personality measurement and testing: an overview

Personality Measurement andTesting: An Overview

Gregory J. Boyle, Gerald Matthews and Donald H. Saklofske

A colleague recently remarked:

Psychologists who specialize in the study of person-ality and individual differences spend a lot of timecoming up with various descriptions of people, likeMachiavellianism, external locus of control, open-ness to experience, and neuroticism. Even moreeffort is spent trying to measure these ideas withtests like the MMPI-2, brief anxiety scales, andRorschach Inkblots. But do they really tell us any-thing about human behaviour in general or aboutthe individual? Does it make a difference in how weview people, select them for jobs, or guide therapychoices and assist in evaluating outcomes?

This is a very loaded question, and the onethat appears to challenge both the technicaladequacy of our personality measures, butespecially the construct and criterion validityor effectiveness of personality instruments indescribing individual differences, clinicaldiagnosis and guiding and evaluating inter-ventions. Technically, there are very fewactual ‘tests’ of personality – the Objective-Analytic Battery being an exception. Mostso-called ‘tests’ of personality are in fact,self-report scales or rating scales based onreports of others. Such scales quantify subjec-tive introspections, or subjective impressionsof others’ personality make-up. At the same

time, it is a relevant question and one that wewill continue to face in the study of personal-ity and the application of the findings,including assessment of personality, withinpsychological practice areas such as clinicaland school psychology, and within settingssuch as the military, business and sports psy-chology, among others. Volume 1 in this two-volume series is devoted to a critical analysis ofthe theories, models and resulting research thatdrive the personality descriptions and assess-ment discussed in Volume 2. Demonstratingboth the construct and practical validity ofpersonality descriptions is essential to psy-chology as a scientific discipline and empirically grounded practice/profession.

THE STATUS OF PSYCHOLOGICAL ASSESSMENT

In a recently published paper focusing onpsychological assessment, the followingclaim was made:

Data from more than 125 meta-analyses on testvalidity and 800 samples examining multimethod

1

9781412946520-Ch01 5/28/08 12:21 PM

assessment suggest four general conclusions: (a) Psychological test validity is strong and com-pelling, (b) psychological test validity is comparableto medical test validity, (c) distinct assessmentmethods provide unique sources of information,and (d) clinicians who rely exclusively on interviewsare prone to incomplete understandings. (Meyer et al., 2001: 128)

The authors also stated that multiple meth-ods of assessment in the hands of ‘skilled cli-nicians’ further enhanced the validity of theassessments so that the focus should nowmove on to how we use these scales in clini-cal practice to inform diagnosis and prescrip-tion. This is a remarkable accomplishment, ifaccurate, and even a bold claim that has notgone unchallenged. Claims (a) and (b) havebeen attacked on various grounds (e.g. seecritiques by Fernández-Ballesteros, 2002;Garb et al., 2002; Hunsley, 2002; Smith,2002). Furthermore, the debate about theclinical or treatment validity of psycholog-ical assessment and the added or incremen-tal value of multimethod assessment isargued by some not to rest on solid empiricalground (e.g. Hunsley, 2002; Hunsley andMeyer, 2003), in spite of such carefullyargued presentations on the utility of inte-grative assessment of personality with bothadults (e.g. Beutler and Groth-Marnat, 2003)and children (e.g. Riccio and Rodriguez,2007; Flanagan, 2007). In fact, this is verymuch the argument put forward by support-ers of RTI (response to intervention) in chal-lenge to the view that diagnostic assessment,using multiple assessment methods, shouldpoint the way to both diagnosis and inter-vention planning (see special issue ofPsychology in the Schools, 43(7), 2006).

While the Meyer et al. review focused onall areas of psychological assessment, itdoes suggest that the theories and models, aswell as research findings describing variouslatent traits underlying individual differ-ences have produced sufficient informationto allow for reliable and valid measurementand in turn, application of these assessmentfindings to understanding, predicting andeven changing human behaviour associatedwith intelligence, personality and conation

(see Boyle and Saklofske, 2004). Whilethere has been considerable progress, butcertainly not a consensus in the models andmeasures used to describe intelligence and cognitive abilities, the other main indi-vidual differences’ areas of personality and conation have travelled a somewhatdifferent path to their current position inpsychological assessment.

Calling this a remarkable accomplishmentalso has to be put in the context of time.Psychological science is only slightly morethan 125 years old. As a profession thatapplies the research findings from both exper-imental and correlational studies in diagnosis,intervention and prevention in healthcare set-tings, schools, business and so on, psychologyis even younger. Specializations that are heav-ily grounded in psychological assessmentsuch as clinical, school, counselling andindustrial-organizational psychology onlybegan to appear more or less in their presentform in the mid-twentieth century. While itcan be debated, the success of the Binet intel-ligence scales in both Europe and NorthAmerica in the early 1900s, followed by thewidespread use of ability and personalityinstruments for military selection duringWorld War I in the US, and the growing inter-est in psychoanalysis complimented bydevelopment and use of projective measuresto tap ‘hidden’ personality structures, pro-vided the strong foundation for the contem-porary measurement and assessment ofpersonality.

A BRIEF HISTORICAL NOTE ONPSYCHOLOGICAL ASSESSMENT

However, history shows that the descriptionand assessment of individual differences is notnew to psychology. Sattler (2001) and Aiken(2000) have provided brief outlines of keyevents in cognitive and educational assess-ment during the several hundred years prior tothe founding of psychology, and one canclearly sense that the ‘tasks’ of psychological

2 THE SAGE HANDBOOK OF PERSONALITY THEORY AND ASSESSMENT

9781412946520-Ch01 5/28/08 12:21 PM

measurement were being determined duringthis time. Prior to the creative scientific stud-ies by Galton in the nineteenth century, thefirst psychological laboratory established in1879 by Wundt, and psychology’s earliestefforts at measuring the ‘faculties of themind’ during the Brass Instruments era (e.g.James McKeen Cattell), there is a long his-tory documenting efforts to describe thebasis for human behaviour and what makesus alike all others and yet unique in otherways. As early as 4,000 years ago in China,there is evidence of very basic testing pro-grams for determining the ‘fit’ for variouscivil servants followed by the use of writtenexams some 2,000 years ago that continuedin various forms through to the start of thetwentieth century. Efforts to understand andassess human personality also have a longhistory that predates the study of psychology.Centuries before the psychoanalytic descrip-tions of Freud, who argued for the impor-tance of the unconscious and suggested thatthe putative tripartite personality structure of the id, ego and superego were shaped by a developmental process reflected in psychosexual stages, the Roman physicianGalen contended that human personality was a function of the body secretions(humors). Galen subsequently outlined thefirst personality typology characterized bythe choleric, melancholic, sanguine andphlegmatic types.

