological Association.

5. Brief reports and commentsThese allow publication of research studies and theoretical, critical or review comments with an essential contribution to make. They should be limited to 2000 words, including references. The abstract should not exceed 120 words and should be structured under these headings: Objective, Method, Results, Conclusions. There should be no more than one table or figure, which should only be included if it conveys information more efficiently than the text. Title, author name and address are not included in the word limit.

6. Supporting InformationBJC is happy to accept articles with supporting information supplied for online only publication. This may include appendices, supplementary figures, sound files, videoclips etc. These will be posted on Wiley Online Library with the article. The print version will have a note indicating that extra material is available online. Please indicate clearly on submission which material is for online only publication. Please note that extra online only material is published as supplied by the author in the same file format and is not copyedited or typeset. Further information about this service can be found at http://authorservices.wiley.com/bauthor/suppmat.asp

7. Copyright and licensesIf your paper is accepted, the author identified as the formal corresponding author for the paper will receive an email prompting them to login into Author Services, where

59

via the Wiley Author Licensing Service (WALS) they will be able to complete the license agreement on behalf of all authors on the paper.

For authors signing the copyright transfer agreementIf the OnlineOpen option is not selected the corresponding author will be presented with the copyright transfer agreement (CTA) to sign. The terms and conditions of the CTA can be previewed in the samples associated with the Copyright FAQs.

For authors choosing OnlineOpenIf the OnlineOpen option is selected the corresponding author will have a choice of the following Creative Commons License Open Access Agreements (OAA):

- Creative Commons Attribution Non-Commercial License OAA- Creative Commons Attribution Non-Commercial -NoDerivs License OAA

To preview the terms and conditions of these open access agreements please visit the Copyright FAQs and you may also like to visit the Wiley Open Access Copyright and Licence page.

If you select the OnlineOpen option and your research is funded by The Wellcome Trust and members of the Research Councils UK (RCUK) or the Austrian Science Fund (FWF) you will be given the opportunity to publish your article under a CC-BY license supporting you in complying with your Funder requirements. For more information on this policy and the Journal’s compliant self-archiving policy please visit our Funder Policy page.

8. Colour illustrationsColour illustrations can be accepted for publication online. These would be reproduced in greyscale in the print version. If authors would like these figures to be reproduced in colour in print at their expense they should request this by completing a Colour Work Agreement form upon acceptance of the paper. A copy of the Colour Work Agreement form can be downloaded here.

9. Pre-submission English-language editingAuthors for whom English is a second language may choose to have their manuscript professionally edited before submission to improve the English. A list of independent suppliers of editing services can be found at http://authorservices.wiley.com/bauthor/english_language.asp. All services are paid for and arranged by the author, and use of one of these services does not guarantee acceptance or preference for publication.

10. Author ServicesAuthor Services enables authors to track their article – once it has been accepted – through the production process to publication online and in print. Authors can check the status of their articles online and choose to receive automated e-mails at key stages of production. The author will receive an e-mail with a unique link that enables them to register and have their article automatically added to the system. Please ensure that a complete e-mail address is provided when submitting the manuscript. Visit http://authorservices.wiley.com/bauthor/ for more details on online

60

production tracking and for a wealth of resources including FAQs and tips on article preparation, submission and more.

11. The Later StagesThe corresponding author will receive an email alert containing a link to a web site. A working e-mail address must therefore be provided for the corresponding author. The proof can be downloaded as a PDF (portable document format) file from this site. Acrobat Reader will be required in order to read this file. This software can be downloaded (free of charge) from the following web site: http://www.adobe.com/products/acrobat/readstep2.html.

This will enable the file to be opened, read on screen and annotated direct in the PDF. Corrections can also be supplied by hard copy if preferred. Further instructions will be sent with the proof. Excessive changes made by the author in the proofs, excluding typesetting errors, will be charged separately.

12. Early ViewBritish Journal of Clinical Psychology is covered by the Early View service on Wiley Online Library. Early View articles are complete full-text articles published online in advance of their publication in a printed issue. Articles are therefore available as soon as they are ready, rather than having to wait for the next scheduled print issue. Early View articles are complete and final. They have been fully reviewed, revised and edited for publication, and the authors’ final corrections have been incorporated. Because they are in final form, no changes can be made after online publication. The nature of Early View articles means that they do not yet have volume, issue or page numbers, so they cannot be cited in the traditional way. They are cited using their Digital Object Identifier (DOI) with no volume and issue or pagination information. E.g., Jones, A.B. (2010). Human rights Issues. Human Rights Journal. Advance online publication. doi:10.1111/j.1467-9299.2010.00300.x

61

Part 1

Part 2

Appendix BSummary of procedure

Click on study URL

Information about study

Informed consent

Yes No

Close window do not complete study

Measure of perfectionism

Manipulation of failure imagery

T1 measure of state shame, guilt and pride

Manipulation check of failure imagery

Instructions for writing task

Control Self-esteem Self-compassion

T2 measure of state shame, guilt and pride

Feedback and demographics questions

Debrief

62

Appendix C

Proof of ethical review

A

p

pe

n

di

x

D

63

All measures and manipulations are presented below in the order they were completed by participants.

Information sheet

Principal InvestigatorNatalie Almond, Natalie.Almond@surrrey.ac.uk

Co-InvestigatorDr Laura Simonds, L.Simonds@surrey.ac.uk

Information sheet Project title: Writing, personality and appraisals

My name is Natalie Almond and I am a Trainee Clinical Psychologist, based in the Psychology department at the University of Surrey. This means that I am currently studying for a doctorate qualification in Psychology. As part of my training I have to complete a research project. I am researching how writing can influence the way that individuals feel and think, and I am particularly interested in how certain personality factors may influence the relationship between writing and appraisals such as thoughts and feelings. This study is important in helping us to consider whether writing used in psychological therapy is equally beneficial to people with different personality characteristics.

If you choose to complete this study you will be directed to an online study. You will be asked to complete four different questionnaires and to perform an imagery task and a writing task. In the imagery task you will be provided with some instructions and ask to try and picture an image, and hold it in mind before answering a number of questions about this image. It is important that you take time to complete the imagery task as described. In the writing task you will be provided with some prompts of what to write about.

This study will take approximately 30 minutes. In return for your participation you will be asked to choose whether you wish to be awarded 1 lab token OR entry into a Prize Draw.If you choose Prize Draw you will be entered into a prize draw for Amazon vouchers, one randomly selected person will receive vouchers worth £50 and two randomly selected people will receive vouchers worth £25. The draw will take place on February 20th 2015 and you will be contacted by email if you win one of the prizes. You will then need to collect the prizes in person from the researcher.

The data that you provided by participating in this study will be anonymised and will remain confidential. You have the right to withdraw from this study for two weeks after completing the study. You do not have to give a reason to withdraw from the study.  

This study has received a favourable ethical opinion by the University of Surrey FAHS Ethics Committee. If you have any complaint or concerns about any aspects of the way you have been dealt with during the course of this study, then please contact Dr Laura Simonds at the email address given above.

64

[ ] I have read the information above and wish to proceed with this study

65

Consent sheet

Principal InvestigatorNatalie Almond, Natalie.Almond@surrrey.ac.ukCo-InvestigatorDr Laura Simonds, L.Simonds@surrey.ac.uk

Consent formProject title: Writing, personality and appraisals ·I the undersigned voluntarily agree to take part in the study on Writing, personality and attitudes. ·I have read and understood the Information Sheet provided.  I have been given a full explanation by the investigators of the nature, purpose, location and likely duration of the study, and of what I will be expected to do. I have been given the opportunity to ask questions on all aspects of the study and have understood the advice and information given as a result. ·I agree to comply with any instruction given to me during the study and to co-operate fully with the investigators. I shall inform them immediately if I suffer any deterioration of any kind in my health or well-being, or experience any unexpected or unusual symptoms. ·I understand that all personal data relating to volunteers is held and processed in the strictest confidence, and in accordance with the Data Protection Act (1998). I agree that I will not seek to restrict the use of the results of the study on the understanding that my anonymity is preserved. ·I understand that I am free to withdraw from the study for up to two weeks after completing the study, without needing to justify my decision and without prejudice. I understand that after two weeks I will not be able to withdraw my data. ·I acknowledge that in return for signing up for this study I shall earn one lab token OR that I will be entered into a prize draw where one person chosen at random will win £50 and 2 people chosen at random will win £25 in Amazon vouchers. ·I confirm that I have read and understood the above and freely consent to participating in this study.  I have been given adequate time to consider my participation and agree to comply with the instructions and restrictions of the study. Full Name:

Email address (required for lab token OR entry to prize draw):

Please select which of the following you wish to receive for your participation[ ]Lab tokens (course credit for psychology students)[ ] Entry in to the prize draw

66

Measure of Perfectionism

APS-R (Slaney, Rice, Mobley, Trippi, & Ashby, 2001).

The following items are designed to measure attitudes people have toward themselves, their performance, and toward others. There are no right or wrong answers.

Please try to respond to all of the items. Use your first impression and do not spend too much time on individual items in responding.

Respond to each of the items using the scale below to describe your degree of agreement with each item.

1. I have high standards for my performance at work or at school.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

2. I am an orderly person.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

3. I often feel frustrated because I can’t meet my goals.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

4. Neatness is important to me.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

5. If you don’t expect much out of yourself, you will never succeed.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

6. My best just never seems to be good enough for me.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

7. I think things should be put away in their placeStrongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

8. I have high expectations for myself.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

9. I rarely live up to my high standards.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

67

10. I like to always be organized and disciplined.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

11. Doing my best never seems to be enough.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

12. I set very high standards for myself.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

13. I am never satisfied with my accomplishments.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

14. I expect the best from myself.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

15. I often worry about not measuring up to my own expectations.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

16. My performance rarely measures up to my standards.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

17. I am not satisfied even when I know I have done my best.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

18. I try to do my best at everything I do.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

19. I am seldom able to meet my own high standards of performance.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

20. I am hardly ever satisfied with my performance.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

21. I hardly ever feel that what I’ve done is good enough.Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

22. I have a strong need to strive for excellence.Strongly Disagree Slightly Neutral Slightly Agree Strongly

68

Disagree Disagree Agree Agree

23. I often feel disappointment after completing a task because I know I couldhave done better.

Strongly Disagree

Disagree Slightly Disagree

Neutral Slightly Agree

Agree Strongly Agree

69

Failure imagery task

The following task is timed and you will not be able to progress to the next page until the time allowed has passed.

We would like you to imagine a piece of coursework you have been working on and putting a lot of effort into as part of your university studies.

Imagine the grade you want to get and how you would feel if you achieved it . . . . close your eyes or look down and try to imagine this for 30 seconds."

(this page is shown for 50 seconds, allowing 20 seconds to read the instructions and 30 to complete them after this time has elapsed a button to progress to the next page appears)

Now imagine that when your essay is marked and comes back to you, the grade it is given is two grades lower than you wanted. Close your eyes or look down and try to imagine this for 1 minute." During this minute imagine:How would you feel about this lower grade?

What would this lower grade mean to you?

What would you think about yourself?

(this page is shown for 80 seconds, allowing 20 seconds to read the instructions and 60 to complete them, after this time has elapsed a button to progress to the next page appears)

70

State Shame and Guilt scale

(Marschall, Sanftner, & Tangney, 1994).

The following are some statements which may or may not describe how you are feeling right now, following imagining the scenario.

Please rate each statement using the 5-point scale below. Remember to rate each statement based on how you are feeling right at this moment.

1. I feel good about myself

Not feeling this way at all

Feeling this way somewhat

Feeling this way very strongly

1 2 3 4 5

2. I want to sink into the floor and disappear.

Not feeling this way at all

Feeling this way somewhat

Feeling this way very strongly

1 2 3 4 5

3. I feel remorse, regret.

Not feeling this way at all

Feeling this way somewhat

Feeling this way very strongly

1 2 3 4 5

4. I feel worthwhile, valuable.

Not feeling this way at all

Feeling this way somewhat

Feeling this way very strongly

1 2 3 4 5

5. I feel small.

Not feeling this way at all

Feeling this way somewhat

Feeling this way very strongly

1 2 3 4 5

6. I feel tension about something I have done.

Not feeling this way at all

Feeling this way somewhat

Feeling this way very strongly

1 2 3 4 5

7. I feel capable, useful.

Not feeling this way at all

Feeling this way somewhat

Feeling this way very strongly

1 2 3 4 5

71

8. I feel like I am a bad person.

Not feeling this way at all

Feeling this way somewhat

Feeling this way very strongly

1 2 3 4 5

9. I cannot stop thinking about something bad I have done.

Not feeling this way at all

Feeling this way somewhat

Feeling this way very strongly

1 2 3 4 5

10. I feel proud.

Not feeling this way at all

Feeling this way somewhat

Feeling this way very strongly

1 2 3 4 5

11. I feel humiliated, disgraced.

Not feeling this way at all

Feeling this way somewhat

Feeling this way very strongly

1 2 3 4 5

12. I feel like apologizing, confessing.

Not feeling this way at all

Feeling this way somewhat

Feeling this way very strongly

1 2 3 4 5

13. I feel pleased about something I have done.

Not feeling this way at all

Feeling this way somewhat

Feeling this way very strongly

1 2 3 4 5

14. I feel worthless, powerless.

Not feeling this way at all

Feeling this way somewhat

Feeling this way very strongly

1 2 3 4 5

15. I feel bad about something I have done.

Not feeling this way at all

Feeling this way somewhat

Feeling this way very strongly

1 2 3 4 5

72

Manipulation check

Please answer the items below by clicking your answer on the scale.

How easy was it to imagine this situation?

Not easy to imagine

Very easy to imagine

1 2 3 4 5 6 7 8 9 10

How clear was your visualisation of the situation?No clear image

Very clear image

1 2 3 4 5 6 7 8 9 10

How easy was it to imagine how you would feel in this situation?Not easy to imagine

Very easy to imagine

1 2 3 4 5 6 7 8 9 10

How much of the allotted time did you spend thinking about this situation? None of the time

All of the

time0% - - - t o - - - 100%

Enter you answer as a number between 0 and 100 below- there is no need to use a % sign.