Interest in such processes as memory andreaction time, and efforts to assess and distin-guish between mental retardation and mentalillness were already underway before theestablishment of Galton’s psychometric lab-oratory in London and Wundt’s and Cattell’spsychophysical laboratories in Germany andthe US respectively. While much of this workwas focused on the study of intelligence andcognitive abilities, it laid the foundation forpsychological testing and assessment thathas shaped the face of psychology today.Probably the greatest impetus for test devel-opment came as a result of the success of theBinet intelligence tests, first in France andthen in the US. The use of tests to classify

school children according to ability was followed by the development and use of theArmy Alpha and Beta tests to aid in theselection of recruits (in terms of their cogni-tive abilities) for military service in the US Army. However at that same time, it wasalso recognised that there was a need to iden-tify military recruits who might be prone to,or manifest the symptoms of, psychologicaldisorders. Woodworth (1919) created thePersonal Data Sheet that presented exami-nees with a questionnaire not unlike thosefound on scales tapping psychiatric disordersto which a ‘yes–no’ response could be made.While there was not a control or check for‘faking good–faking bad’ protocols, themeasure was deemed to be a success. Thus,well before the Minnesota MultiphasicPersonality Inventory (MMPI), constructedby Hathaway and McKinley (1940, 1943)and its revised version (MMPI-2, 1989), as well as the California PsychologicalInventory (CPI; see Gough, 1987), and other more recent personality measures,Woodworth’s (1919) Personal Data Sheet,was followed shortly after by other personal-ity scales such as the Thurstone PersonalitySchedule (Thurstone and Thurstone, 1930)and the Bernreuter Personality Inventory(Bernreuter, 1931), which may be consid-ered the earliest personality measures, atleast employing a contemporary question-naire format. Of interest is that other meas-ures being constructed around the same timehighlighted the divergent views on personal-ity assessment methods at the time includingthe Rorschach Inkblot Test (Rorschach,1921) and the Human Figure Drawings(Goodenough, 1926) and the SentenceCompletion Tests (Payne, 1928).

PSYCHOLOGICAL SCIENCE VERSUS PSEUDOSCIENCE

The basis by which current psychologicalassessment methods and practices can beseparated from other attempts to describe the

PERSONALITY MEASUREMENT AND TESTING: AN OVERVIEW 3

9781412946520-Ch01 5/28/08 12:21 PM

latent traits and processes underlying differ-ences in human behaviour is the very factthat they are grounded in scientific research,as outlined in the editors’ introduction to Vol. 1. It is science that forces a method ofstudy, including objectivity, experimentationand empirical support of hypotheses, andrequires the creation and testing of theories.Psychology requires the operationalizing ofvariables and factors to be used in a descrip-tion of human behaviour. In contrast to pseu-dosciences that operate outside of thisframework and rest their case in beliefs, per-sonal viewpoints, and idiosyncratic opinions,psychology also demands replication and,where possible, quantification of measures.

Measurement is the cornerstone of psy-chology and has spawned a number of meth-ods for gathering the very data that maydemonstrate the usefulness or lack of useful-ness of a theory or provide the informationneeded to describe a particular human per-sonality characteristic or even diagnose apersonality disorder or clinical condition.Pseudosciences such as astrology, palmistryand phrenology, which compete with psy-chological views of personality, do notrequire such objective evidence to supporttheir claims; rather, vague ‘theories’ aretreated as fact and so-called evidence is oftentautological. Thus a strength of psychology isthat it has as its basis measurement thatincludes varying methods of gathering data totest theoretical ideas and hypotheses, as wellas strict adherence to psychometric measure-ment principles such as reliability, validityand standardization (cf. Boyle, 1985).

FOUNDATIONS OF PERSONALITYMEASUREMENT AND ASSESSMENT

As mentioned above, it was concurrent withthe advent of World War I that a major effortto assess personality characteristics was firstwitnessed. Prior to that time, the closestmeasure of personality would likely be con-sidered as the word association techniques

used by Jung. Today almost everyone isfamiliar with personality measures, self-report questionnaires and rating scales thatmost often appear in the form of a statementor question (e.g. ‘I am a very nervousperson’; ‘I enjoy activities where there are alot of people and excitement’) that the clientanswers with a ‘yes–no,’ ‘true–false’ or anextended scale such as a 5 or 7 point orgreater Likert-type scale with anchors suchas ‘always true of me–never true of me’ or‘definitely like me–not at all like me.’ Thesehighly structured measures contrast with themore ambiguous, subjective and open-endedtechniques most often found in projectiveinstruments such as the Rorschach Inkblot orThematic Apperception tests.

Indeed, there is a longstanding tensionbetween objective and subjective strategiesfor personality assessment (see Cattell andJohnson, 1986; Schuerger, 1986, Vol. 2). Useof questionnaires based on subjectiveinsights and self-reports has dominated thefield, but one may wonder how much thisdominance reflects the convenience and lowcost of questionnaire assessment. Advocatesof objective testing may legitimately ques-tion the validity of subjective experience andthe apparent ease with which desirableresponses may be faked. Table 1.1 sets outthe key issues dividing the two camps; bothhave strengths and weaknesses. We do nottake a position on which approach is ulti-mately to be preferred; the chapters in Vol. 2illustrate the vitality of both subjective andobjective measurement approaches. Ideally,multimethod measurement models in whichsubjective and objective indices converge oncommon latent traits are to be desired, butcurrent measurement technology remainssome way from achieving this goal.

Given the more common use of standard-ized personality measures such as, for exam-ple, the Minnesota Multiphasic PersonalityInventory-2 (MMPI-2), the CaliforniaPsychological Inventory (CPI), the SixteenPersonality Factor Questionnaire (16PF), the Eysenck Personality Questionnaire -Revised (EPQ-R), and the NEO Personality


9781412946520-Ch01 5/28/08 12:21 PM

Inventory – Revised (NEO-PI-R), a briefdescription of the strategies underlying theirconstruction will be presented here. Kaplanand Saccuzzo (2005) provide a usefuldescription of the various strategies employedin constructing personality measures.Deductive strategies employ both face validity(logical–content strategy) and theory-drivenviews of personality. However, the assump-tion that an item followed by a ‘yes’ response,on the basis of content alone (‘I am fre-quently on edge’) taps anxiety or the broaderneuroticism dimension found on scalesassessing the Big Five (NEO-PI-R) or thethree Eysenckian dimensions (EPQ-R) mayor may not be accurate. And for instrumentsthat employ a face-validity perspective, therational approach to constructing items tomeasure particular characteristics may provide the client motivated by other alterna-tive needs with the opportunity to provide

inaccurate and biased responses (e.g. seeBoyle et al., 1995; Boyle and Saklofske,2004). For example, a scale purportedly tapping aggression with items such as ‘I often start fights’ or ‘I have never backeddown from a chance to fight’ may be sotransparent as to increase the likelihood that examinees will also be more able tocreate a ‘false’ impression, depending ontheir motivation (e.g. early parole or lightercourt sentence, malingering).

The foundational basis for many contem-porary personality scales includes empiricalstrategies that employ the responses of vari-ous criterion groups (e.g. anxious vs. non-anxious adolescents) to determine how theydiffer on particular items and scales. Forexample, the very successful psychopathol-ogy scales, namely the MMPI (Hathawayand McKinley, 1940, 1943) and the revisedMMPI-2 published in 1989, and the Millon


Table 1.1 Objective vs. subjective assessments of personality – some key issuesObjective testing perspective Subjective testing perspective

Meaningfulness of People often lack insight into their Self-reports are a class of behaviours that mayself-reports true personalities. Personality usefully index latent personality traits.

may be shaped by unconscious forces As Cattell (1973) pointed out, self-reports (psychoanalysis) or by situationally may be treated as behaviours whose specific implicit learning processes. meaning can be established through

research (Q data) rather than as veridical insights (Q data).

Role of response bias Self-reports often reflect no more Response bias may be assessed independentlythan trivial response styles from latent traits. Furthermore, some(e.g. acquiescence), (e.g. or deliberate ‘biases’ may be integral to personality andimpression management faking). worth investigating as substantive traitsTechniques for assessment of (Paulhus, 2002).response bias may themselves be open to manipulation.