73

Writing tasks

General instructions for all conditions

In the next task we would like you to continue to think about the assignment that you did not do as well on as you had anticipated.We would like you to write about this by answering the questions below by typing in to the text boxes provided. Having thought about not doing as well on an assignment as you anticipated we would like you to:

Writing control condition

1. List the building names at the University of Surrey.

2. Write 4 to 5 sentences describing the University of Surrey campus.

3. Write 4 to 5 sentences describing what it is like to be a student at the University of Surrey.

OR

Self-esteem writing

1. List your positive characteristics—indications that you are competent and valuable.

2. Write 4 to 5 sentences explaining how what happened was not entirely your fault and to think about the event in a way that makes you feel better about yourself.

3. Write 4 to 5 sentences describing why the event does not really indicate anything about the kind of person you are.

OR

Self-compassion writing

1. List what you think a friend might think and feel about themselves following the experience of a similar event.

2. Write 4 to 5 sentences that express understanding, kindness, and concern to yourself in the same way that you might express concern to a friend who had undergone the same experience.

3. Write 4 to 5 sentences describing your experience of the event in a factual way without making any judgements or evaluations about yourself or the event.

74

Repetition of State Shame and Guilt Scale

Feedback and demographics

We would be grateful if you would complete a few final questions about this study.

1) What do you think this study was about?

2) Were you suspicious at any point that the study was looking at something other than what was stated? [ ] Not at all[ ] A little[ ] A lot

Please could you tell us your:

Gender:

[ ] Male[ ] Female[ ] Nationality

First language:

[ ] English[ ] Other

Religion:

_______________

Age:

_________________

Subject studied at University:

_____________________

Year of study:[ ] Undergraduate Year 1[ ] Undergraduate Year 2[ ] Undergraduate Year 3[ ] Masters

75

[ ] Doctoral study[ ] Other

76

Debrief

In this study we were interested in finding out, if people’s personality influenced how they experienced the imagery task. The aspect of personality that we were interested in was Perfectionism. Perfectionists are individuals who have unrealistically high standards that they adhere to rigidly and for these individual’s their self-worth is defined in relation to meeting these high standards. We were interested in what feelings were created by imagining failing to meet your own standards on a coursework assignment and whether those who were more perfectionistic experienced different feelings, or felt the same feelings more strongly. We anticipated that those who were more perfectionistic, may experience more shame and guilt than those who were less perfectionistic. We were also interested in whether writing (particularly compassionate writing) would reduce the extent to which individuals experienced shame and guilt and whether there would be less reduction in shame and guilt for those who were more perfectionistic. It was important that we did not tell you that we would ask you to think about failing as this may have made it very difficult to recruit participants who had more perfectionistic tendencies! If you have felt upset or distressed as a result of completing this study then please visit the Centre for Well-being at the University of Surrey. Information about the Centre and how to contact them can be found here

Centre for well-being webpage (copy paste in to your browser)  http://portal.surrey.ac.uk/portal/page?_pageid=729,1&_dad=portal&_schema=PORTAL

77

Appendix E

Histograms for discrepancy (maladaptive perfectionism)

Figure E1. Distribution of discrepancy for those who did not complete the study.

Figure E2. Distribution of discrepancy for those who completed the study (final

sample).

ZSkewness= 1.83

ZKurtosis= .37

ZSkewness= .88

ZKurtosis= -1.05

78

Appendix F

Histograms of variables for correlations

Figure F1- Histogram of discrepancy.

Figure F2- Histogram of Time 1 pride

ZSkewness= .88

ZKurtosis= -1.05

ZSkewness= 2.61

ZKurtosis= -.60

79

Figure F3- Histogram of Time 1 shame

Figure F4- Histogram of Time 1 guilt

ZSkewness= .68

ZKurtosis= -2.01

ZSkewness= -.11

ZKurtosis= -1.90

80

Figure F5- Histogram of imageability check.

Figure F6- Histogram of percentage time spent on imagery.

ZSkewness= -2.23

ZKurtosis= -.05

ZSkewness= -2.76

ZKurtosis= .70

81

Figure F7- Histogram of Time 2 shame.

Figure F8- Histogram of Time 2 guilt.

ZSkewness= 4.15

ZKurtosis= .59

ZSkewness= 2.24

ZKurtosis= -1.08

82

Figure F9- Histogram of Time 2 pride.

ZSkewness= .70

ZKurtosis= -1.20

83

Appendix G

Table G1. Bivariate Spearman’s Correlations for Time 1 Variables.

Imageabili

ty

Percentage

time

Time 1

shame

Time 1

guilt

Time 1

pride

Time 2

shame

Time 2

guilt

Time 2

pride)

Discrepancy

(n=95)

rs = .15

p = .16

rs = .30**

p = .003

rs = .44***

p < .001

rs = .37***

p < .001

rs = -.33**

p = .001

rs = .43***

p < .001

rs = .36***

p < .001

rs = -.27**

p = .008

Imageability

(n=93)

rs = .25*

p = .02

rs = .21

p = .05

rs = .10

p =.33

rs = -.27*

p =.04

rs = .09

p = .42

rs = .06

p = .56

rs = -.08

p = .45

Percentage time

(n=95)

rs = .39***

p < .001

rs = .37***

p < .001

rs = -.21*

p = .04

rs = .12

p = .05

rs = .10

p = .35

rs = .03

p = .79

Time 1 shame

(n=95)

rs = .77***

p < .001

rs = -.69***

p < .001

rs = .61***

p < .001

rs = .47***

p < .001

rs = -.31**

p = .002

Time 1 guilt

(n=95)

rs = -.49***

p < .001

rs = .61***

p < .001

rs = .63***

p < .001

rs = -.21*

p = .04

Time 1 pride

(n=95)

rs = -.46***

p < .001

rs = -.310**

p = .002

rs = .53***

p < .001

Time 2 shame

(n=95)

rs = .79***

p < .001

rs = -.54***

p < .001

Time 2 guilt

(n=95)

rs = -.48***

p < .001

84

Note- Asterisks denote the significance level * p < .05, ** p < .01 and *** p <.001

85

Appendix H

Histograms of residuals from one- way ANOVA’s

Figure H1- Histogram of standardised residuals for discrepancy.

Figure H2- Histogram of standardised residuals for imageability.

ZSkewness= .71

ZKurtosis= -.97

ZSkewness= -2.22

ZKurtosis= -.03

86

Figure H3- Histogram of standardised residuals for percentage time spent

Figure H4- Histogram of standardised residuals for Time 1 shame.

ZSkewness= .57

ZKurtosis= -2.05

ZSkewness= -3.14

ZKurtosis= 1.80

87

Figure H5- Histogram of standardised residuals for Time 1 guilt.

Figure H6- Histogram of standardised residuals for Time 1 pride.

ZSkewness= 2.71

ZKurtosis= -.32

ZSkewness= -.06

ZKurtosis= -1.72

88

Appendix I

Table I1- Levine’s test for homogeneity of variance for one-way ANOVA on writing

task for Time 1 variables.

Levene Statistic df 1 df 2 Sig.

Discrepancy

(Maladpative

perfectionism)

4.00 2 92 < .05

Imageability 1.04 2 92 > .05

Percentage time spent

on failure scenario

0.23 2 92 > .05

Time 1 shame 2.78 2 92 > .05

Time 1 guilt 1.54 2 92 > .05

Time 1 pride 1.03 2 92 > .05

89

Appendix J

Table J1- Kruskal-Wallis tests on writing condition for Time 1 variables.

H (DOF) p

Discrepancy (Maladpative

perfectionism)

.19 2 > .05

Imageability .39 2 > .05

Time 1 shame .06 2 > .05

Time 1 pride 3.76 2 > .05

90

Appendix K

Histograms of the residuals for all ANCOVAs on shame, guilt and pride.

Figure K1- Histogram of the residuals for Time 2 shame in ANCOVA on writing condition with Time 1 shame and discrepancy as covariates.

Figure K2- Histogram of the residuals for Time 2 guilt in ANCOVA on writing condition with Time 1 guilt and discrepancy as covariates.

ZSkewness= .73

ZKurtosis= .21

ZSkewness= .87

ZKurtosis= -.42

91

Figure K3- Histogram of the residuals for Time 2 pride in ANOVA on writing condition with Time 1 pride and discrepancy as covariates.

ZSkewness= 1.39

ZKurtosis= -.46

92

Appendix L

Table L1- Levine’s test for homogeneity of variance for ANOVAs and ANCOVAs

on shame, guilt and pride. (N=95)

Levene

Statistic

df 1 df 2 Sig.

Time 2 shame .032 2 92 > .05

Time 2 guilt 4.51 2 92 < .05

Time 2 pride .484 2 92 > .05

93

Appendix M

Histograms of residuals for ANOVA on LIWC variables

Figure M1 Histogram of residuals word count for writing task

Figure M2- Histogram of residuals first-person singular words percentage score

ZSkewness= .87

ZKurtosis= .02

ZSkewness= 2.46

ZKurtosis= 2.24

94

Figure M3- Histogram of residuals third-person singular words percentage score.

Figure M4- Histogram of residuals third-person plural words percentage score.

ZSkewness= 21.00

ZKurtosis= 58.32

ZSkewness= 6.94

ZKurtosis= 14.84

95

Figure M5- Histogram of residuals of past-tense words percentage score.

Figure M6- Histogram of residuals of present-tense words percentage score.

ZSkewness= 3.55

ZKurtosis= 6.82

ZSkewness= -1.14

ZKurtosis= .45

96

Figure M7- Histogram of residuals of future-tense words percentage score.

Figure M8- Histogram of residuals of positive emotion words percentage score.

ZSkewness= 1.67

ZKurtosis= 2.31

ZSkewness= 5.99

ZKurtosis= 8.80

97

Figure M9- Histogram of residuals of negative emotion words percentage score.

Figure M10- Histogram of residuals of work words percentage score.

ZSkewness= 6.76

ZKurtosis= 14.12

ZSkewness= .18

ZKurtosis= -1.07

98

Figure M11- Histogram of residuals of achievement words percentage score.

ZSkewness= 9.06

ZKurtosis= 21.04

99

Appendix N

Table N1- Levine’s test statistics for ANOVAs conducted on LIWC variables

Levene Statistic df 1 df 2 Sig.

Word count 1.98 2 92 > .05

First-person singular 4.00 2 90 < .05

Third-person singular 7.07 2 90 < .01

Third-person plural 18.67 2 90 < .001

Past-tense 9.63 2 90 < .001

Present-tense 0.45 2 90 > .05

Future-tense 24.43 2 90 < .001

Positive emotions 0.22 2 90 > .05

Negative emotions 10.46 2 90 < .001

Work 3.76 2 90 < .05

Achievement 0.84 2 90 > .05

100

Appendix O

Table O1. Mean Ranks and Statistics for Kruskal-Wallis tests investigating the effect of writing condition on LIWC variables.

Dependant variable Mean Ranks H (DOF) p

Control Self-

esteem

Self-

compassion

1st person singular (% of text) 22.78 69.32 49.90 43.93 (2) <. 001

3rd person plural (% of text) 34.62 37.14 65.50 32.10 (2) <. 001

Past-tense (% of text) 18.03 55.02 65.41 53.94 (2) <. 001

Future-tense (% of text) 22.10 53.61 63.06 43.02 (2) <. 001

Positive emotions (% of text) 49.95 51.07 41.21 2.60 (2) >. 05

Negative emotions (% of text) 22.12 42.66 71.80 56.06 (2) <. 001

Work (% of text) 64.13 42.61 35.83 18.83 (2) <. 001

Achievement (% of text) 27.77 58.46 54.31 22.87 (2) <. 001

101

MRP Research Proposal

Can compassionate-imagery reduce perfectionist’s experience of shame?

102

Introduction

Background and Theoretical Rationale

Perfectionism is associated with a variety of physical and psychological

disorders (Burns, 1980; Pacht, 1984; Shafran & Mansell, 2001). Egan, Wade and

Shafran (2011) have argued that perfectionism is a transdiagnostic process and have

provided evidence that perfectionism is associated with a range of symptoms

associated with psychological distress and that treating perfectionism can be helpful

in reducing psychological distress. Shafran and Mansell (2001) have described

perfectionists as individuals holding unrealistically high standards that they adhere to

rigidly and note that the individual’s self-worth is defined in relation to meeting

these high standards. Despite this straightforward definition, there are currently three

scales with different focuses that are typically used to measure perfectionism in the

research literature these are: Multifactorial perfectionism scale (F-MPS; Frost et al.,

1990), Multifactorial perfectionism scale (H-MPS; Hewitt & Flett, 1991), and the

Almost Perfect Scale- Revised (APS-R: Slaney, Rice, Mobley, Trippi & Ashby,

2001).

Factor-analytic investigations of these perfectionism scales typically reveals

two higher-order factors which have been variously described as adaptive vs.

maladaptive perfectionism (Enns, Cox, & Clara, 2002; Rice, Ashby & Slaney, 1998;

Suddarth & Slaney, 2001), positive striving vs. maladaptive concerns (Bieling,

Isralei, & Antony, 2004), personal standards vs. evaluative concerns (Dunkley,

Blankstein, & Berg, 2012) and perfectionistic strivings vs. perfectionistic concerns

(Dunkely, Blankstein, Masheb, & Grillo, 2006; Stöber & Otto, 2006). Reviewing

these factor analyses Stöber & Otto, (2006) suggest that there are two distinct factors

underlying perfectionism; the adaptive factor captures having high personal

103

standards, and striving for excellence and the maladaptive factor seems to capture

concern over mistakes, doubts about actions, fear of disapproval by others, and

discrepancy between expectations and results. The maladaptive factor seems to

capture self-evaluation, and self-criticism about whether an individual is meeting

their own standards and it is this factor that is associated with predictors of

maladjustment such as negative affect (see Stöber & Otto, 2006 for a review).

Therefore it is this aspect of perfectionism that is of interest.

Tangey (2002) argues that when people evaluate themselves against a set of

standards they often experience the self-evaluative emotions (shame, guilt,

embarrassment and pride). Tangey notes that perfectionists are continually self-

evaluating (especially maladaptive perfectionists) and predicts they will experience

these emotions more than others. There is good evidence that socially-prescribed

perfectionism (an aspect of maladaptive perfectionism) is positively correlated with

proneness to experiencing shame and guilt (Lutwak & Ferrari, 1996: Tangey, 2002)

and with feelings of shame and guilt (Klibert, Langhinrichsen-Rohling, & Saito,

2005). Fedewa, Burns and Gomez (2005) also demonstrated that negative

perfectionism showed positive correlations with proneness to shame and guilt, and

with state shame and guilt. Stöber, Harris and Moon, (2007) extended these findings

demonstrating that healthy perfectionists experienced more state pride and less state

shame and state guilt than unhealthy perfectionists and non-perfectionists.

Collectively, these findings suggest that shame and guilt are associated with

maladaptive or negative aspects of perfectionism (see also Ashby, Rice, & Martin,

2006). In summary, individuals who endorse maladaptive perfectionism are thought

to be engaging in more self-evaluation and self-criticism, and these individuals

experience more shame and guilt than others.

104

The literature on perfectionism and self-criticism suggest that they develop

in a similar manner. Drawing on attachment theory Gilbert and Procter (2006) note

that insecure attachment styles developed in childhood can lead the individual to

view others as a source of threat to the self, and that this results in heightened self-

monitoring, self-blaming, self-criticism and striving to meet other people’s

expectations of the self. Gilbert and Procter also note that modelling, safety

strategies/behaviours with hostile others, shame (Andrews, 1998; Gilbert, 1998),

inabilities to process anger (Ferster, 1973), and lack of internal schema of others as

safe/supportive (Mikulincer & Shaver, 2004), all potentially contribute to the

development of self-criticism. Interestingly similar factors are proposed to contribute

to the development of perfectionism (See Almond & Strauss, 2013) and in both cases

insecure attachment styles are implicated. Although self-criticism and associated

shame may develop through a range of processes, self-criticism is typically

negatively correlated with abilities to be compassionate and self-affiliative (Gilbert,

Clark, Hempel, Miles, & Irons, 2004; Gilbert, McEwan, Gibbons, Chotai, Duarte, &

Matos, 2011).

Gilbert (2009) proposes that there are three type of emotion-regulation

system (achieving and activating, threat-protection and affiliative-soothing) that

interact to regulate our emotions dependent on context. Compassion-focussed

therapy aims to bolster the affiliative and soothing system, which is often under-

developed in those experiencing psychological distress who may tend to rely on

achievement and activating strategies and/or threat-protection strategies.

Compassion-focussed therapy has been shown to produce significant reductions in

depression, anxiety, self-criticism, shame, and increase the individual’s ability to be

self-soothing (Gilbert & Procter, 2006; Judge, Cleghorn, McEwan & Gilbert, 2012).

105

Perfectionists are likely to rely achievement/activating and threat-protection

emotion-regulation systems, and tend to experience self-criticism and shame. It is

plausible that compassionate-mind training for perfectionists, could increase

soothing-affiliative emotion-regulation and reduce their experience of self-criticism

and shame.

Compassionate-mind interventions are often 12-14 weeks in duration,

however, given perfectionists are proposed to experience self-criticism at the point of

failing to meet standards and this self-criticism plays a role in maintaining

perfectionism (Shafran, Egan, & Wade, 2010), it is pertinent to consider whether a

brief compassion intervention at the point of failure, could reduce perfectionists

experience of shame and self-criticism. Addressing this question is the focus of this

project.

One aspect of compassionate-mind training lends itself to a brief intervention

is compassionate-imagery (Gilbert & Irons, 2004; Gilbert, Baldwin, Irons & Palmer,

2006). Indeed, Rockcliff, Gilbert, McEwan, Lightman, & Glover (2006) provide

evidence that a single dose of compassionate-imagery can have soothing effects. One

commonly used form of compassionate-imagery is to develop an image that

represents or embodies warm and compassionate qualities, this is sometimes referred

to as an ideal or perfect nurturer (Lee, 2005). Therefore this research will investigate

whether such compassionate-imagery can reduce perfectionist’s experience of shame

and self-criticism following failing to meet their standards.

Research Question

Can compassionate-imagery reduce maladaptive perfectionist’s experience of

shame, and guilt, following failure?

Main Hypotheses

106

Individuals who score more highly on the measure of maladaptive

perfectionism will report lower levels of shame and guilt following exposure to

compassionate-imagery compared to non-compassionate imagery.

Individuals who score more highly on the measure of maladaptive

perfectionism will show a greater reduction in their ratings of shame and guilt

following exposure to compassionate-imagery than following exposure to non-

compassionate imagery.

Method

Participants

A power analysis was conducted to determine the sample size required for

this study (alpha= 0.05, and power= 0.80). A medium effect size was anticipated for

this study and based on Cohen’s (1988) descriptions of F squared effect sizes, the

effect size was estimated to be= .015. There would be three predictors discrepancy,

compassionate-imagery condition and their interaction. This power analysis

suggested a sample of 77 would be required.

The experimental design of this study uses imagery, therefore it is plausible

that some participants will not actually engage in this imagery and should be

removed from the sample. To account for this possibility the data will be checked as

recruitment is on-going so that more data can be collected to replace those who have

not engaged with the task appropriately.

Participants will all be University of Surrey students (both undergraduates

and postgraduates). Psychology students will be recruited initially as they are easily

accessible to the researcher. The study will be advertised via the Psychology

research panel. The recruitment will be widened out to all University of Surrey

students if sufficient numbers cannot be recruited from within Psychology.

107

Participants who are not students at the University of Surrey will not be included in

the sample.

There will likely be approximately 240 undergraduate psychology students

and 90 postgraduate psychology students to collect data from. From my own

experience collecting data in University settings, the response rate is usually

relatively high in Psychology students who will be required to collect data for their

own study at some point in their student careers. Recruiting enough participants

would require 2/7 of the psychology student population to complete the study. From

my own experience conducting research in University settings, online studies are

typically preferable to this sample, therefore it seems reasonable that 2/7 students

could be recruited. The University of Surrey website suggests that in 2011-2012 their

were 15,209 undergraduate and postgraduate students, therefore expanding the study

out to the whole student population should ensure that the required sample size could

be collected.

Design

The study has a 2 (Compassionate-imagery: compassionate-imagery vs. non-

compassionate-imagery) between-subjects design. Maladaptive perfectionism

(measured used APS-R discrepancy scale) will be a measured independent variable

and will be treated as a continuous variable.

The dependent variables are the emotions of shame and guilt.

Measures/Interviews/Stimuli/Apparatus

All measures and manipulations can be seen in Appendix 1 in full.

Perfectionism measure

Perfectionism will be measured using the APS-R (Slaney, Rice, Mobley,

Trippi & Ashby, 2001). This is a 23-item scale with three subscales Discrepancy,

108

High standards, Order. The APS-R is the only perfectionism scale that was designed

to capture the adaptive elements of perfectionism in one scale (High standards) and

the maladaptive components of perfectionism in a separate scale (Discrepancy).

APS-R has acceptable construct, concurrent, convergent and discriminant validity

(Rice, Ashby & Slaney, 2007) the three factors were reliable with a confirmatory

factor analysis producing Cronbachs alphas that ranged from .85- .92 for the

subscales (Slaney et al., 2001) and test-retest reliability is good (Rice & Aldea,

2006).

Manipulation of failure

Participants will be asked to complete modified version of an imagery task

taken from Gilbert et al., (2006). Participants will be asked to imagine completing a

piece of university coursework and finding out that they do not do a well as they

expect to. This is an appropriate manipulation of failure because it concerns a task

that most students would likely want to succeed in, and using imagery in this way

means that the experimenter can manipulate a discrepancy between what was hoped

for and what was achieved. Both of these can be hard to achieve in tasks where

failure is induced experimentally via task-completion and negative feedback.

Measures of shame and guilt

Shame and guilt will be measured using the State Shame and Guilt Scale

(SGSS: Marschall, Sanftner, & Tangney, 1994). Robins, Noftle, and Tracey (in

press) note that the subscales have good cronbachs alphas; shame (5 items; α = .89),

guilt (5 items; α = .82), pride (5 items; α = .87). No other psychometric information

is available but this measure is often used in the published literature (Fedewa et al.,

2005; Stöber et al., 2007).

Manipulation of compassionate-imagery

109

Similar to Gilbert et al., (2006) participants will then be asked to complete a

subsequent imagery task. This task will manipulate the use of compassionate versus

non-compassionate-imagery. The compassionate-imagery task is a composite of a

task described in Gilbert & Procter (2006) and from Lee and James’ (2011) perfect

nurturer task. The non-compassionate imagery task is matched to the compassionate-

imagery task as closely as possible removing all references to compassion. The task

can be seen in full in the appendix but requires participants to read the instructions

and type notes as they develop a compassionate/ non-compassionate image and then

bring this image to mind for three minutes.

Manipulation check of experience of failure-imagery / compassionate-imagery

Given that participants will be completing tasks using imagery it will be

important to check to what extent they have engaged in /completed the imagery

tasks, how easy or difficult they found it to complete these imagery tasks and how

clear the image was. Participants will be asked these questions (adapted from Gilbert

et al., 2006) once after the failure-imagery and once following the manipulation of

compassionate/ non-compassionate-imagery. The items can be seen in the Appendix.

Measure of mood

A 20-item self-report measure of positive and negative affect (PANAS:

Watson, Clark, and Tellegen (1988) will be used to determine whether the

manipulations of compassionate and non-compassionate-imagery have divergent

effect on mood more generally. The PANAS is considered to have good reliability

Cronbachs alpha .89 for the Positive Affect scale, and .85 for the Negative Affect

scale (Crawford & Henry, 2004).

Procedure

An opportunity sample of students will be recruited from the University of

110

Surrey by advertising the study on the Psychology research panel where participants

will receive research credit for completion of the study. Participants will also be

recruited by advertising the study throughout the University campus using posters

and emails. A prize draw will be offered to incentivise participation.

Students who chose to complete the study will be directed to a URL to

complete the study online. Participants will be initially asked to complete the

measure of perfectionism. All participants will then complete the failure-imagery

task followed by measure of shame and guilt. At this point the procedure diverges by

condition, with half of the participants being asked to complete a manipulation of

compassionate-imagery and half non-compassionate-imagery. All participants will

then repeat the measure of shame and guilt, before completing the measure of mood,

and the imagery manipulation checks. Finally participants will provide demographic

information, and will be thanked for participation and debriefed.

Ethical considerations

The University of Surrey Faculty of Arts and Humanities Ethics Committee

will be asked to review the ethics of conducting this research.

Deception

It will be necessary to deceive participants about the true purpose of the

study. Participants will be told that the study investigates the effects of imagery on

their appraisals of situations. If participants are told that the study is about failure it

is likely that this may deter those who are more perfectionistic in particular and it is

necessary to recruit students with a wide range of perfectionism scores.

Distress

It is possible that asking students who are perfectionist to think about failure

could cause a small amount of distress, in particular, self-criticism and shame.

111

However, it is important to note that not all participants will be perfectionists and so

not all participants will experience distress. However, the purpose of this experiment

is to investigate whether self-compassionate-imagery can reduce the distress

experienced by perfectionists, so in order to test the efficacy of such an intervention

it is necessary to induce a modicum of distress. Furthermore, investigating this topic

via imagery, may allow participants to distance themselves from this distress more

easily, as this is not a real experience of failure.

R&D Considerations

There are no R and D considerations for this project.

Proposed Data Analysis

For all scales the subscale means will be calculated by reversing items as

necessary and replacing any missing items with the subscale mean score. The

discrepancy scale of the APS-R will be considered as a continuous independent

variable that represents the extent to which participants endorse maladaptive aspects

of perfectionism where higher scores represent more maladaptive perfectionism.

Moderated regression analyses (Aiken & West, 1991) regressions will be conducted

to assess whether discrepancy (continuous) and compassionate-imagery condition

(categorical) and their interaction predict initial shame and guilt following the

failure-imagery manipulation for all participants. It is anticipated based on the

previous research evidence that shame, and guilt should be predicted by discrepancy.

It would be preferable that compassionate-imagery condition did not predict initial

shame or guilt because this would suggest that there were differences between our

experimental groups levels of shame and guilt prior to the manipulation of

compassionate-imagery.

Moderated regression analysis will be used to determine whether

112

discrepancy, self-compassionate-imagery condition, and their interaction predict the

experience of 1) shame and 2) guilt following the compassionate-imagery

manipulation. This will require the continuous variables to be centred around zero,

and compassionate-imagery condition to be dummy coded (0,1) and an interaction

term calculated by computing the product of these variables.

The design of the study will also make it possible to consider whether the

change in shame and guilt (from post failure manipulation to post compassionate-

imagery manipulation) is predicted by discrepancy, compassionate-imagery

condition and their interaction again using a moderated regression analysis.

However, this would only be considered if shame and guilt are not predicted by

compassionate-imagery condition following failure imagery.

Service User and Carer Consultation / Involvement

I have presented my research plans to a member of the university/ service

user and carer panel via my oral presentation of this research proposal. They raised

concerns about how ethnicity may predict the experience of perfectionism; I will

measure ethnicity as one of the demographic variables collected at the end of my

study. I plan to consult the carer/service user representatives to help me consider the

clarity of my materials in particular how easy the imagery manipulations are to

follow.

Feasibility Issues

It may be difficult to recruit the required numbers of participants from the

psychology research panel. Therefore, I will plan to recruit from the university

student population in addition to the psychology panel and include this in my initial

ethics proposal so that I can recruit more broadly if required without needing to

reapply for ethical approval.

113

Analysing the data using a moderated regression approach allows for

consideration of perfectionism/ discrepancy as a continuous variable so that one can

consider how increases in perfectionism predict the other variables. It will be

important to ensure that there are a comparable range of discrepancy scores in each

of the compassionate-imagery conditions. This will be achieved by regularly

downloading the data from the online questionnaire. The way in which participants

are recruited/ allocated to condition could then be adjusted to account for their

perfectionism score to distribute this equally across conditions.

It is possible that participants will not actually engage in the imagery

manipulations as requested, therefore, a question will be included in the imagery

manipulation check to assess how much of the allotted time participants spent on the

imagery task.

Dissemination strategy

Following analysis of the findings, if appropriate the results will be submitted

to an academic conference for presentation and/ or to an academic journal for

publication.

Study Timeline

Oct13-Dec13

Jan14-Mar14

Apr14- Jun14

Jul14-Sep14

Oct14-Dec14

Jan 15- Feb 15

MRP Course approvalEthics submissionEthics approvalDraft introductionDraft methodsData collectionData analysisDraft resultsDraft discussionFinalise draftComplete draft to supervisor

114

References

Almond, N. R., & Strauss, C., (2013). Perfectionism, parental bonding and

attachment. Unpublished manuscript, University of Surrey.

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting

interactions. Newbury Park, London, Sage.

Andrews, B. (1998). Shame and childhood abuse. In P. Gilbert, & B. Andrews (Eds),

Shame: Interpersonal behavior, psychopathology and culture (pp. 176–190).

New York: Oxford University Press.