Biological basis of If personality is biologically based, it is Traits may be higher-level emergent personalpersonality as the unlikely that self-reports map directly qualities that are not isomorphic with anybasis for measurement onto the brain systems controlling traits. single brain system (Zuckerman, 2005). Thus,

Research should work towards direct it is difficult to capture traits in their entiretyassessment of individual differences in using biological indices. Specific biologicalneural functioning and their molecular- theories also have only mixed support fromgenetic sources. empirical tests (Matthews and Gilliland, 1999).

Status of objective, It is questionable whether subjective Historically, the reliability and validity of implicit and projective experience possesses the scaling leading projective tests has been controversial.tests properties necessary for quantitative The new generation of implicit measures do

measurement models (cf. Barrett, 2005). not yet have the extensive nomological netTests based on objective behaviours of traits assessed by questionnaire. Such traitsmay be intrinsically superior to currently possess superior criterion, constructsubjective reports in supporting and consequential validity.adequate measurement.

9781412946520-Ch01 5/28/08 12:21 PM

Clinical Multiaxial Inventories I, II and III(Millon, 1977, 1987, 2006) as well as the‘normal’ personality trait scale, the CaliforniaPersonality Inventory or CPI (Gough, 1987),are examples of instruments grounded in thisapproach to test construction and clinicaluse. Criterion-keyed inventories employ theapproach that is less tied to what an item‘says’ or any a priori views of what it mightbe assessing, but rather whether the item dis-criminates or differentiates a known extremegroup (e.g. clinical groups such as depressed,schizophrenic, etc.) from other clinical andnormal respondents.

In other instances, statistical techniques,particularly factor analysis, are also used toinfer or guide psychologists in determiningthe meaning of items and, thus, to define themajor personality trait dimensions. Cattell’sSixteen Personality Factor Questionnaire or16PF began as a large set of items based on alengthy trait list that were then reduced to 36‘surface traits’ and then further to 16 sourcetraits, said to describe the basic dimensionsof personality structure (see Boyle, 2006). In turn, structural equation modelling (see Cuttance and Ecob, 1987) allows per-sonality structure to be examined in thelarger context of other psychological vari-ables to portray a more comprehensive andintegrated description of human behaviour.

Finally, theory-driven measures draw fromdescriptions of ‘what should be’ or ‘folk con-cepts’ (e.g. CPI) and use this as the basis forconstructing personality instruments, an exam-ple being the Edwards Personal PreferenceSchedule based on Murray’s description ofhuman needs. The major personality theoriesthat have influenced the measurement of per-sonality include psychoanalysis (e.g.Rorschach Inkblot Test; Vaillant’s (1977)Interview Schedule for assessing defencemechanisms), phenomenology (Rogers andDymond’s Q-sort), behavioural and sociallearning (Rotter’s I-E Scale) and trait concep-tions (Cattell’s 16 PF; Eysenck’s EPQ-R; andCosta and McCrae’s NEO-PI-R).

Certainly, the personality scales andassessment techniques most often employed

today, in both research and clinical practice,include a combination of all the aboveapproaches. The Eysenckian measures (e.g. MPI, EPI, EPQ/EPQ-R), the Cattellianmeasures (e.g. 16PF, HSPQ, CPQ; CAQ), as well as Big Five measures such as theNEO-PI-R have relied on empirical andfactor analytic input into scale construction.Thus, the argument may be made that theNEO-PI-R, in spite of varying criticisms (seeBoyle, Vol. 1), is a popular instrument forassessing putative trait dimensions labelledextraversion, neuroticism, conscientiousness,agreeableness and openness to experience.However, as Boyle et al. (1995: 432)reported, the NEO-PI-R measures less thantwo-thirds of the known trait variance withinthe normal personality sphere alone. Indeed,the proponents of any one of the major per-sonality measures we have listed wouldclaim that the measure concerned is based ontheory, supported by research findings and isof practical value in clinical psychology andother applied fields.

TYPES OF PERSONALITYASSESSMENT

When one thinks of personality assessment,what usually comes to mind is the self-reportquestionnaire. This almost exclusive relianceon questionnaires asking the respondent toanswer a series of questions is showing signsof change and will continue to do so asgenetic, biological and neurological markers,for particular personality traits come to thefore over time. At this time, the emphasis ison multimethod assessment approaches toensure a convergence of results related topersonality (and other) assessment as well asdiagnosis of cognitive and affective disor-ders, includes case history and other extantdata, interview, observation, behavioural anda pot pourri of informal assessment strate-gies, in combination with standardized testsand questionnaires. However, it wouldappear from the Meyer et al. (2001) review,


9781412946520-Ch01 5/28/08 12:21 PM

that standardized, norm referenced measures(standard set of questions, method of admin-istration, scoring) are the most valid and reliable of the currently available methodsfor assessing personality constructs.

The use of questionnaires and self-reportinventories has dominated the field of per-sonality measurement. In contrast to per-formance measures used in the assessment ofcognitive ability (intelligence tests) and theassessment of skills through the use of, forexample, driving tests, musical competitionsand electrical apprenticeship practica, per-sonality assessment has largely employedsomewhat subjective self-report techniquesor reports of others using questionnaires,checklists and rating scales. While question-naire methods predated projective scales,their development was spurred by the needfor standardized scales that would minimizehuman error in administration, scoring andinterpretation. Use of such measures alsoallowed quantification of the personalitydimensions being examined. Accordingly,psychologists could not only determine thedirection (e.g. introversion vs. extraversion)but also the magnitude (e.g. very high scoreon extraversion at say, the 98th percentile) ofa particular trait. This in turn, allowed forfurther refinements in assessment as well asreplicability and cross-validation of theinstruments themselves. Standardized per-sonality instruments are most often associ-ated with the assessment of personality traits(see Matthews et al., 2003) including thosedescribed by H.J. Eysenck, R.B. Cattell, P.T. Costa and R.R. McCrae, D.N. Jacksonand others.

Projective measures are grounded in thetenets of dynamic psychology, beginningwith the early psychoanalytic work of Freud.These measures were developed as a way ofprobing into unconscious content and moti-vations and to give a ‘window’ into the basicpersonality of the client. Here it is the subjec-tivity that is celebrated both in terms of thestructure-free format that clients are given to respond to, often ambiguous stimuli, that will presumably allow for the expression of

personality but also the openness of interpre-tation afforded the clinician who is wellgrounded in the views and ‘clinical experi-ences’ of dynamic psychology. While thereare a number of projective measures rangingfrom the Szondi and Blacky Pictures to theRosenzweig Picture Frustration Test, sen-tence completion techniques and House-Tree-Person Test that were created during theearly and middle part of the last century, theThematic Apperception Test (Morgan andMurray, 1935) and Rorschach Inkblot Test(Rorschach, 1921) remain the most oftenused projective measures today. Even withsome waning in the interest of subjective/projective measures, in recent years, the well known Draw-A-Person and BenderGestalt tests, among others, have been furtherextended to include the assessment of psy-chopathology and affective indicators (e.g.Draw A Person: Screening Procedure forEmotional Disturbance; Naglieri et al., 1991).Langens and Schmalt (Vol. 1) discuss morerecent work that builds on the TAT.