Ashby, J., Rice, K., & Martin, J. (2006). Perfectionism, shame, and depressive

symptoms. Journal of Counseling & Development, 84, 148-156. doi:

10.1002%2Fj.1556-6678.2006.tb00390.x

Bieling, P. J., Israeli, A. L., & Anthony, M. M. (2004). Is perfectionism good, bad,

or both? Examining models of the perfectionism construct. Personality and

Individual Differences, 36, 1373-1385. doi: 10.1016%2FS0191-

8869%2803%2900235-6

Burns, D. D. (1980). Feeling good: the new mood therapy. New York: New

American Library.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale,

NJ: Erlbaum.

Crawford, J. R., & Henry, J. D. (2004). The Positive and Negative Affect Schedule

(PANAS): Construct validity, measurement properties and normative data in

a large non-clinical sample. British Journal of Clinical Psychology, 43, 245–

265. doi: 10.1348%2F0144665031752934

Dunkley, D. M., Blankstein, K. R., & Berg, J.-L. (2012). Perfectionism Dimensions

and the Five‐factor Model of Personality. European Journal of Personality,

115

26, 233–244. 10.1002/per.829

Dunkley, D. M., Blankstein, K. R., Masheb, R. M., & Grilo, C. M. (2006). Personal

standards and evaluative concerns dimensions of “clinical” perfectionism: A

reply to Shafran et al. (2002, 2003) and Hewitt et al. (2003). Behaviour

Research and Therapy, 44, 63–84. doi:10.1016%2Fj.brat.2004.12.004

Egan, S. J., Wade, T. D., & Shafran, R. (2011). Perfectionism as a transdiagnostic

process: A clinical review. Clinical Psychology Review, 31, 203–212. doi:

0.1016/j.cpr.2010.04.009.

Enns, M. W., & Cox, B. J. (2002). The nature and assessment of perfectionism: A

critical analysis. In G. L. Flett & P. L. Hewitt (Eds.), Perfectionism: Theory,

research, and treatment (pp. 33-62). Washington, DC: American

Psychological Association.

Fedewa, B. A., Burns, L. R., & Gomez, A. A. (2005). Positive and negative

perfectionism and the shame/guilt distinction: Adaptive and maladaptive

characteristics. Personality and Individual Differences, 38, 1609-1619. doi:

10.1016/j.paid.2004.09.026

Ferster, C.B. (1973). A functional analysis of depression. American Psychologist, 28,

857–870. doi: 10.1037%2Fh0035605

Frost, R. O., Marten, P., Lahart, C. M., & Rosenblate, R. (1990). The dimensions of

perfectionism. Cognitive Therapy and Research, 14, 449–468. doi:

10.1007%2FBF01172967

Gilbert, P. (1998). What is shame? Some core issues and controversies. In P. Gilbert,

& B. Andrews (Eds), Shame: Interpersonal behavior, psychopathology and

culture (pp.3–36). New York: Oxford University Press.

Gilbert, P. (2009). The compassionate mind. Londo: Constable & Robinson.

116

Gilbert, P., Baldwin, M., Irons, C., Baccus, J., & Clark, M. (2006). Self-criticism and

self-warmth: An imagery study exploring their relation to depression. Journal

of Cognitive Psychotherapy, 20, 183–200. doi: 10.1891%2Fjcop.20.2.183

Gilbert, P., Clarke, M., Kempel, S., Miles, J.N.V., & Irons, C. (2004). Criticizing

and reassuring oneself: An exploration of forms style and reasons in female

students. British Journal of Clinical Psychology, 43, 31–50. doi:

10.1348/014466504772812959

Gilbert, P., & Irons, C. (2004). A pilot exploration of the use of compassionate

images in a group of self-critical people. Memory, 12, 507–516. doi:

10.1080/09658210444000115

Gilbert, P., McEwan, K., Gibbons, L., Chotai, S., Duarte, J., & Matos, M. (2011).

Fears of compassion and happiness in relation to alexithymia, mindfulness and

self-criticism. Psychology and Psychotherapy, 84, 239–255. doi:

10.1111%2Fj.2044-8341.2011.02046.x

Gilbert, P., & Procter, S. (2006). Compassionate mind training for people with high

shame and self-criticism: A pilot study of a group therapy approach. Clinical

Psychology and Psychotherapy, 13, 353–379. doi: 10.1002/cpp.507

Hewitt, P. L., & Flett, G. L. (1991). Perfectionism in the self and social contexts:

Conceptualization, assessment, and association with psychopathology. Journal

of Personality and Social Psychology, 60, 456–470. doi: 10.1037%2F%2F0022-

3514.60.3.456

Judge, L., Cleghorn, A., McEwan, K., & Gilbert, P. (2012). An exploration of group-

based compassion focused therapy for a heterogeneous range of clients

presenting to a Community Mental Health Team. International Journal of

Cognitive Therapy, 5(4), 420–429, doi: 10.1521%2Fijct.2012.5.4.420

117

Klibert, J. J., Langhinrichsen-Rohling, J., & Saito, M. (2005). Adaptive and

maladaptive aspects of self-oriented versus socially prescribed perfectionism.

Journal of College Student Development, 46, 141–156. doi:

10.1353%2Fcsd.2005.0017

Lee, D. A., (2005). The perfect nurturer: A model to develop a compassionate mind

within the context of cognitive therapy p 326-351. In P. Gilbert (Ed.)

Compassion: Conceptualisations, Research and Use in Psychotherapy. Hove:

Routledge

Lee, D. A., and James, S. (2011). The compassionate-mind guide to recovering from

trauma and PTSD. Oakland, CA: New harbinger publications.

Lutwak, N., & Ferrari, J. R. (1996). Moral affect and cognitive processes:

Differentiating shame from guilt among men and women. Personality and

Individual Differences, 21, 891-896. doi: 10.1016%2FS0191-

8869%2896%2900135-3

Marschall, D. E., Saftner, J., & Tangney, J. P. (1994). The State Shame and Guilt

Scale. George Mason University, Fairfax, VA.

Mikulincer, M., & Shaver, P.R. (2004). Security-based self representations in

adulthood: Contents and processes. In N.S. Rholes, & J.A. Simpson (Eds), Adult

attachment: Theory, research, and clinical implications. (pp. 159–195). New

York: Guilford.

Pacht, A. R. (1984). Reflections on perfection. American Psychologist, 39, 386-390.

doi:10.1037%2F%2F0003-066X.39.4.386

Rice, K. G., & Aldea, M. A. (2006). State dependence and trait stability of

perfectionism: A short-term longitudinal study. Journal of Counseling

Psychology, 53, 205-213. doi: 10.1037%2F0022-0167.53.2.205

118

Rice, K. G., Ashby, J. S., & Slaney, R. B. (1998). Self-esteem as a mediator between

perfectionism and depression: A structural equation analysis. Journal of

Counseling Psychology, 45, 304-314. doi: 10.1037%2F%2F0022-

0167.45.3.304

Rice, K. G., Ashby, J. S., & Slaney, R. B. (2007). Perfectionism and the five-factor

model of personality. Assessment, 14, 385-398. doi:

10.1177/1073191107303217

Robins, R. W., Noftle, E. E., & Tracey, J. L. (in press). Assessing Self-Conscious

Emotions: A Review of Self-Report and Nonverbal Measures. In Tracy, J. L.,

Robins, R. W., & Tangney, J. M. (Eds.) The Self-Conscious Emotions:

Theory and Research. New York: Guilford.

Rockliff, H., Gilbert, P., McEwan, K. Lightman, S., & Glover, D. (2008). A pilot

exploration of heart rate variability and salivary cortisol responses to

compassion-focused imagery. Clinical Neuropsychiatry, 5 (3), 132-139. doi:

not avaliable

Shafran, R., Egan, S., & Wade, T. (2010). Overcoming perfectionism. London:

Constable and Robinson.

Shafran, R., & Mansell, W. (2001). Perfectionism and psychopathology: A review of

research and treatment. Clinical Psychology Review, 21, 879−906. doi:

10.1016%2FS0272-7358%2800%2900072-6

Slaney, R. B., Rice, K. G., Mobley, M., Trippi, J., & Ashby, J. S. (2001). The

revised Almost Perfect Scale. Measurement and Evaluation in Counseling and

Development, 34, 130–145. doi: not available.

Stöber, J., Harris, R. A., & Moon, P. S. (2007). Perfectionism and the experience of

pride, shame, and guilt: Comparing healthy perfectionists, unhealthy

119

perfectionists, and nonperfectionists. Personality and Individual Differences,

43, 131-141. doi: 10.1016/j.paid.2006.11.012

Stöber, J., & Otto, K. (2006). Positive conceptions of perfectionism: Approaches,

evidence, challenges. Personality and Social Psychology Review, 10, 295–319.

doi: 10.1207%2Fs15327957pspr1004_2

Suddarth, B. H., & Slaney, R. B. (2001). An investigation of the dimensions of

perfectionism in college students. Measurement and Evaluation in

Counseling and Development, 34, 157-165. doi: not available.

Tangney, J. P. (2002). Perfectionism and the self-conscious emotions: Shame, guilt,

embarrassment, and pride. In P. L. Hewitt & G. L. Flett (Eds.),

Perfectionism: Theory, research, and treatment (pp. 199-215). Washington,

DC: American Psychological Association.

Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of

brief measures of positive and negative affect: The PANAS scales. Journal of

Personality & Social Psychology, 47, 1063-1070. doi: 10.1037//0022-

3514.54.6.1063

120

MRP Literature review

Perfectionism, parental bonding and attachment.

121

Journal choice and justification

I plan to submit a version of this literature review to Clinical Psychology

Review because the aims and scope of the journal are suitable to the topic of the

review I have written. The journal states that it publishes reviews on topics that have

a clear relationship to research or practice in Clinical Psychology. The literature

review includes a section that justifies why the topic is relevant to Clinical

Psychology. Furthermore, the journal has a good impact factor (7.07) suggesting that

publishing in this journal would mean that the research would be read and cited by

others and promote good exposure for my work.

122

Abstract

This review aims to determine whether there is a relationship between

perfectionism and parental bonding or attachment. Perfectionism has been associated

with psychological disorders since the 1980s but recent reviews have highlighted the

importance of classical perfectionism as a risk factor for developing and maintaining

psychological disorders. Despite the importance of perfectionism, there has been

little work that has investigated the origins of perfectionism. One factor that has been

mooted to be important to the development of perfectionism is attachment, however,

there has been very little examination of whether attachment or parental bonding is

indeed related to classical perfectionism. Addressing this question forms the focus of

this review. A literature search identified 10 peer-reviewed articles that had assessed

this relationship. Systematically reviewing this literature provided good evidence

that certain elements of perfectionism (Concern over mistakes, Doubt over Actions,

Discrepancy, and Self-Oriented Perfectionism) were correlated with early

experiences of parental bonding and attachment style whereas other elements of

perfectionism (High Standards, Personal Standards) were not. Whilst the review

identifies that there is a relationship between elements of perfectionism and parental

bonding or attachment it also highlights the need to fully consider the causal

direction and mechanisms underlying these relationships. The review also

highlighted the need to consider these relationships using diverse samples, and a

range of methods that allow for a test of causality. Future work should seek to

address these issues.

123

Introduction

Perfectionism

Burns (1980) defined perfectionism as a set of interrelated cognitions

including expectations, interpretations of events, and evaluations of the self and

others. At this time perfectionism was considered to be a pervasive and debilitating

condition. Perfectionism has been associated with a variety of physical and

psychological disorders including alcoholism, irritable bowel syndrome, abdominal

pain, and Type A coronary-prone behaviour, a range of anxiety disorders,

depression, anorexia nervosa, and obsessive compulsive personality disorder, (Burns,

1980; Pacht, 1984; Shafran & Mansell, 2001). Indeed Egan, Wade and Shafran

(2011) have argued that perfectionism is a transdiagnostic process and have provided

evidence that perfectionism is associated with, and can be helpful in treating, a range

of symptoms associated with psychological distress. Consequently, perfectionism is

an important area of study for clinical psychology and given this importance there

has been much discussion about how perfectionism should be defined.

Frost, Marten, Lahart, and Rosenblate, (1990) noted that perfectionism

research has suffered due the lack of a precise definition of perfectionism. Initial

work tended to focus on the presence of high standards, which as Hamachek, (1978)

notes does not distinguish between individuals who could be considered adaptive

perfectionists (those with high standards and are able to relax these standards when

the situation permits) and maladaptive perfectionists (those with high standards who

criticise their performance as never good enough and are unable to relax their

standards). Attempts to better define perfectionism have resulted in a proliferation of

measures and subscales, which has led Shafran and Mansell (2001) to call for better

measures of perfectionism that were more closely related to what they termed the

124

“classical concept” of perfectionism. Shafran and Mansell describe perfectionists as

individuals holding unrealistically high standards that they adhere to rigidly and note

that the individual’s self-worth is defined in relation to meeting these high standards.

Below we briefly review three commonly used measures of perfectionism to

highlight how perfectionism will be defined in this review.

Measures of perfectionism

The three measures of perfectionism that seem to be most commonly used in

the literature are: Multifactorial perfectionism scale (F-MPS; Frost et al., 1990),

Multifactorial perfectionism scale (H-MPS; Hewitt & Flett, 1991), and the Almost

Perfect Scale- Revised (APS-R: Slaney, Rice, Mobley, Trippi & Ashby, 2001).

F-MPS

The F-MPS consists of six factors, Concern over Mistakes (CoM), Doubts

about Actions (DoA), Personal Standards (PS), Parental Expectations (PE), Parental

Criticism (PC), and Organization (OR). CoM includes negative responses to

mistakes, equating mistakes with failure and beliefs about loosing other people’s

respect following failure. DoA refers to a sense that projects are not completed

properly. PS was defined as the setting of very high standards and the inflated

importance placed on these high standards for self-evaluation. PE was the tendency

to believe that one's parents set very high goals and PC the tendency to believe that

one’s parents are overly critical. OR was defined as including an emphasis on the

importance of order and the preference for order.

H-MPS

The H-MPS consists of three factors; Self-Oriented Perfectionism (SOP),

Other-Oriented Perfectionism (OOP) and Socially-Prescribed Perfectionism (SPP).

Hewitt and Flett (1991) describe SOP to consist of behaviours such as setting

125

exacting high standards for oneself, stringently evaluating and critiquing one's own

performance and striving to avoid failures and attain perfectionism. OOP consists of

beliefs and expectations about the capabilities of others. SPP is about the

individual’s perception of their significant other’s expectations of them being

perfect.