As a reaction to the psychodynamic influ-ence in psychology and further drawing fromthe earlier success of Pavlov and Watson’swork in describing and changing behaviour,Skinner’s model of operant conditioning wasextensively embraced following World WarII and for the following 30 years. Here, thereis no interest in inferring latent traits under-lying the expressions of human behaviour, orsearching for unconscious mechanisms (theso-called ‘Black Box’) that might help explainindividual differences. Rather personality isviewed or operationalized as observablebehaviour reflecting the interaction betweenthe person and his/her environment. Thus abehaviour that has been identified as poten-tially relevant for intervention (e.g. hittingothers; talking out of turn) is observed interms of frequency, duration and so on, in thecontext of its antecedent and consequent con-ditions. Thus it can be determined if thebehaviour requires change, and if so, the pre-scriptive approach for doing so is to changethose antecedent (environmental factors suchas a noisy and distracting classroom) and/or


9781412946520-Ch01 5/28/08 12:21 PM

consequent (e.g. reinforcement) conditionsthat would maintain the behaviour in question.Furthermore, this method has considerablepredictive utility regarding the likelihood ofthe occurrence of particular behaviours.Based on systematic behavioural observation,there is no need to infer personality factors oran underlying personality structure. However,it is the use of observational data, most salientin the behavioural approaches, that is alsocentral to the clinical and research study ofpersonality.

Interviews have been a mainstay of psy-chological information and continue to formthe cornerstone of such specialized areas ascounselling psychology. Clinical psychol-ogy, industrial/organizational psychologyand many other branches of applied psychol-ogy employ both structured and more open-ended interviews to gather critical informationabout a client’s personal history, worries andconcerns, career aspirations, mental healthproblems and so on. While self report per-sonality questionnaires are essentially a formof structured interview, the use of interviewtechniques in general are considered to beless reliable and valid in diagnosis and treat-ment planning. However, in the service of a multimethod approach to personalityassessment, interview data can have bothexploratory and confirmatory usefulness. Toparaphrase Gordon Allport, if you want toknow what people think or feel, ask them!

In more recent years, explorations of thebiological and neurological bases of humanbehaviour, from fields such as behaviourgenetics and neuropsychology have con-tributed significantly to the study of person-ality. These contributions are extensivelydescribed in Vol. 1 in chapters by Stelmackand Rammsayer (psychophysiology) andalso by Johnson et al. (behaviour genetics).While many personality theories are firmlygrounded in brain-behaviour and geneticexplanations (e.g. Eysenck’s E and N fac-tors), tests of these hypothetical links arenow much more possible with the use ofMRI and fMRI, as well as metabolic, neuro-transmitter and genetic measures.

In line with the dominant tradition of thefield, many of the contributions to Vol. 2 areconcerned with questionnaire assessments.The various uni- and multidimensional personality questionnaires may be evaluatedagainst agreed standards for determining theefficacy of a given psychological measure-ment instrument (AERA/APA/NCME TestStandards, 1999). These standards lay out aframework for interpreting reliability andvalidity, so that the questionnaire developerhas the following obligations:

1 To provide evidence for the reliability of themeasure in question and information on thestandard error of measurement.

2 To demonstrate that a meaningful relationshipexists between the test’s content and the con-struct that it is intended to measure (similar to‘content validity’).

3 To provide theoretical and empirical analysessupporting (or disconfirming) relationshipsbetween the construct and the responses pro-vided by the test-taker (e.g. checking thatresponses are not driven by social desirability orother biasing factors).

4 To demonstrate that the internal structure of theconstruct is as suggested by the underlying theo-retical framework (e.g. whether it is uni- or mul-tidimensional, whether it is hierachical instructure, etc.).

5 To localize the construct within a nomologicalnet; that is, other individual differences variablesto which the assessment relates, as specified bytheory. This criterion relates to ‘construct validity’,including establishing both convergent and dis-criminant evidence, test-criterion relationships,and investigating how validity generalizes acrosssamples, situations and cultures.

Readers may determine for themselveshow well the leading questionnaires matchup to these test standards. We have indicatedpreviously in this introduction the need foralternatives to questionnaires, includingobjective tests. Volume 2 also addresses thesealternatives, in reviewing psychophysiologi-cal techniques that may lend themselves toassessment, and also implicit, objective andprojective tests. Historically, it has oftenproved difficult to obtain evidence for relia-bility and validity that matches corresponding


9781412946520-Ch01 5/28/08 12:21 PM

evidence for questionnaire assessments, butthe chapters here provide optimism that a newera of computer-interactive objective testingof personality constructs may be at hand.

INTRODUCTION TO VOLUME TWO

This volume contains a series of in-depth andcritical chapters on the broad topics of per-sonality measurement and assessment writ-ten by leading experts. The chapters aregrouped into several themes including gen-eral methodological issues, multidimensionalpersonality instruments, assessing biologi-cally based and self-regulative traits, fol-lowed then by projective and objectivepersonality measures, and lastly by measuresassessing abnormal personality.

GENERAL METHODOLOGICAL ISSUES

It is often said in relation to psychologicalassessment that the key to moving forwardwith psychometrically sound measurementrests with the definitions that are determinedto best represent a particular domain ofbehaviour, psychological disturbance (orwellbeing), or underlying traits such as extra-version or neuroticism. From the start, werealize that this is a daunting task for psy-chologists that will invariably require aninterdisciplinary perspective and effort. Johnet al. (1988) quite rightly asserted that ‘per-sonality psychology has not yet established agenerally accepted taxonomy of its subjectmatter which includes all variation in theovert social behaviour and the internal expe-riences of individuals’ (1988: 171). This isbased on the view that personality attributes,like so many other psychological constructs,including intelligence, are abstract conceptsthat are not directly observable, but rather areinferred. The search for a generally supportedtaxonomy would provide the needed basis forpersonality research, in spite of differences intheoretical orientation by bringing ‘an order’to the huge collection of personality variables

that have been created and studied over theyears. In turn, this has direct relevance towhat we measure in our personality instru-ments and how we can use this information inunderstanding individual differences.

Saucier (Vol. 2) contends that how wedefine, organize and measure personality canbe guided from lexical studies of natural lan-guage. In turn, these studies have formed thebasis for a personality structure that runs the gamut from a single factor solution some-what akin to ‘g’ in intelligence theory andmeasurement, to seven lexical factors. Whilelanguage may partly determine the numberof factors that emerge in an examination ofpersonality structure based on human lexi-cons, the issue becomes even more apparentwhen we attempt to develop measures toassess personality. The question can beasked: are personality characteristics univer-sal? If so, then other than their expression orthe actual behaviours observed to infer a per-sonality characteristic, the universality ofpersonality traits for example should allowfor the translation and adaptation of aninstrument from one language and culture toanother. But as we search further into the cultural and linguistic fabric of differingsocieties, we find unique examples of per-sonality factors that do not seem to have anequivalent elsewhere. Comparing and con-trasting cultures that are defined by an inde-pendent versus interdependent view of theself, the Japanese concept of ‘omoiyari’would seem to exhibit some relationship withprosocial behaviour and empathy as definedin Western psychology. However, it is alsounique because of the intuitive aspect(‘sasshi’) that is valued so highly in societiesthat are grounded in an interdependent viewof the person. The chapter by van de Vijverand van Hemert (Vol. 2) describes the criti-cal aspects of the methodology required incross-cultural research and instrument con-struction and then follows this with some ofthe advances in the cross-cultural measure-ment of personality. It readily becomesapparent that the search for both universalpersonality factors and potentially unique


9781412946520-Ch01 5/28/08 12:21 PM

clusters of personality variables will not be uncovered by simply comparing theresponses to scales administered in two dif-ferent countries or cultures, even if themeasures are translated.

In response to the diversity of views onpersonality that have resulted from varioustheoretical, research and measurement per-spectives in psychology and allied disci-plines, Jackson (Vol. 2) proposes a ‘hybridmodel’ that should serve both heuristic andpractical functions. Integrating biological,experiential and social-cognitive theories,Jackson describes how this model departsfrom earlier views that appears to have frag-mented rather than unified the study of per-sonality (e.g. viewing approach-avoidance asorthogonal constructs; separating tempera-ment and character). Of particular interest topractitioners is Jackson’s contention that theproposed hybrid model will guide the imple-mentation of various psychological treatmentinterventions. This has been a major concernof clinical, school/educational, counsellingand I/O psychologists, as well as those whopractice psychology in health, military, sport,forensic and other venues. As one psycholo-gist known to the authors quipped recently,‘What good does it do for the psychologistand client to know the client’s scores on theBig Five or to tell the client that s/he is astable introvert’! Predicting successful andunsuccessful outcomes with and withoutinterventions will provide personality psychology with the status accorded to intelligence and intelligence tests.

MULTIDIMENSIONAL PERSONALITYINSTRUMENTS

The second group of chapters is focused onan examination of some of best-known andmost often used measures of personality. In contrast to scales that are intended toassess psychopathology such as the MMPI-2,these measures reflect an eclectic underpin-ning of theory, trait descriptions, and factoranalysis that rather describe the structure of

personality. As the late Professor HansEysenck so often reminded us, the psychoti-cism or the P factor in his theory of personal-ity, and also assessed on the EPQ andEPQ-R, is not a measure of psychotic behav-iour or psychopathology. Rather it reflects atough minded/tender minded dimension thatmay predispose a person to psychopathy or schizophrenia.

The well known California PsychologicalInventory is now over 50-years-old and isconsidered to be very much akin to a ‘folkdescription’ of personality in contrast toinstruments either driven by theory orderived empirically from factor analysis.There is some disagreement about the actualfactor structure and whether this measure isbest described within the currently popularfive-factor model (FFM). Also, some of thescales on the CPI 260 and 434 are less reli-able than is minimally ideal. However, explo-ration of the current 20 CPI scales hasresulted in some new scales summarized byBoer et al. (Vol. 2). The CPI is one of themore often used measures in the businesssector by I/O psychologists for personnelselection but has also been used extensivelyin counselling and forensic settings.

Factor analysis has been a driving influ-ence on the development of both intelligenceand personality instruments. For example,the widely recognized three strata structureof intelligence described by Carroll wasbased on an analysis of 456 factor-analyticstudies of intelligence. On the other hand,proposed models of intelligence, for exampleby Spearman and Thurstone many years ear-lier, have been tested with factor analysis(both exploratory and confirmatory) to deter-mine if the proposed structure can be repli-cated with large data sets. Certainly, many ofthe trait descriptions of personality are in partderived from factor analysis (e.g. Cattell’s 16PF), or the theoretical structure is supportedwith the aid of factor analysis (e.g. Eysenck’sPEN model reflected in the EPQ/EPQ-R).

The Comrey Personality Scales (CPS) area very good example of how factor analysishas been employed over time to create the


9781412946520-Ch01 5/28/08 12:21 PM

eight factors found in this measure (Comrey,1994; Comrey and Lee, 1992). As is nowexpected with all scales that employ a ques-tionnaire format and self-ratings, a validityand response bias scale are included to assistin determining various biases that would thenchallenge the accuracy of the report and itsclinical usefulness. Comrey (Vol. 2) has pro-vided solid empirical evidence in support ofthe factor structure as well as the validity andclinical use of these scales. A question some-times asked about personality scales is:‘What does it mean to be extraverted … whatdoes this tell the psychologist or even theclient?’ Comrey has provided clear clinicaldescriptions of what it means to score high orlow on measures such as orderliness vs. lackof compulsion, or trust vs. defensiveness.A.L. Comrey’s psychometric work has stoodthe test of critical scrutiny over many yearsand remains an exemplary contribution to thescientific (factor analytic) construction ofmultidimensional personality scales, along-side the parallel outstanding contributions of R.B. Cattell and H.J. Eysenck (see Boyle, Vol. 1).

Probably one of the very best examples ofthe early use of factor analysis to define andmeasure personality characteristics as well asto expand and refine the scale is found in the16PF developed by R.B. Cattell that was firstpublished in 1949 (see Cattell et al., 1970),with the fifth revision being published in1993. Cattell (1973) described personality ascomprised of three levels. Starting with the16 primary personality traits, factor analysisproduced the second-order global factors thatvery much interface with the current Big Fivepersonality factors (Boyle, 1989b; Krug andJohns, 1986). Such a model does allow for apersonality description at several levels(strata) but also contributes to an understand-ing of individual differences as can be seen,for example, with global extraversion. Such ascale permits both research on large-scalepopulation comparisons (e.g. cross-culturalcomparisons) but also at the level of the indi-vidual who has requested counselling forinterpersonal problems or work-related

stress. The chapter by Heather Cattell andAlan Mead (Vol. 2) also portrays the significantrole played by the 16PF in defining variantsof the Big Five (see Boyle et al., 1995) butfurther delves into the debate surrounding thecorrelated vs. orthogonal relationship ofthese factors, drawing our attention to the‘power’ of factor analysis (oblique solutions)in defining the relational structure of person-ality. Another interesting issue is the relation-ship between the psychometric cornerstonesof reliability and validity and how they inter-act to an optimal level on measures such asthe 16PF. Of particular interest to the readersof this chapter are the comprehensive refer-ences to the clinical use and applications ofthe past and current versions of the 16PF.

Boyle and Barton (Vol. 2) have extendedthe chapter by Heather Cattell to first remindus that Raymond B. Cattell (as indexed byjournal citations) is one of the most influen-tial psychologists of the twentieth century(Haggbloom et al., 2002: 142). We are alsoreminded of the huge compendium of researchand measurement instruments in the person-ality field alone that Cattell gave us includingthe Sixteen Personality Questionnaire, theHigh School Personality Questionnaire, the Adolescent Personality Questionnaire, theChildren’s Personality Questionnaire, theEarly School Personality Questionnaire, the Preschool Personality Questionnaire, theCentral Trait-State Kit, the Objective-Analytic Battery and the Clinical AnalysisQuestionnaire along with its more recent ver-sion, the PsychEval Personality Questionnaire(Cattell et al., 2003). This chapter thenturns to an analysis of personality meas-ures using Barton’s nine-parameter modelthat targets key ‘who, what, how’ questionsand echoes Cattell’s call for the developmentof personality measures that go beyond theuse not only of subjective L-data and Q-datameasures, but also draw from the objectivetesting of behaviour (T-data). Such anapproach will provide psychologists with themultimethod assessment framework neededto converge on the most accurate and meaningful description of an individual’s


9781412946520-Ch01 5/28/08 12:21 PM

personality. On another note, as one readsthrough these two volumes and possiblybecomes concerned about the diversity of personality models and measurementapproaches, we are reminded of a statementby Eysenck (1984), also named as one of themost influential contributors to twentiethcentury psychology (see Haggbloom et al.,2002). In an analysis of Cattell’s personalitytheory and measures, Eysenck stated that ‘the Cattell and Eysenck constructs and theo-ries should be seen not as mutually contra-dictory, but as complementary and mutuallysupportive’ (1984: 336).

The Big Five personality factors have dom-inated the personality trait literature overrecent years. More references are seen to theputative constructs labelled: extraversion,neuroticism, openness to experience, agree-ableness and conscientiousness as measured,for example, by the NEO-PI-R than to anyother set of personality traits, in spite of thesubstantial lack of agreement among psychol-ogists (e.g. Cattell, 1995; Eysenck, 1992;McAdams, 1992; Schneider et al., 1995).However, Costa and McCrae have provided avery detailed ‘inside’ look at the constructionof the NEO-PI-R in relation to the position oftrait psychology, and criticisms of earlierpersonality measurement including othertrait measures based on Eysenck’s P, E, Nmodel and Cattell’s 16PF. Moving beyond N,E and O, and influenced by Norman andGoldberg’s factors defining the structure ofpersonality, Costa and McCrae engaged in anextensive research programme that resultedin the NEO-FFI, NEO-PI and the morerecent NEO-PI-R and NEO-PI 3 (a detailedcritique of the factor analytic methodologyemployed in construction of the NEO-PI-R isprovided by Boyle et al., 1995: 431–433).