APS-R

The APS-R consists of 3 factors Discrepancy (D), High standards (HS) and

Order (OD). D measures the inconsistency between an individual’s high personal

standards and their perceptions of success in meeting those standards. HS is very

similar to the measure of PS described above, and OD is very similar to the measure

of OR described above.

Defining perfectionism

This brief review of perfectionism measures highlights the multitude of

constructs that have been considered to form part of perfectionism. It is notable that

across these different measures there are some similarities, a focus on having high

personal standards (PS, HS, SOP) appears on all measures. Likewise all measures

attempt to capture some element that relates to monitoring or evaluating whether

these standards have been met (CoM, DoA, SOP, D). Similarly, a number of the

scales consider expectations and criticism from others (SPP, PE and PC).

Shafran and Mansell (2001) note that SOP, PS and CoM come closest to

measuring the classical concept of perfectionism defined as the pursuit of

excessively high standards that are rigidly adhered to and define self-worth.

Therefore in this review perfectionism is defined in relation to Shafran and Mansell’s

definition (and subscale choice) of the classical concept of perfectionism. Shafran

and Mansell did not consider the APS-R in their review as it was published

126

contiguously, however, given the similarities between HS and PS and between SOP

and D noted above it is proposed that HS and D could equally be viewed as

measuring the classical concepts of perfectionism and so these will be included in

this review. Shafran and Mansell argue that DoA does not measure of perfectionism

given its overlap with OCD symptomology. Yet, there is good evidence that CoM

and DoA may be overlapping concepts: Stöber’s (1998) factor analysis of the F-MPS

suggests that CoM and DoA should be collapsed on to one factor, therefore DoA is

considered as an element of perfectionism in this review.

Subscales assessing expectations of perfection in others (e.g., OOP) and

subscales assessing the expectations and criticism of significant others and parents

(e.g., PE, PC and SPP) will not be included in this review. PE and PC in particular

are not appropriate because they are retrospective (Shafran & Mansell, 2001) and

confound the assessment of current perfectionism (Rhéaume et al., 2000).

Furthermore the way in which children are parented may be related to the

development of perfectionism (Frost et al., 1990; Rice, Lopez, & Vegara, 2005)

rather than a current assessment of perfectionism. The development of perfectionism

is considered below.

Development of perfectionism

There has been much theorising around the role of parenting in the

development of perfectionism. Barrow and Moore (1983) theorised that four

different types of early experience may contribute to the development of

perfectionism: overly critical and demanding parents; excessively high parental

standards and expectations (indirect criticism); perfectionistic parents modelling the

behaviour for children; and parental approval being absent, inconsistent or

conditional. Similarly, in a theoretical review chapter, Flett, Hewitt, Oliver and

127

Macdonald (2002) described four potential models that may explain the development

of perfectionism: 1) Social expectations, 2) Social learning, 3) Anxious rearing, and

4) Social reaction/harsh environment. The social expectations model suggests that

approval is contingent on the child being perfect, the social learning model suggest

that the child learns to imitate the parent who is perfectionistic, the anxious rearing

model suggests that anxious parenting promotes a focus on mistakes and negative

consequences of mistakes which leads to perfectionistic striving and anxious over-

concern. Finally, the social reaction/ harsh environment model suggests that a variety

of difficult early experiences such as abuse, maltreatment, withdrawal of love, or

exposure to a chaotic family environment may lead to the use of perfectionism as a

coping strategy. The literature reviewed above on the development of perfectionism

shares clear similarities with the literature on the development of attachment as Rice

and Mirazdeh (2000) have noted.

Attachment and perfectionism

An attachment is the representation of an affective bond between individuals,

and contains the behaviours that are useful for maintaining contact with the

significant other(s) (Ainsworth, 1989; Bowlby, 1969; 1988; Sroufe, 1979). Secure

attachments are optimal, and emerge when parents have been emotionally accessible

to, and nurturing of, their children. Insecure attachments result from unpredictable,

harsh, or unsupportive parent-child interactions or overly involved and anxious

parenting and these types of experiences are clearly represented in the theorizing

around the development of perfectionism as noted above. Children with insecure

attachment bonds have difficulties managing developmental challenges and

experience a variety of personal and interpersonal adjustment problems (Bowlby,

1973). It is possible that the development of perfectionism represents one of these

128

interpersonal adjustment problems.

Rice and Mirazdeh (2000) suggest that children who experience inconsistent

parental approval may develop an insecure attachment and learn to emphasize the

importance of their performance rather than their emotional needs supporting the

development of perfectionism. In such cases being perfect provides a means to

receive or maintain recognition from parents or to escape or avoid critical or punitive

responses from parents.

There have been few research studies that have considered the relationship

between attachment and perfectionism. Flett et al. (2002) reviewed a couple of

studies that were of variable quality; with aspects of parental relations being

contained in both the measure of perfectionism and attachment, and the use of

atypical measures of perfectionism or attachment. Although Flett et al. argue there is

a relationship between attachment and perfectionism this is based on evidence of an

association between SPP and fearful and preoccupied attachment styles. This

evidence is questionable because SPP is arguably not part of the classical conception

of perfectionism because it takes into account significant others expectations.

Therefore it is possible that this evidence of a relationship between SPP and

attachment is confounded.

The literature reviewed thus far suggests that there may be a relationship

between attachment and perfectionism but before this can be considered it is

important to define the scope of attachment experiences that will be included in this

review.

Measuring attachment

A child’s attachment to their primary caregiver can be assessed in children

during childhood using the strange situation (Ainsworth, 1978), and retrospectively

129

in adults using the adult attachment interview (AAI: George, Kaplan, & Main, 1984).

It is unlikely that there will be many, if any, studies where a child’s attachment style

has been assessed and recorded in childhood alongside concurrent or subsequent

measures of perfectionism, in part due to an absence of measures of perfectionism in

children. Furthermore, both the strange situation and the AAI require researchers to

undergo specific training in their administration and scoring and so consequently

studies using these methods may be infrequent in the literature. These assessment

tools also produce a categorical measure of attachment, and as such these measures

have been criticised because they do no take into account variation between

individuals within the same category (Mikiluncer & Shaver, 2007). Categorical

measures may make it more difficult to identify a relationship between attachment

and perfectionism, therefore this review will only consider dimensional measures of

attachment.

In an attempt to access early experiences of an individuals’ parenting

retrospectively some researchers have used the parental bonding instrument (PBI:

Parker, Tupling, & Brown, 1979). The PBI assesses the participant’s perceptions of

their mother’s and father’s parenting in terms of two dimensions, care and

overprotection. Although this is not a measure of attachment, the PBI does capture

the individual’s perceptions about their early experiences and these early experiences

of parenting are thought to contribute to the individual’s attachment style.

Attachment can also be measured in terms of an adult’s attachment to adults

that they are in close relationships with. There is evidence that attachment security

remains moderately stable over the first nineteen years of life (Fraley, 2002) and that

adult attachment is in part determined by early attachment to the primary caregiver

(Hazen & Shaver, 1987). Adult attachment can therefore be considered as a proxy

130

for early attachment.

In summary, perfectionism is an important concept for mental health and

well-being and the literature theorises a relationship between attachment and

perfectionism. However, there have been no been no published reviews nor

investigation of this relationship and consequently it is unclear if there is a

relationship. This review addresses this gap in the literature by investigating whether

there is a relationship between attachment orientation and perfectionism.

Method of the literature search

The literature search was conducted on the 15th January 2013 using Psycinfo,

Scopus and Web of Knowledge (WoK) databases. The search consisted of two

components combined with an “AND” the first was attachment OR “parental

bonding” and the second perfection*. There were no date limitations imposed on the

search. Differences between databases meant that in Psycinfo the search was

conducted on Abstracts, in Scopus the search was conducted on Article title, abstract

and keywords, in Web of Knowledge (WoK) the search was conducted on topic

(which considers titles, abstracts, keywords and keywords plus). In Scopus the

Health sciences, Life sciences, and Social sciences and humanities databases were

included, in WoK default settings were used which included all the databases

available to WoK, and in Psycinfo the Psycinfo database alone was used.

Inclusion/ Exclusion criteria

In order to be included in the final set of papers it was required that sources

were peer-reviewed to ensure that sources were of high quality. In Psycinfo this was

achieved by selecting the peer-reviewed only option. In Scopus and WoK this was

achieved by limiting document type to those most likely to have been peer-reviewed,

(in Scopus, article or review, and in WoK article, review, case report or clinical

131

trial). Sources identified by Scopus and WoK were checked and excluded if they

were not peer-reviewed or if this could not be determined. The search results were

combined and duplicates removed leaving 107 results. The inclusion/exclusion

criteria are explained briefly below, the number of results lost at each stage is shown

in Figure 1.

The full-text version of the source had to be available in English to be

included. All results were checked by the author and were excluded if the full-text

version was not available in English. The following inclusion criteria were addressed

in turn. The sources identified had to: 1) contain both a measure of attachment and a

measure of perfectionism in the same study and report the relationship between these

variables, 2) the measure of attachment had to be dimensional and there had to be

evidence of convergent validity with other attachment measures as defined by

Ravitz, Maunder, Hunter, Sthankiya, and Lancee (2010) and 3) the measure of

perfectionism had to include the at least one of the subscales noted to represent the

classical definition of perfectionism in the introduction (that is SOP, HS, PS, D,

CoM or DoA) or a composite of the subscales, and the measure had to have

convergent validity with another measure of perfectionism as noted by Enns and Cox

(2002). Papers that did not meet any of the above criteria were excluded. The final

set consisted of 10 papers that met these criteria.

Results

The results will be presented in two sections. The first section will consider

studies where early experiences of attachment or parental bonding have been

measured and the results are summarised in Table 1. The second section will

consider studies where adult attachment has been measured and the results are

132

Figure 1. Schematic representation of the literature selected for inclusion in the

review.

133

summarised in Table 2. In both sections the results will consider each perfectionism

subscale in turn, reviewing and evaluating what evidence there is for a relationship

between that perfectionism subscale and attachment. Subscales will be presented so

that those that are conceptually similar appear sequentially. All studies had large

sample sizes (n ranged from 138 to 604) and so statistical power was unlikely to be a

concern therefore sample sizes for each study are recorded only in the Tables.

Perfectionism and Parental Bonding Inventory

Personal standards (PS)

Enns, Cox, and Clara (2002) investigated the relationship between score on

the two indices of the PBI-care and overprotection (Parker et al., 1979) and PS in

Canadian University students. In this study the care dimension was assessed in terms

of lack of care1. Participants were asked to complete the PBI once for each parent

and the PBI scores were calculated by averaging responses for mother and father for

each participant. There were no significant correlations. PS was not associated with

either perceived lack of care from parents or perceived overprotection by parents.

Self-oriented perfectionism (SOP)

Two studies investigated the relationship between SOP and the two indices of

the PBI in outpatients with depression (Enns, Cox, & Larsen, 2000) or Canadian

college students (Enns et al., 2002). In both studies the PBI was completed once for

each of the participant’s parents. Enns et al. (2000) analysed their results by

considering PBI scores for each parent separately whereas Enns et al. (2002)

examined responses on the PBI towards parents by averaging responses for mother

and father. Furthermore, Enns et al. (2000) analysed responses by participant gender

1 The PBI is usually measured so that higher scores indicate care, however, in this study the authors chose to recode the answers so that higher scores indicated a lack of care.

134

whereas Enns et al. (2000) averaged across participant gender. For the sake of

brevity when reporting Enns et al. (2000) findings only women’s responses about

their mothers are considered. Women were chosen because they represented the

majority of participants and responses about mothers were chosen on the assumption

that women are usually the primary caregiver. It is noteworthy that Enns et al. report

that there were no statistical differences between men and women’s responses to the

PBI. In Enns et al. (2000) the PBI dimension is coded as specified by Parker et al.,

(1979) however, in Enns et al. (2002) the care dimension was assessed in terms of

lack of care1.

Enns et al. (2000; 2002) both demonstrate a weak correlation between

participant’s SOP and their perceptions of their mothers or parents being

overprotective. The findings on the care dimensions were also consistent: Enns et al.

(2000) found a weak negative correlation between perceptions of maternal care and

SOP and Enns et al. (2002) found a weak positive correlation between perceptions of

parental lack of care1 and SOP. Thus both studies support the idea that higher

perceptions of parental care are associated with lower levels of SOP.

The consistency between studies is noteworthy because one study considered

an average of parents PBI scores whereas the other considered the scores separately.

Furthermore, different samples were used in the two studies and the samples are

divergent in age (Enns et al., 2002 M= 20 years, SD= 4.4 years; Enns et al., 2000 M=

40 years, SD= 10.0 years). However, the findings on the care subscale should be

interpreted with some caution. In both studies the correlations between SOP and

(lack of) care are significant only at the p< .05 level, in Enns et al. (2000) they

compute 64 correlations therefore one would expect 3.2 of the 64 correlations to be

significant at the p> .05 level. Given the risk of a type 1 error, considering these

135

results with a more stringent p level would be advisable and would suggest that there

was no significant evidence of a relationship between SOP and (lack of) care on the

PBI.

Self-oriented perfectionism and Personal standards: Summary of

findings

SOP and PS subscales both measure HS so it is pertinent to consider these

findings alongside each other. There is little evidence that PS are associated with

early experiences of parents in terms of lack of care or overprotection. SOP however

is positively associated with perceptions of parental overprotection. The SOP

subscale assesses not only setting exacting high standards for oneself, but also

stringently evaluating and critiquing one's own performance and striving to avoid

failure. Therefore it is possible that it is these later aspects that are responsible for the

differing pattern of results on SOP compared to PS. Without further research that

adopts a more stringent approach, it is unclear whether perception of parental care

(or lack thereof) is significantly associated with SOP.

Concern over mistakes (CoM)

Enns et al. (2000; 2002) also considered the relationship between CoM and

the two dimensions of the PBI. Both studies found evidence of a significant (at

p< .01) weak to moderate positive correlation between perceptions of parental (or

maternal) overprotection and CoM. Again there was consistency between the two

studies on the care dimension with Enns et al. (2000) finding evidence for a weak

negative correlation between perceptions of maternal care and CoM and Enns et al.

(2002) finding evidence for a weak positive correlation between perceptions of

parental lack of care1 and CoM. It is interesting to note that the positive correlation in

Enns et al. (2002) is stronger and significant at a more stringent level on this

136

subscale providing good evidence that the greater the perception of parental lack of

care, the greater the CoM reported.