While considerable research pertainingto scale reliability and validity has beenundertaken and some of the key findingsare included in this chapter, the question ofaccuracy in self-report measures has beenaddressed quite simply in the NEO-PI andNEO-PI-R questionnaires. A last item asksrespondents to say if they have answereditems honestly and accurately and if all items

have been responded to and the answer sheetcompleted correctly. However, this responsecan easily be manipulated. Even if a personthinks at a conscious level that s/he hasanswered ‘honestly’, it does not follownecessarily that the responses are accurate,particularly if the individual has poor self-insight. This simple approach is quite incontrast to many other scales such as the16PF, MMPI, PAI and BASC that haveincluded a number of ‘validity’ checks; how-ever the computer scored version of theNEO-PI-R does give further indications ofsuch potentially relevant indicators ofresponse accuracy and bias in reporting suchas the number of missing items. Although inrecent times the proposed Big Five personal-ity factors have tended to take ‘centre stage’in personality research, the use of the NEO-PI-R in applied settings is tempered by Costaand McCrae’s view that ‘more research isstill needed to optimize its application’.

As stated above, the debate over thenumber of traits that would ‘best’ define per-sonality, and in turn, that will have the great-est application to ‘real world’ settings,ranging from personnel selection to thera-peutic intervention prescriptions, has beenheard for many years. Based on both taxo-nomic descriptions and factor-analytic inves-tigations, it would appear that the three mostoften cited positions are those reflected in H. Eysenck’s (1991) paper, ‘Dimensions ofPersonality: 16, 5 or 3? Criteria for a taxo-nomic paradigm’. Eysenck’s personalitytheory, while having undergone variousrevisions (as outlined by O’Connor, Vol. 2),has stayed true to the position that the threemajor personality dimensions of extra-version, neuroticism and psychoticism aresufficient to account for individual differ-ences across a wide spectrum of humanbehaviours. However, the proportion of thepersonality sphere variance accounted for byEysenck’s three broad factors almost cer-tainly is less than optimal (cf. Mershon andGorsuch, 1988). Eysenck’s model has resultedin the publication of a number of scalesbeginning with the MPI tapping extraversionand neuroticism followed by the EPI and


9781412946520-Ch01 5/28/08 12:21 PM

even-tually the EPQ and EPQ-R thatincluded the P scale (see Furnham et al., Vol. 2). Children’s versions of the EPI andEPQ scales have been concurrently con-structed by S. Eysenck and are referred to asthe ‘junior’ versions (e.g. JEPQ; Eysenck andEysenck, 1975). While Eysenck’s modelclearly described a number of primary traitsfrom which the second-order factors of E, Nand P emerged, there was less effort investedin developing scales to measure each ofthese. However, over time the components ofextraversion, initially focusing on impulsivityand sociability, were split off with sociabilityremaining as part of E along with such otherprimary traits as sensation seeking and ven-turesomeness. While the Eysencks developedseveral scales to assess impulsivity, venture-someness and empathy, there was also someeffort to select those items from the EPQdefining the three-factor space to predictcriminal propensity and antisocial behaviour.The EPI and EPQ are still used extensively inresearch studies and have been translated andadapted for use in many different countries(Boyle and Saklofske, 2004).

Entering the taxonomic debate regarding thenumber of personality factors that are neededto account for individual differences in behav-iour, the Zuckerman–Kuhlman PersonalityQuestionnaire (ZKPQ) has redefined the factorspace described by the Big Five. AsZuckerman (Vol. 2) explains in his chapter, theZKPQ was developed to derive a personalitystructure appropriate for measuring basic per-sonality traits with their roots in biologicaltraits. Zuckerman provides a detailed discus-sion of the evolution of the ZKPQ into its cur-rent five-factor structure and the labelling ofthe factors as ImpSS (impulsive sensationseeking), N-Anx (neuroticism–Anxiety), Agg-Hos (aggression-hostility) and Sy (sociability).While the factor structure and psychometricintegrity of the ZKPQ and also the short-form(Zuckerman, 2005) has been replicated incross-cultural studies, some studies have alsoattested to the potential for use in a variety ofsettings ranging from risk taking in universitystudents (Zuckerman and Kuhlman, 2000) tothe reactions of migrant groups to moving into

a new and different culture (Schmitz, 2004).However, because of its underpinnings in thepsychobiology of personality, Zuckerman con-tends that possibly one of the greatest uses ofthe ZKPQ should be to explore the underlyingbasis of personality in the brain.

Ashton et al. (Vol. 2) have offered yetanother model that comprised of six factors,assessed using the HEXACO-PI (Ashton etal., 2006). They have argued that studies ofmore than a dozen languages show that sixpersonality factors appear common to all.The name of this model (HEXACO) servesas an acronym for the names of the factorsincluding: honesty–humility (H), emotional-ity (E), extraversion (X), agreeableness (A),conscientiousness (C) and openness to expe-rience (O). In their chapter, Ashton and Leewrite that, ‘Despite its lexical origins, theHEXACO model uses the name openness toexperience rather than intellect/imagina-tion/unconventionality.’ The major additionto this model, in comparison with the pro-posed Big Five factors, and resulting scale,again in comparison with the NEO-PI, is theH factor. However, Ashton and Lee further sug-gest that the six personality factors described intheir model reflect two dimensions represent-ing altruistic in contrast to antagonistic tenden-cies, and engagement within different areas ofendeavour. It is the contention of the authorsthat the honesty–humility and emotionality fac-tors are what give the HEXACO model anadvantage over the currently popular Big Five.This chapter provides Croatian, Greek, andFilipino data supporting the cross-culturalfactor structure.

Tellegen and Waller (Vol. 2) describe theprocess of constructing a measure of personal-ity from both deductive and externalapproaches to scale construction. However,they argue that an ‘exploratory’ approach usedduring the 10 year construction period of theMultidimensional Personality Questionnaire(MPQ) (Tellegen, 1982, 1995) has the advan-tage of permitting changes during the researchand development phases of test construction.The authors state that, ‘The intent in con-structing the MPQ was to clarify and demar-cate major dimensions in the self-descriptive


9781412946520-Ch01 5/28/08 12:21 PM

personality trait domain’ and offer strongsupport for their scale (e.g. heuristic virtues,substantial heritabilities, links with neurobe-havioural personality models, shared factorloadings with other major personality scalessuch as the EPQ, CPI, PRF, NEO, andCattell’s 16 PF, as well as respectable scalereliabilities). An interesting and controver-sial issue is raised in this chapter regardingthe congruence between self-ratings and rat-ings of others, as well as with external criteria.When comparing subjective self-report per-sonality ratings with ratings by knowledge-able others (e.g. spouse, friends/peers,employer/employees), it is not uncommon tosee some, and possibly, considerable diver-gence. This could be construed as measure-ment error related to the varying reliabilitiesof the scales or that the descriptors (e.g.items and scales) are aimed more at assessinglatent traits versus overt behavioural mani-festations that are less readily observed bysignificant others. Score discrepancies mightalso reflect biases in responding (socialdesirability, or malingering) by the ‘client’leading to a difference in self-ratings versusratings by others. Tellegen and Waller, infact, argue that the very nature of the MPQwill possibly result in ‘reliable discrepancieswhich are potentially informative and shouldnot be dismissed as simply demonstrating thefallibility of self-report’. One has to appreci-ate the view taken in this chapter about theuse of and feedback from personality meas-ures such as the MPQ in everyday appliedsettings: ‘With use of appropriate norms,feedback contributes to self-clarification bytranslating discrete self-statements intocoherent and telling characterizations.’

ASSESSMENT OF BIOLOGICALLYBASED TRAITS

The third section of this volume is titled‘Assessment of biologically based traits’ andincludes five chapters focusing on the biolog-ical underpinnings of personality structureand measurement. A major criticism of many

personality measures is that they are often sophenotypical in content and purpose that theymiss describing the causal and underlyingcorrelates of key personality traits and fac-tors. Saklofske and Eysenck (1994) statedthat ‘trait models of personality are some-times criticized for apparently pretending toexplain differences in behaviour by simplypostulating the existence of traits based onthat behaviour.… Factor analysis and othercorrelational methods are not meant to tellus anything about causality but to act astools for the discovery of a proper person-ality taxonomy. Having solved the problem,we may then go on to carry out the more dif-ficult task of finding out why some peopleare more sociable, others shy, why somepeople are extraverted, others introverted.’

Brain-behaviour studies began with theearlier EEG studies and are now driven bytechnologically sophisticated techniquessuch as fMRI used in neuropsychologicalresearch. The fields of psychobiology andpsychophysiology using early measures such as EMG, GSR and HR, have morerecently engaged in direct studies of neuro-transmitters such as dopamine and serotonin.The specialty areas such as behaviour genetics, initially studying human behaviourbetween individuals of varying degrees ofgenetic relatedness (e.g. kinship studies) nowhave access to DNA data. All are essential for a full understanding of human personalitythat links the phenotypic expression of personality with the underlying genotypicfoundations.

Furedy’s chapter (Vol. 2) provides an histor-ical backdrop for the relevance of psychophys-iological measurement in the study ofpersonality. He then describes how the follow-ing psychophysiological measures may be con-sidered by personality researchers: ‘peripheralvs. central measures, baseline vs. response-to-challenge measures, tonic vs. phasic measures;uniphasic vs. multiphasic measures; lo-tech vs.hi-tech measures; physiological “respectabil-ity” vs. psychological validity; temporal vs.localization measures; specific vs. reactive sen-sitivity; psycho-physiological vs. behavioural


9781412946520-Ch01 5/28/08 12:21 PM

measures; reliability vs. validity.’ But beforepsychologists who rely on self-report question-naires, and observation and interview databegin to feel that their measures are less ade-quate and not a ‘direct’ measure of personality,Furedy provides a very good example of thehigh reliability but low validity of the poly-graph for classification purposes (e.g. truthfulvs. deceptive individuals).

Extending this work, de Geus andNeumann (Vol. 2) provide two very goodreasons for the significance of psychophysio-logical measurement in furthering our under-standing of personality. They point to thelimitations of an over-reliance on self- andother-report ‘paper and pencil’ measures thatmay be prone to various biases, distortionsand psychometric shortcomings. In contrast, psychophysiological indices have the advan-tage that ‘voluntary control over the recordedbiological signals is limited if not absent’. At more of a construct validity level, theseauthors argue that personality may only becompletely understood by also describing thebiological processes underlying the majordimensions of personality. Focusing on thetwo most agreed-upon personality traits,extraversion and neuroticism, compellingevidence is provided to show their psychoph-siological underpinnings, ranging from ERPto fMRI data. In a similar tone to Furedy,however, de Geus and Neumann also agreethat the ‘reliability of psychophysiologicalmeasures is currently less convincing thanthose for paper-and-pencil measures andvalidity has been far more rigorously testedfor the latter’. They suggest two reasons forthe ‘shortfall’ of psychophysiological data to provide a solid foundation on which tobuild comprehensive understandings of per-sonality. First, many studies rely only on asingle measure (e.g. cerebral blood flow,EEG asymmetry) rather than examining thecomplex and often interactive nature of mul-tiple causal pathways. Second, ‘mainstreamneuroscience is still very much focused onuniversal affective and cognitive brainprocesses at the expense of individual differ-ences … by not taking individual differences

into account, or considering them a mere nuisance variable, many neuroscience studiesmay have failed to detect a link between a brain structure and the putative affectiveand cognitive processes in which it isinvolved.’

Congdon and Canli (Vol. 2) focus theiranalysis of the biological basis of personalityon the ‘primary’ personality factor of impulsivity. While impulsivity is considereda multidimensional construct found in manypersonality descriptions, included in many personality scales, and identified in variouspsychopathological (e.g. bipolar disorder,DSM-IV-TR Axis-II disorders) and neurobe-havioural (e.g. AD/HD) conditions, ‘the factthat patients are classified based on a taxon-omy that is not biologically based poses aserious challenge to efforts to investigate thebiological basis of impulsivity.’ Noting theshortcomings of fitting impulsivity into thelarger framework of personality, theydescribe research from non-invasive neu-roimaging and molecular genetic studies thathave separately provided support for a bio-logical foundation. However, their relianceof these studies on heterogeneous diagnosticcategories and self-report measures, ‘obscureany effect that a genotype may have on thephenotype of interest, especially when thesize of the effect is small’. Thus the authorsargue for an ‘endophenotype’ approach thatwould combine neuroimaging and moleculargenetic approaches and show its efficacy in an investigation of dopaminergic genevariation on impulsivity.

Strelau and Zawadski (Vol. 2) follow upon a lifetime of work on temperamentwhich Strelau defines as a stable set of personality traits, essentially present frombirth or early infancy. Although tempera-ment has a neurobiochemical basis, changesmay occur due to external conditions. Thischapter briefly outlines the more significanttheoretical views of temperament anddescribes some of the key measures thathave evolved within the psychometric tradi-tion. The very fact that there are some 30temperament instruments and more than 80


9781412946520-Ch01 5/28/08 12:21 PM

temperament scales reported in the literature,suggests that either the construct is so broadas to not be particularly useful in theory,research and practice, or that there is con-siderable overlap. Strelau and Zawadskistate that ‘the results of factor analysis con-firm the expectation of a broad five-factordomain of personality with temperamentalscales located mainly in two “arousability”factors: “emotionality/neuroticism” and“extraversion/activity” ’. However, Strelauand Zawadski argue that much work is stillrequired to add specificity to the very broadconcept of temperament and to develop reliable and valid measures.

ASSESSMENT OF SELF-REGULATIVETRAITS

‘Styles of self-regulation are integral aspectsof personality’ (Matthews et al., 2000: 199).They further argue that the integration of per-sonality traits and self-regulation requires aresolution of two divergent viewpoints: ‘Thetrait approach views personality as stableacross time and across different situations …much of the literature on self-regulationadopts a social-cognitive perspective thatconceptualizes personality as the outcome ofidiographic, contextually sensitive cognitiveprocesses’ (2000: 171). While most primary(e.g. impulsivity) and higher-order personal-ity traits (e.g. extraversion) relate to styles ofself-regulation, the chapters in section 3highlight this critical feature of humanbehaviour.