Doubts over action (DoA)

Enns et al. (2000; 2002) considered the relationship between DoA and the

two subscales of the PBI in clients with depression, and in Canadian college students

retrospectively. Whilst Enns et al (2000) found no evidence for an association

between female depressed participant’s DoA and their perceptions of care from their

mothers, Enns et al. (2002) found evidence of a significant positive but weak

association between college student’s DoA and their perceptions of lack of care from

their parents (Enns et al., 2002). Enns et al. (2000; 2002) both found evidence of a

(weak to moderate) positive correlation between participant’s DoA and their

perceptions of their parents being overprotective. This relationship was stronger in

the Enns et al. (2002) study where the participants were college students and the

perceptions of mothers and fathers being overprotective were averaged.

Collectively these two studies provide evidence that participants perceptions

of overprotection by parents is associated with DoA, however it is less clear to what

extent there is an association between care and DoA. It is unclear whether the

absence of a significant association between perceptions of maternal care and DoA

in the Enns et al. (2000) study is borne of considering just women and their relation

to their mothers because when perceptions of parental lack of care (mothers and

fathers averaged) are considered in Enns et al. (2002) there is a significant positive

association with DoA. However, it is unclear why this issue of gender and parents

would impact on the DoA subscale and not any of the other subscales. Likewise the

absence of an association between perceptions of care and DoA in the Enns et al.,

(2000) study (compared to the Enns et al., 2002 study) may result from the sample

137

comprising outpatients with depression but again it is unclear why the DoA subscale

would be the only subscale affected by the differing samples. One potential

explanation is provided by evidence that severity of depression is positively

correlated with DoA (Enns & Cox, 1999) Therefore, it is possible that the depressed

participants in the Enns et al. (2000) had inflated DoA scores and that there was little

variation in DoA scores, which would make it difficult to detect a correlation with

perfectionism. However, Enns et al. (2000) do not provide any mean scores for the

perfectionism subscales so it is difficult to assess whether this suggestion can explain

the divergence in results across the two samples. Furthermore given their conceptual

similarities it is unclear why the same issue would not have beset the results on CoM

also.

Concern over mistakes and Doubts over action: Summary of findings

In keeping with the idea that CoM and DoA subscales may represent slightly

different aspects of the same construct (Stöber, 1998), the findings for the two

subscales are considered alongside each other.

The findings for perceptions of parental (or maternal) overprotection on the

PBI and both CoM and DoA are very consistent. There are significant moderately

strong, positive correlations between parental overprotection and CoM and between

parental overprotection and DoA in both studies considered (see Table 1). The

results on the care dimension of the PBI are more variable. Whilst there is consistent

evidence for an association between perceptions of care with CoM across the two

studies, the relationship between perceptions of care and DoA is inconsistent.

138

Table 1. Correlation r-values between perfectionism and parental bonding subscales

SampleAttachment

target

PBI

(lack

of) care

PBI

over-

protectio

n

SOP

Enns et al. (2000) OD (n=138) Mother -.22* .36**

Enns et al. (2002) US (n=261) Parents (.14*) .18**

PS

Enns et al. (2002) US (n=261) Parents (.04) .10

CoM

Enns et al. (2000) OD (n=138) Mother -.25* .42**

Enns et al. (2002) US (n=261) Parents (.35**) .38**

DoA

Enns et al. (2000) OD (n=138) Mother -.10 .36**

Enns et al. (2002) US (n=261) Parents (.33**) .45**

Note- *p < .05, **p < .01, ***p < .001. US- University students. OD- Outpatients

with depression. Parentheses represent correlations where the PBI care measure was

coded to represent lack of care.

Overall summary of perfectionism and parental bonding inventory

The examination of the relationships between parental bonding and

dimensions of perfectionism has provided good evidence that participants’

perception of higher levels of parental overprotection is associated with higher levels

of SOP, CoM and DoA. Furthermore, this examination has also provided good

evidence that participants’ perception of higher levels of parental lack of care is

associated with higher levels of SOP, and CoM. Notably, there was no evidence that

perceptions of parental overprotection, nor parental care were associated with PS.

139

This is interesting and may reflect the notion that high personal standards are

associated with both adaptive and maladaptive perfectionism. Although most of the

findings reviewed are consistent and consider differing samples (in terms of

diagnosis and age), care should be taken when interpreting these results because they

represent the findings of only two separate studies, both of which were conducted in

the same country. These results may be generalizable to other Western cultures, but

there is not any evidence to support this yet.

Adult attachment

Personal standards (PS)

Two studies also investigated the relationship between personal standards

and adult attachment dimensions measured using the Experiences in Close

Relationships (ECR, Brennan, Clark, & Shaver, 1998) in University students in the

USA (Rice, Lopez, & Vergara, 2005) and in a sample of athletes recruited from

Universities and sports clubs from across the UK (Shanmugam, Jowett, & Meyer,

2012). In both studies there was no significant correlation between PS and

attachment anxiety nor between PS and attachment avoidance.

High standards (HS)

Two studies considered the relationship between HS and adult attachment

dimensions (measured using the ECR) in University students in the USA

(Iannantuono & Tylka, 2012 Rice et al. 2005). Both studies found no evidence of an

association between HS and attachment anxiety. Iannantuono and Tylka found no

evidence of an association between HS and attachment avoidance, and Rice et al.

found a significant negative correlation (at p < .05) between HS and attachment

avoidance. This correlation was very weak, but suggested that higher scores for HS

was associated with lower scores for attachment avoidance.

140

Ulu and Tezer (2010) considered the relationship between HS and adult

attachment dimensions using the Relationship Scales Questionnaire (RSQ; Griffin &

Bartholomew, 1994) in University students in Turkey. The RSQ asks participants to

rate their agreement with four subscales that represent four different adult attachment

styles secure (most optimal), preoccupied, dismissing and fearful. The RSQ had been

translated in to Turkish previously and was found to be reliable and valid (Sümer &

Güngör, 1999). The authors adapted the RSQ so that it produced dimensional

measures attachment anxiety and attachment avoidance by performing the following

calculations avoidance = (fearful + dismissive) − (preoccupied + secure); anxiety =

(preoccupied + fearful) − (secured + dismissing). This approach was supported by a

factor analysis (Kurdek, 2002). Ulu and Tezer found no evidence of an association

between HS and attachment anxiety nor any evidence of an association between HS

and attachment avoidance as measured by the RSQ. Considering the results from

these three studies these it appears that there is very little evidence of a correlation

between attachment dimensions and HS. However, given the adaption made to the

RSQ it is important to be wary in interpreting the absence of significant findings in

this study.

Only one study provided evidence of a correlation between attachment

avoidance and HS, and it is unclear if this correlation is meaningful. Rice et al.

measured 13 variables and considered the 78 correlations of these variables, without

adjusting the significance level to reduce type 1 error. Results are only reported in

terms of whether they reach conventional significance levels of p < .05, so it is

unclear whether these findings are significant if a more stringent p value is adopted.

Self-oriented perfectionism (SOP)

Reis and Grenyer (2002) investigated the relationship between SOP and

141

attachment in University students in Australia. Attachment to partner was measured

using the RSQ. There was no significant association between SOP and secure, nor

SOP and dismissing styles. However, there was a significant weak positive

correlation between SOP and preoccupied attachment subscales and between SOP

and fearful attachment subscales. More recent measures of adult attachment styles

(e.g., ECR; Brennan, et al., 1998) consider attachment to be best represented by the

crossing of two dimensions: attachment anxiety (anxiety about abandonment) and

attachment avoidance (avoidance of intimacy). These two dimensions can be applied

to the categorical views of attachment style such that: secure represents low

attachment anxiety, low attachment avoidance; preoccupied represents high

attachment anxiety, low attachment avoidance; dismissing represents low attachment

anxiety, high attachment avoidance; and fearful represents high attachment anxiety,

high attachment avoidance. Thus Reis and Grenyer’s findings could be interpreted to

suggest that higher attachment anxiety is associated with SOP because higher

attachment anxiety is associated with both preoccupied and fearful attachment styles.

However, the RSQ does not categorise individuals into attachment styles, it simply

asks them to rate their agreement with items pertaining to each of the four styles and

so this interpretation should be viewed cautiously. Furthermore, the RSQ has lower

reliability and validity than other measures of attachment according to Ravitz et al.

(2010) and within this study the internal reliability of the secure and dismissive

scales were low (alpha=.48 and .50 respectively), so this could explain the lack of

correlations in these subscales. In summary, the idea that attachment anxiety is

associated with SOP should be viewed cautiously unless this finding is replicated in

other studies.

Rice et al. (2005) investigated the association between SOP and adult

142

attachment style using the ECR (Brennan et al., 1998), in University students in the

USA. There was no significant association between SOP and attachment anxiety nor

between the SOP and attachment avoidance. Overall there is a mixed picture as to

whether SOP is associated with adult attachment. There was some tentative evidence

that there may be an association between attachment anxiety and SOP but Rice et al.

(2005) did not replicate this finding.

Composite Personal standards (SOP and PS)

Dunkley, Berg, and Zuroff, (2012) investigated the relationship between

perfectionism and attachment in Canadian University students. They used a two

factor model of perfectionism derived from a factor analysis of the Depressive

experiences Questionnaire, (DEQ: Blatt, D'Afflitti, & Quinlan, 1976) the F-MPS

(Frost et al., 1990) and the H-MPS (Hewitt & Flett, 1991). The two-factors were

Personal standards (comprised of SOP and PS) and Self-criticism (comprised of

DoA, COM, SPP, and DEQ). The self-criticism factor is comprised of subscales of

perfectionism that we have excluded from our analysis (SPP) for reasons outlined

above, therefore we focus only on the Personal standards composite scale (PScomp).

Dunkley et al. (2012) measured attachment using a version of the RSQ

(Griffin & Bartholomew, 1994) specifically focusing on the subscales from Collin

and Read (1990) which are fear of closeness, fear of dependency and fear of loss.

Brennan, et al., (1998) note that fear of closeness and fear of dependency both

correlate with attachment avoidance and fear of loss correlates with attachment

anxiety.

The results reveal significant weak positive correlations of PScomp with fear

of dependency and fear of closeness (both of which are associated with attachment

avoidance) and a significant but very weak correlation between PScomp and fear of

143

loss (associated with attachment anxiety). Consequently these results could be

viewed as evidence that PScomp is weakly associated with attachment avoidance

and very weakly associated with attachment anxiety.

Self-oriented perfectionism, personal standards and high standards:

Summary of findings

Given that SOP, PS and HS subscales are all measure high personal standards

it is pertinent to consider these findings in relation to each other. Examining the

results collectively there is little evidence that adult attachment dimensions are

related to high personal standards. Considering those studies that used the ECR and

the RSQ there is no evidence that attachment anxiety is associated with high

personal standards, and scant evidence that attachment avoidance is associated with

high personal standards (only one very weak negative correlation at p < .05 in a

study that did not adjust for multiple correlations).

The remaining studies that considered whether there is an association

between attachment and HS and PS used measures of attachment that are more

difficult to consider simply in terms of attachment avoidance and anxiety. Rice et al.

(2005) provided evidence that SOP was significantly positively associated with

preoccupied and fearful styles of attachment, albeit weakly, and this pattern may

reflect that higher levels of attachment anxiety are associated with higher personal

standards. Dunkley et al. (2012) found a weak but highly significant association

between a composite of SOP and PS with fear of dependency and fear of closeness

(both avoidance), and weaker and less significant relationship with fear of loss

(anxiety).

Collectively these findings seem to suggest that there is little evidence for

associations between high personal standards and attachment. However, there is

144

better evidence of an association between SOP and attachment anxiety and

avoidance. SOP is not a pure measure of high standards, it also measures stringently

evaluating and critiquing one's own performance and striving to avoid failure. This

suggests that it may be the additional concepts captured by SOP that are not apparent

in HS or PS subscales, which may be responsible for this relationship. However, this

interpretation should be viewed cautiously because the studies that provide evidence

for a relationship between SOP and attachment both use less reliable measures of

attachment (Ravitz et al., 2010). It is possible that the discrepancy subscale of the

APS may be similar to the SOP in terms of capturing the evaluation and critique of

an individual’s own performance and so considering the results on the discrepancy

subscale may provide supportive evidence for a relationship between monitoring and

evaluating performance and attachment.

Discrepancy (D)

Four studies considered the relationship between D and adult attachment

dimensions (using the ECR) in University students in the USA (Iannantuono &

Tylka, 2012; Rice et al., 2005; Wei, Heppner, Russell, & Young, 2006; Wei,

Mallinckrodt, Russell, & Abraham, 2004). All of these studies found significant

positive but weak correlations between attachment avoidance and D and all

demonstrated significant positive correlations (that varied in strength from weak to

moderately strong) between attachment anxiety and D (Iannantuono & Tylka, 2012;

Rice et al., 2005; Wei et al., 2004; 2006) 2. Ulu and Tezer (2010) also found evidence

2 Wei et al. (2004; 2006) were testing a structural equation model with other variables and so they split attachment avoidance and attachment anxiety variables in to 3 parcels. Therefore for Wei et al. (2004), there were three correlations with each attachment dimension for each perfectionism subscale. In Wei et al. (2006), discrepancy was also split into 3 parcels resulting in nine correlations considering discrepancy and attachment anxiety and nine correlations considering discrepancy and attachment avoidance.

145

of significant weak positive correlation between D and attachment avoidance and

significant moderate positive correlation between D and attachment anxiety when

attachment was measured using an adapted RSQ in a sample of Turkish University

students.

The studies reviewed provide very good evidence of significant positive

correlations between discrepancy and both attachment anxiety and attachment

avoidance. The relationship between discrepancy and attachment anxiety seems to be

stronger than the relationship between discrepancy and attachment avoidance. This

evidence is consistent with the same relationships found with two different measures

of attachment, and in both US and Turkish samples, however, it is noteworthy that

all samples are University students and so replication with an alternative sample to

test the generalizability of this finding would be useful.

Discrepancy and Self-oriented perfectionism: Summary of findings

It is worth comparing the findings on discrepancy and SOP because the

subscales share some conceptual similarities. Although the findings on SOP are

tentative there was evidence that SOP is significant positively correlated with higher

levels of attachment anxiety (preoccupied and fearful) and PScomp (SOP and PS)

scale is significantly positively correlated with attachment anxiety (fear of loss) and

with attachment avoidance (fear of dependency and fear of closeness). These

findings are strengthened by the conceptual replication of these findings on the

discrepancy subscale.

Concern over mistakes (CoM)

Rice et al. (2005) and Wei et al. (2004) considered the relationship between

CoM and adult attachment dimensions (measured using the ECR) in University

students in the USA. Both studies revealed significant weak positive correlations

146

between CoM and attachment anxiety and CoM and attachment avoidance.

Doubts over action (DoA)

Two studies considered the relationship between DoA and adult attachment

dimensions (measured using the ECR) in University students in the USA (Rice et al.,

2005; Wei et al., 2004). Both studies revealed significant weak positive correlations

between DoA and attachment anxiety and DoA and attachment avoidance.

Concern over mistakes and Doubts over action: Summary of findings

In keeping with Stöber’s (1998) proposal that CoM and DoA are related

constructs it is pertinent to consider these results alongside each other. All of the

results reviewed on these subscales provide consistent evidence of significant weak

positive associations between CoM and DoA with both attachment anxiety and

attachment avoidance. However, it is notable that all of these studies used the ECR

and used University students as a sample so these results are replicable but lack

generalizability.

Overall summary of perfectionism and adult attachment

The examination of the relationships between attachment and dimensions of

perfectionism has provided good evidence that both higher levels of attachment

anxiety and avoidance are associated with higher levels of CoM, DoA and D. There

was some evidence that the relationship between attachment anxiety was stronger,

particularly in relation to the discrepancy subscale. Notably, there was little evidence

that either attachment anxiety or avoidance were associated with PS or HS but there

was some evidence that attachment dimensions are associated with SOP. This is

interesting and may reflect the notion that high standards are associated with both

147

Table 2. Correlation r-values between perfectionism and adult attachment subscales

Sample Attachment target

ECR-anxious

ECR-avoidance

RSQ-anxious

RSQ-avoidance

Fear of dependency

Fear of loss

Fear of closeness Secure Preoccup

iedDissimissing Fearful

SOP

Reis and Grenyer (2002) US(n=245) Partner  - - - - - - - -.09 .24** -.02 .17***

Rice et al. (2005) Students(n=241) Partner .08 .03 - - - - - - -  - -

PS

Rice et al. (2005) Students(n=241) Partner .04 <. 01 - - - - - - - - -

Shanmugam et al. (2012)US, athletes(n=411)

general, partners, coaches teammates

.40 .80 - - - - - - - - -

HSIannantunono and Tylka (2012)

US(n=249) Partners -.08 -.05  - -  - -  - - - - -

Rice et al. (2005) Students(n=241) Partner -.08 -.12* - - - - - - - - -

Ulu and Tezer, (2010) US(n=604) Partner - - .10 -.20 - - - - - - -

PScomp

Dunkley et al. (2012) US(n=163)

Close others - - - - .36*** .16* .30*** - - - -

DIannantunono and Tylka (2012)

US(n=249) Partners .44*** .26***  - - - - - - - - -

Rice et al. (2005) Students(n=241) Partner .36* .36* - - - - - - - - -

Sample Attachme ECR- ECR- RSQ RSQ- Fear Fear Fear Secu Preocc Dissi Fearf

148

nt target anxious

avoidance

-anxious

avoidance

of dependency

of lossof closeness

re upied missing ul

Ulu and Tezer (2010)US(n=604)

Partner - - .40** .25** - - - - - - -

Wei et al. (2006)US(n=372)

Partner

.48***

.48***

.49***.49*** .49*** .49***.52*** .53*** .55***

.17***

.18***.19*** .19***.20*** .21***.21*** .22***.23***

 - - - - - - - - -

Wei et al. (2004) US(n=310)

Partner .38** .39**.41**

.32**

.24**

.21**- - - - - - - - -

COM

Rice et al. (2005)

Students(n=241)

Partner .28* .25*  - - - - - - - - -

Wei et al. (2004)US(n=310)

Partner.37**.34** .33**

.27**

.21**

.20**- - - - -  - - - -

DOA

Rice et al. (2005)

Students(n=241)

Partner .35* .28*  - - - - - - - - -

Wei et al. (2004)US(n=310)

Partner .32**.33**.32**

.34**

.27**

.23**- - - - - - - - -

Note- *p < .05, **p < .01, ***p < .001. US-University students. Dashes represent correlations that were untested or unreported.

149

adaptive and maladaptive perfectionism. Although most of the findings reviewed are

very consistent, they predominantly consider University students in Western cultures

and so care should be taken when interpreting these results because they warrant

replication in more diverse samples.

Discussion

In summary, the literature reviewed provides good evidence for a relationship

between some elements of perfectionism and parental bonding and good evidence for

a relationship between some elements of perfectionism and adult attachment.

Specifically, Self-oriented perfectionism (SOP), Concern over mistakes (CoM) and

Doubts over action (DoA) are positively associated with perceived parental

overprotection, and SOP and CoM are positively associated with perceived parental

lack of care. Similarly, CoM, DoA, and Discrepancy (D) are positively associated

with attachment avoidance and anxiety. There is some evidence that SOP is positively

associated with attachment dimensions, including fear of loss, dependency and

closeness, and preoccupied, and fearful attachment styles. Interestingly, there is little

evidence that Personal standards (PS), or High standards (HS) are associated with

parental bonding or adult attachment. It is interesting to note the consistency of the

findings on parental bonding and attachment, especially when it is noted that there

were no measures of discrepancy taken in the studies of parental bonding. This

consistency is notable because the parental bonding inventory is not a measure of

attachment per se, but a retrospective measure of early experience that may capture

some of the experiences that lead to the formation of attachment security or

insecurity.

Theoretical implications

Defining perfectionism

150

Hamachek (1978) highlighted that high standards were necessary but not

sufficient to define maladaptive perfectionism. Similarly Shafran and Mansell (2001)

noted that the perfectionism was characterised by the pursuit of excessively high

standards and a rigid adherence to them. Recently Dunkley, Zuroff, & Blankstein,

(2003) factor analysed the F-MPS, H-MPS with the DEQ in an attempt to identify the

factors underlying perfectionism. This produced 2 factors: High standards (SOP and

PS) and Self-criticism (DEQ, DoA, CoM, SPP). Interestingly those measures of

perfectionism that tap self-critical/evaluative perfectionism seem to be more

associated with psychological problems than perfectionism defined by high standards

(Dunkley, Blankstein, Masheb, & Grilo, 2006; Stöber & Otto, 2006). This review

considers different subscales to Dunkley et al., and seems to highlight three

dimensions of perfectionism High standards (PS and HS), Monitoring, evaluation and

criticism of whether standards are met (D and SOP), and Concerns and doubts (CoM

and DoA). Interestingly the last two factors seem to capture self-critical evaluative

aspects of perfectionism and the subscales proposed to be in these factors are those

most reliably associated with attachment dimensions, supporting the idea that insecure

attachment is associated with maladaptive perfectionism. Considering the findings of

this review in a broader theoretical context suggests that high standards are not

associated with adult attachment nor with maladaptive perfectionism, but that self-

critical and evaluative aspects of perfectionism are associated with both attachment

and maladaptive perfectionism.

Development of perfectionism

The findings of this review strongly suggest that aspects of maladaptive

perfectionism are associated with adult attachment style. Scholars have previously

argued that a variety of different types of early experience may contribute to the

151

development of perfectionism, (Barrow & Moore, 1983; Flett et al., 2002).

Specifically, anxious rearing where parents promote a focus on mistakes and are

overly critical, or provide inconsistent approval could all lead to the development of

an anxious attachment style. Whereas, parental approval being absent or inconsistent,

a harsh environment with difficult early experiences such as abuse, maltreatment or

withdrawal of love, could lead to the development of a more avoidant attachment

style. Indeed in both cases perfectionism could develop as a coping strategy in order

to try and maintain and achieve love and affection, or in an attempt to avoid

punishment or withdrawal of love. However, the findings of this review do not

demonstrate a causal direction or explain this relationship but represents an important

step in considering the relationship between attachment and perfectionism.

Limitations and implications for future research

Diversity of samples

Both studies that considered parental bonding and perfectionism were

conducted in Canada, either with depressed patients or with University students. This

variation in samples allowed for some variation in age, but in both samples women

were a majority. All of the studies that considered adult attachment and perfectionism

were conducted in University students, this means the samples were predominantly

well-educated, White, female and between the ages of 18-22. Furthermore, most of

the studies were conducted in Westernised cultures, therefore, future work should

look to increase the diversity of the populations considered, targeting populations that

are experiencing psychological distress and those from different cultural backgrounds.

Self-report measures

All of the studies included in this review used self-report measures of parental

bonding, attachment, and perfectionism. This is in part borne of the inclusion criteria

152

identified. A dimensional measure of attachment was specified, however, a by-

product of this decision is that this may have excluded studies that contained

interviewer-assessed measures of attachment because these are typically scored

categorically. Similarly all of the measures of perfectionism included in this report are

self-report measures of perfectionism, this is representative of the measures of

perfectionism available. The use of self-report measures means that biases such as

social desirability can affect the validity of the results. Future work should consider

whether the relationship between attachment, parental bonding and perfectionism is

replicated on interviewer-assessed measures of attachment.

Correlational designs

All of the results reviewed are correlational in nature. As a result the causal

directions and specific mechanisms are unclear and unspecified. Some of the studies

reviewed used more complex regression modelling such as SEM or moderated

regressions in an attempt to better specify these relationships. However, studies that

use these types of techniques tended to address questions that are outside the remit of

this review. It is clear that some elements of perfectionism are related to parental

bonding and adult attachment, however the causal pathways between attachment and

perfectionism warrant further research.

Cross-sectional data

All of the studies reviewed considered cross-sectional data on attachment and

perfectionism. Given that attachment style develops early in life and is moderately

stable until early adulthood (Fraley, 2002) a longitudinal design may be better able to

capture the causal relationships between attachment and perfectionism. The studies

that use the PBI arguably capture some early experience as the PBI is a retrospective

measure of earlier parenting. However, this measure is not without fault and its

153

retrospective nature means that it may be biased by the individual’s memory of their

parenting. Future work should attempt to consider the relationships between

attachment and perfectionism in a longitudinal manner to help elucidate the causal

relationship.

Experimental investigations

None of the studies identified by this review adopted an experimental

approach. An experimental investigation could provide a means of assessing the

causal relationship between attachment and perfectionism. Although it is not possible

to manipulate attachment style, a variety of mechanisms have been used to activate

attachment style, usually via priming an attachment threat (e.g., Crisp et al., 2009;

Mikiluncer, Birnbaum, Woddis & Nachmias, 2000). Attachment moderated

behaviours are often observed under conditions of attachment-related threat (Fraley &

Shaver, 1998) so it would be possible to investigate whether following an attachment

threat, those individuals with an insecure attachment style, endorse aspects of

perfectionism more strongly and engage in more perfectionistic behaviour.

Conclusion

This review assessed whether perfectionism was related to early experiences

of parenting and attachment style. The literature reviewed provided evidence that

Concern over mistakes, Doubts over actions, Discrepancy, and Self-oriented

perfectionism were correlated with early experiences of parental bonding and

attachment style whereas high standards and personal standards were not. However,

the studies reviewed were limited in their generalizability and the causal nature of the

relationship is underspecified and so future work should seek to address these issues.

154

References

Ainsworth. M. D. (1978). Patterns of Attachment: A Psychological Study of the

Strange Situation. Hillsdale, NJ: Lawrence Erlbaum Associates.

Ainsworth, M. D. S. (1989). Attachments beyond infancy. American Psychologist, 44,

709-716. doi: 10.1037%2F%2F0003-066X.44.4.709

Barrow, J. C., & Moore, C. A. (1983). Group interventions with perfectionist

thinking. Personnel and Guidance Journal, 61, 612– 615.

doi:10.1111%2Fj.2164-4918.1983.tb00008.x

Blatt, S. J., D'Afflitti, J. P., & Quinlan, D. M. (1976). Experiences of depression in

normal young adults. Journal of Abnormal Psychology, 85, 383-389.

doi:10.1037%2F%2F0021-843X.85.4.383

Bowlby, J. (1969). Attachment and loss, Vol. 1: Attachment. New York: Basic Books.

Bowlby, J. (1973). Attachment and loss, Vol. 2: Separation. London: Hogarth Press.

Bowlby, J. (1988). A secure base. New York: Basic Books.

Brennan, K. A., Clark, C.L., & Shaver, P. R. (1998). Self-report measurement of adult

attachment: an integrative overview. In J. A. Simpson & W. S. Rholes (Eds.)

Attachment theory and close relationships (pp. 46–76). New York: Guilford.

Burns, D. D. (1980). Feeling good: the new mood therapy. New York: New American

Library.

Collins, N. L., & Read, S. J. (1990). Adult attachment, working models, and

relationship quality in dating couples. Journal of Personality & Social

Psychology, 58, 644–63. doi: 10.1037%2F%2F0022-3514.58.4.644

Crisp, R. J., Farrow, C. V., Rosenthal, H. E. S., Walsh, J., Blisset, J., & Penn, M. K.

(2009). Interpersonal attachment predicts identification with groups. Journal of

Experimental Social Psychology, 45, 115–122. doi10.1016/j.jesp.2008.09.006

155

Dunkley, D. M., Zuroff, D. C., & Blankstein, K. R. (2003). Self-critical perfectionism

and daily affect: Dispositional and situational influences on stress and coping.

Journal of Personality and Social Psychology, 84, 234–252. doi:10.1037%2F

%2F0022-3514.84.1.234

Dunkley, D. M., Berg, J.-L., & Zuroff, D. C. (2012). The Role of Perfectionism in

Daily Self-Esteem, Attachment, and Negative Affect. Journal of Personality 80

(3), 633-663. doi: 10.1111/j.1467-6494.2011.00741.x

Dunkley, D. M., Blankstein, K. R., Masheb, R. M., & Grilo, C. M. (2006). Personal

standards and evaluative concerns dimensions of “clinical” perfectionism: A

reply to Shafran et al. (2002, 2003) and Hewitt et al. (2003). Behaviour Research

and Therapy, 44, 63–84. doi:10.1016%2Fj.brat.2004.12.004

Egan, S. J., Wade, T. D., & Shafran, R. (2011). Perfectionism as a transdiagnostic

process: A clinical review. Clinical Psychology Review, 31, 203–212. doi:

0.1016/j.cpr.2010.04.009.

Enns, M.W., & Cox, B. J. (1999). Perfectionism and depression symptom severity in

major depressive disorder. Behavior Research & Therapy, 37, 783-94.

doi:10.1016%2FS0005-7967%2898%2900188-0

Enns, M. W., & Cox, B. J. (2002). Nature and assessment of perfectionism. In G. L.

Flett & P.L. Hewitt (Eds.), Perfectionism: Theory, research and treatment (pp.

33−62).Washington, DC: American Psychological Association.

Enns, M.W., Cox, B. J. & Clara, (2002). Adaptive and maladaptive perfectionism:

developmental origins and association with depression proneness. Personality

and Individual Differences 33, 921–935. doi:10.1016%2FS0191-

8869%2801%2900202-1

Enns, M. W., Cox, B. J., & Larsen, D. K. (2000). Perceptions of parental bonding and

156

symptom severity in adults with depression: Mediation by personality

dimensions. Canadian Journal of Psychiatry, 45, 263–268. doi: not available.

Fraley, R. C. (2002). Attachment stability from infancy to adulthood: Meta-analysis

and dynamic modeling of developmental mechanisms. Personality and Social

Psychology Review, 6, 123-151. doi:10.1207%2FS15327957PSPR0602_03

Frost, R. O., Marten, P., Lahart, C. M., & Rosenblate, R. (1990). The dimensions of

perfectionism. Cognitive Therapy and Research, 14, 449–468. doi:

10.1007%2FBF01172967

George, C., Kaplan, N. & Main, M. (1996). Adult attachment interview. Unpublished

manuscript.

Griffin, D. W., & Bartholomew, K. (1994). The metaphysics of measurement: The

case of adult attachment. In K. Bartholomew and D. Perlman (Eds.) Advances in

personal relationships: Vol. 5. Attachment process in adulthood (pp. 17–52).

London: Kingsley.

Flett, G. L., Hewitt, P. L., Oliver, J. M., & Macdonald, S. (2002). Perfectionism in

Children and Their Parents: A Developmental Analysis. In G. L. Flett & P. L.

Hewitt (Eds.). Perfectionism: Theory, research, and treatment (pp. 89-132).

Washington, DC: American Psychological Association.

Hamachek, D. E. (1978). Psychodynamics of normal and neurotic perfectionism.

Psychology: A Journal of Human Behavior, 15, 27– 33. doi: not avaliable

Hazan, C., & Shaver, P. (1987) Romantic Love conceptualized as an attachment

process. Journal of Personality and Social Psychology, 52(3), 511-524.

doi:10.1037%2F%2F0022-3514.52.3.511

Hewitt, P. L., & Flett, G. L. (1991). Perfectionism in the self and social contexts:

Conceptualization, assessment, and association with psychopathology. Journal of

157

Personality and Social Psychology, 60, 456–470. doi:10.1037%2F%2F0022-

3514.60.3.456

Iannantuono, A. C. & Tylka, T. L. (2012). Interpersonal and intrapersonal links to

body appreciation in college women: An exploratory model. Body Image, 9, 227–

235. doi:10.1016/j.bodyim.2012.01.004.

Kurdek, L. A. (2002). On being insecure about the assessment of attachment styles.

Journal of Social and Personal Relationships, 19, 811–834. doi:

10.1177%2F0265407502196005

Mikulincer, M., Birnbaum, G., Woddis, D., & Nachmias, O. (2000). Stress and

accessibility of proximity-related thoughts: Exploring the normative and

intraindividual components of attachment theory. Journal of Personality and

Social Psychology, 78, 509–523. doi:10.1037/0022-3514.78.3.509.

Mikulincer, M., & Shaver, P. R. (2007). Measurement of attachment-related

constructs in adulthood. In Attachment in adulthood: structure, dynamics, and

change (pp. 81–115). New York: Guilford Press.

Pacht, A. R. (1984). Reflections on perfection. American Psychologist, 39, 386-390.

doi:10.1037%2F%2F0003-066X.39.4.386

Parker, G., Tupling, H., & Brown, L. B. (1979). A parental bonding instrument.

British Journal of Medical Psychology, 52, 1–10. doi:10.1111%2Fj.2044-

8341.1979.tb02487.x

Ravitz, P., Maunder, R., Hunter, J., Sthankiya, B., & Lancee, W. (2010). Adult

attachment measures: A 25-year review. Journal of Psychosomatic Research 69,

419–432. doi:10.1016/j.jpsychores.2009.08.006

Reis, S., & Grenyer, B. F. S. (2002). Pathways to anaclitic and introjective depression.

Psychology & Psychotherapy: Theory, research & Practice, 75, 445-459. doi:

158

10.1348%2F147608302321151934

Rhéaume, J., Freeston, M. H., Ladouceur, R., Bouchard, C., Gallant, L., Talbot, F., &

Vallières, A. (2000). Functional and dysfunctional perfectionists: Are they

different on compulsive-like behaviors? Behaviour Research & Therapy, 38,

119–128. doi: 10.1016%2FS0005-7967%2898%2900203-4

Rice, K. G., Lopez, F. G., & Vergara, D. (2005). Parental/social influences on

perfectionism and adult attachment orientations. Journal of Social and Clinical

Psychology, 24, 580–605. doi:10.1521%2Fjscp.2005.24.4.580

Rice, K. G., & Mirzadeh, S. A. (2000). Perfectionism, attachment, and adjustment.

Journal of Counseling Psychology, 47, 238–250. doi: 10.1037%2F%2F0022-

0167.47.2.238

Shafran, R., & Mansell, W. (2001). Perfectionism and psychopathology: A review of

research and treatment. Clinical Psychology Review, 21, 879−906. doi:

10.1016%2FS0272-7358%2800%2900072-6

Shanmugam, V., Jowett, S., & Meyer, C. (2012). Eating Psychopathology amongst

athletes: Links to current attachment styles. Eating Behaviors, 13, 5-12. doi:

10.1016%2Fj.eatbeh.2011.09.004

Slaney, R. B., Rice, K. G., Mobley, M., Trippi, J., & Ashby, J. S. (2001). The revised

Almost Perfect Scale. Measurement and Evaluation in Counseling and

Development, 34, 130–145. doi: not available.

Stöber, J. (1998). The Frost Multidimensional Perfectionism Scale: More perfect with

four (instead of six) dimensions. Personality and Individual Differences, 24(4),

481-491. doi: 10.1016%2FS0191-8869%2897%2900207-9

Stöber, J., & Otto, K. (2006). Positive conceptions of perfectionism: Approaches,

evidence, challenges. Personality and Social Psychology Review, 10, 295–319.

159

doi: 10.1207%2Fs15327957pspr1004_2

Sroufe, L. A. (1979). The coherence of individual development: Early care,

attachment, and subsequent developmental issues. American Psychologist, 34,

834-841. doi:10.1037%2F%2F0003-066X.34.10.834

Sümer, N., & Güngör, D. (1999). Yetişkin bağlanma stillerinin Türk örneklemi

üzerinde psikometrik değerlendirmesi ve kültürler arası bir karşılaştırma. Türk

Psikoloji Dergisi, 14(43), 71-106. doi: not avaliable

Ulu, İ. P., & Tezer, E. (2010). Adaptive and maladaptive perfectionism, adult

attachment, and big five personality traits. The Journal of Psychology:

Interdisciplinary and Applied, 144, 327-340.

doi:10.1080%2F00223981003784032

Wei, M., Mallinckrodt, B., Russell, D. W., & Abraham, T. W. (2004). Maladaptive

perfectionism as a mediator and moderator between attachment and negative

mood. Journal of Counseling Psychology, 51, 201–212. doi: not available.

Wei, M., Heppner, P. P., Russell, D. W., & Young, S. K. (2006). Maladaptive

perfectionism and ineffective coping as mediators between attachment and

subsequent depression: A prospective analyses. Journal of Counseling

Psychology, 53, 67-79. doi: not available.

160

Clinical Experience

An overview of the clinical experiences gained on placement.

161

Year 1

I spent the first year of training in a Community Mental Health and Recovery

Service for working age adults. I worked with individuals with diagnoses of Complex

trauma, PTSD, bipolar affective disorder, low self-esteem, borderline personality

disorder, avoidant personality disorder, psychosis, obsessive compulsive disorder-

hoarding, social phobia, agoraphobia, and emetophobia. I worked with individuals

with underlying autistic spectrum disorders, and with individuals whose attachment

style and relationships were affected by an attachment figure having a probable

autistic spectrum condition. I used a variety of therapeutic interventions including

psychoeducation, CBT, Dialectical Behaviour Therapy, Mindfulness, and

Compassion-focussed therapy. I co-facilitated two groups: a four-week Understanding

Personality Disorder group and an eight-week Mindfulness group. I also conducted a

neuropsychological assessment to consider whether a woman with longstanding

hoarding disorder was experiencing the onset of dementia.

I taught the multi-disciplinary team (MDT) about motivational interviewing

and taught carers about communication skills. I also conducted an audit to consider

whether the team were NICE guideline compliant in their management of individuals

with schizophrenia/psychosis and shared the findings with the team. I attended a

variety of team meetings including business meetings and allocations meetings, and

team reflective meetings. I also attended local psychology meetings, carer support

groups and the Forum for Carers and people who use services.

Year 2

My second placement was in a Child and Adolescent Mental Health Service. I

worked with a number of children, adolescents, parents and teachers to help support

young people with a range of presentations including chronic pain, anxiety, low

162

mood, ADHD, autistic spectrum conditions, obsessive compulsive disorder, and

anger. I used both second wave CBT approaches and third approaches including ACT

and Mindfulness.

I conducted two neuropsychological assessments: one with an adolescent girl

who was presenting with anger, but had attention and memory difficulties and one

with an adolescent boy with a history of epilepsy who described having difficulties

with his memory. I also conducted a school observation and formulated the child’s

difficulties and shared this with the parents and school. I also supervised an assistant

psychologist to complete a school observation. I taught Year 10 pupils about anxiety

in a school-setting with a primary mental health worker. I also attended regular team

meetings, and psychology local area meetings and service-user group meetings and

observed social work consultation meetings.

My third placement was in an Older People’s Community Mental Health

Team. I worked with a gentleman who had been experiencing depression since his

retirement using a CBT approach and with a woman with bipolar affective disorder

who was experiencing depression and anxiety. On an inpatient ward for older adults

with mental health difficulties I provided weekly CBT groups for anxiety, depression

and grief and a recovery group. I taught the ward staff about the nature of the groups

and what would make an appropriate referral.

I also conducted neuropsychological assessments of two women who were

concerned that they had developed dementia. Both women had longstanding mental

health difficulties. I also helped to setup team formulation meetings to help facilitate

the MDT understanding of the service-users difficulties.

Year 3

My fourth placement was in a Chronic Pain out-patient team. In this

163

placement I worked with a number of individuals, and couples with a variety of types

of pain disorders, and co-morbid diagnoses including complex regional pain

syndrome, chronic fatigue, chronic pelvic pain, fibromyalgia, anxiety, depression, and

borderline personality disorder. I used both second-wave CBT approaches and third

wave approaches such as mindfulness and ACT. I co-facilitated an eight-week Living

with Pain course with a physiotherapist and my supervisor.

I lead a case formulation and discussion group with the MDT and designed

and delivered a workshop on Depression and chronic pain to service-users. I also

worked in more generic health settings. I observed a neuropsychologist conduct an

assessment on the stroke ward and I co-facilitated a level two psychological skills

training in CBT for oncology nurses. In addition I worked therapeutically with a

diabetic woman to address her fear of needles. I also supervised an assistant

psychologist to develop some leaflets as a piece of service development work.

My final placement was in Community Team for People with Learning

Disabilities. Within this placement I worked with individuals, their families and

carer’s and with day services and residential homes. I worked with people with a

range of learning difficulties including Angelman’s syndrome, Down’s syndrome,

foetal alcohol spectrum conditions, and Autism and other conditions including

epilepsy, hydrocephalus, and bipolar affective disorder. I developed and shared

formulations with a number of clients for a variety of difficulties including loneliness,

gambling, and anger and conducted a systemic intervention with a care home. I also

conducted a range of neuropsychological assessment to assess service-users strengths

and weaknesses as part of dementia, and autism assessments. I supervised an assistant

psychologist to complete a joint piece of work to understand a challenging behaviour.

164

Assessments

165

Year I Assessments

PROGRAMME COMPONENT

TITLE OF ASSIGNMENT

Fundamentals of Theory and Practice in Clinical Psychology (FTPCP)

Short report of WAIS-IV data and practice administration

Practice case report Assessment and formulation of a young woman with PTSD.

Problem Based Learning – Reflective Account

Reflective account of personal and professional learning from the process of completing a group problem-based learning task.

Research – Literature Review

Perfectionism, parental bonding and attachment.

Adult – case report Assessment and intervention of an individual with complex trauma.

Adult – case report Neuropsychological assessment of memory difficultiesResearch – Qualitative Research Project

What do First Year Clinical Psychology Trainees Think about Clinical Supervision?

Research – Major Research Project Proposal

Can compassionate-imagery reduce perfectionist’s experience of shame?

Year II Assessments

PROGRAMME COMPONENT

TITLE OF ASSESSMENT

Research - SRRP NICE guidelines audit of psychological and psychosocial interventions being offered to those with Schizophrenia in a CMHRS.

Research Research Methods and Statistics testProfessional Issues Essay

Critically explore the statement that clinical psychology should “move away from psychiatric diagnoses…., which have significant conceptual and empirical limitations, and develop alternative approaches which recognise the centrality of the complex range of life experiences…” (Position Statement on the Classification of Behaviour and Experience in Relation to Functional Psychiatric Diagnoses: Time for a Paradigm Shift, BPS, 2013).

Problem Based Learning – Reflective Account

Reflective account of personal and professional learning from the process of completing a group problem-based learning task.

Child and Family – Case Report

Assessment and intervention of a young person with chronic pain, fatigue and anxiety.

Personal and Professional Learning

A reflective account of the personal and professional learning discussion group process.

166

Discussion Groups – Process AccountOlder People – Oral Presentation of Clinical Activity

Case consultation skills.

Year III Assessments

PROGRAMME COMPONENT

ASSESSMENT TITLE

Research – MRP Portfolio

Perfectionism, Failure and Self-conscious Emotions: A role for Self-compassion?

Personal and Professional Learning – Final Reflective Account

On becoming a clinical psychologist: A retrospective, developmental, reflective account of the experience of training.

Specialist – Case Report

Exposure therapy with a young woman with needle phobia and type 1 diabetes.

167