The chapter by Zayas et al. (Vol. 2) com-plements the two chapters in Vol. 1 thatexamine social-cognitive views of personal-ity (Cervone, Vol. 1; Matthews, Vol. 1). Zayasand colleagues further provide a detailedanalysis of the social-cognitive perspectiveon key questions that have confronted thestudy of personality. Earlier views of behav-iour in more simple terms were described asa function of persons interacting with theirenvironment (PxE). However, the growing

realization of the complexity of both thesevariables was most obvious when observingboth individual and intraindividual differ-ences. While there is a predictability tohuman behaviours that is surely grounded inpersonality, there is also the observation thata person’s behaviours will vary across situa-tions. It is not uncommon to hear expressionssuch as ‘he is a situational extravert’ suggest-ing that under particular situations, and thedemands arising from particular circum-stances, a person may behave or act some-what differently than under other conditions.Thus a purely trait perspective does notaccount adequately for such variabilityacross situations, but at the same time, humanbehaviour is not continuously random. Zayaset al. argue that such variability can best beunderstood by knowing what features in agiven situation are ‘psychologically active’ foreach of us. It is the psychologically importantor ‘if features’ of situations that activate bothcognitive and affective processes, which inturn result in thoughts, feeling and actions.This social cognitive perspective outlined byZayas et al. suggests that ‘if … then … pro-files provide clues for identifying individual-ity and personality coherence withinindividuals’ cross-situational variability …This variability need not be considered asource of error to be eliminated’.

As highlighted in the chapter by Fernandez(Vol. 2), anger, hostility and aggression have been studied in psychology since theearly formulations of Freud that elevatedaggression to one of the major human‘instincts’(cf. Fernandez, 2002). The earlywork of Rosenzweig using the Picture-Frustration Test, the questionnaire analysis ofaggression developed by Buss and Durkee(1957), the theoretical analysis of aggressionby Bandura (1983) and the more recent cog-nitive-behavioural descriptions presented byDodge (e.g. Crick and Dodge, 1996) are butsome examples of the interest in understand-ing and measuring aggression, anger andhostility. Fernandez distinguishes the qualita-tive aspects of anger from other emotionssuch as sadness and also the quantitative


9781412946520-Ch01 5/28/08 12:21 PM

aspects of low (annoyance) to high levels(rage) but further contends that ‘anger canassume the form of an emotion, a mood, or atemperament, depending on whether it isphasic, tonic, or cyclic’. While question-naires have served as the major methods ofassessing anger and hostility, the major con-tribution of this chapter by Fernandez is thedescription of six core dimensions in theexpression of anger including direction,locus, reaction, modality, impulsivity andobjective of anger. Using this model,Fernandez shows how anger profiles are cre-ated using the anchor points of these sixdimensions.

In contrast to the strict trait approach fordefining and measuring personality,Horowitz et al. (Vol. 2) have drawn from theinterpersonal model of personality and haveidentified four interpersonal measures(behaviours, traits, interpersonal goals, inter-personal problems) that are further organizedaround the two interpersonal dimensions ofcommunion and agency. This allows for thecreation of a profile (or ‘nomological net’)describing the individual using eight inter-personal variables. The measures derivedfrom this model and their application to per-sonality assessment are illustrated in thischapter but more importantly serve to ‘showhow the four interpersonal measures (theIMI, IAS, IIP, and CSIV) may be usedtogether to clarify other concepts in clinicalpsychology’ such as personality disorders.

Probably one of the major catalysts forstimulating an examination of the interfacebetween personality and intelligence calledfor by Saklofske and Zeidner (1995) is themore recent examination of emotional intel-ligence (EI). The two somewhat divergentviews of EI reflected in the trait formulationwith its closer links with personality traits,and the ability model proposed by Mayer and Salovey (1997) are described in thechapter by Austin et al. (see Vol. 1). Whatsets the so-called ability approach apart fromthe trait EI view is the focus on the interac-tion between emotion and cognition. AsRivers et al. (Vol. 2) outline in their chapter,

‘Emotional intelligence (EI) refers to thecapacity to both reason about emotion anduse emotion to enhance thinking and prob-lem solving.’ It is the skills of perceiving,using, understanding and managing emotionsthat are the foundation of EI. Furthermore,the method of measuring EI can be con-trasted. While trait scales (e.g. Bar-On, 1997;Schutte et al. 1998) mimic traditional self-report personality questionnaires where aperson’s position on the scale(s) is usuallydetermined using normative comparisons, the ability scales (MSCEIT and MSCEIT-YV)discussed in this chapter employ a problem-solving approach applied to emotional situa-tions, using both consensus and expertscoring (MSCEIT) and veridical scoring(MSCEIT-YV).

The MSCEIT is a departure from standardpersonality assessment using self-reportmeasures but rather, like intelligence tests,one that employs problem situations to whichthe respondent’s answers are compared withexpert opinion. Thus, the low correlationsbetween the two forms of EI assessment mayreflect differing conceptual underpinnings ofEI or method variance or both. And the farfrom high correlations with intelligence testsfor both trait and ability measures raises theinteresting question of whether EI should beconsidered ‘an intelligence’ (Austin andSaklofske, 2005). These are key issues raisedby Roberts et al. in their chapter (Vol. 2)which continues to look critically at EImodels and measures following the first andmore recent books by Matthews et al. (2002,2007). Focusing on the ‘intelligence’ aspectof EI, Roberts et al. argue that self-reportmeasures do not assess intelligence and thusshould not be construed as measures of EI, incontrast to ability-based models that are ‘theonly appropriate ones to delineate, and henceinvestigate, emotional intelligence’. Arguingfor a constrained view of EI, the authors thensuggest that the ability (or maximum performance) model reflected in theMSCEIT kind of measures holds the greatestpromise for assessing EI. However, they arecurrently limited by their ‘mono-operation


9781412946520-Ch01 5/28/08 12:21 PM

and mono-method biases’ and will benefit as well from ‘using new paradigms from emotions research, and new test constructiontechniques from I/O psychology’.

IMPLICIT, PROJECTIVE ANDOBJECTIVE MEASURES OF PERSONALITY

The next section of this volume turns to an examination of implicit, objective andprojective personality measurement. Probablyno other topic in the personality assessmentliterature has generated the same level ofdebate as that seen between proponents ofstandardized instruments versus projectivetests. On another level, cross-cultural issueshave also risen to the fore in relation to bothpersonality as well as intelligence measures.The ‘emic–etic’ perspectives on cross-cultural comparisons have raised a number of questions about how well both the con-structs used in one culture to operationalizeand assess, say intelligence or personality,can travel across national, cultural and linguistic borders. The reader may also referto Saucier’s chapter (Vol. 2) on the signifi-cance of the lexicon in determining how aculture describes and values various humancharacteristics.

Hong and Paunonen (Vol. 2) describe two‘non-verbal’ personality measures, the NPQand FF-NPQ. In contrast to the more psycho-dynamic measures such as the TAT, these twomeasures focus on explicit (rather than uncon-scious) personality characteristics, are sam-ples of behaviour reflecting personality traitsrather than ‘signs’ of some underlying person-ality disposition, and use a structured responseformat to ensure objective scoring andincrease scorer reliability. A key advantageargued for this format by Hong and Paunonenis that these measures are likely more portableand flexible when assessing individuals fromdifferent cultural and language backgroundsbecause the problem of translation, and alsoreading skills level, is reduced. In particular,the FF-NPQ should provide an alternative

measure for determining the robustness of theputative Big Five across cultures.

The basic difference in the assumptionsposed by projective versus standardized per-sonality instruments relates to whether person-ality traits and factors are explicitly known tothe person who is self-reporting or instead, thatpersonality is more implicit and may beassessed with techniques referred to as Implicit Association Tests (IATs). Schnabel et al. (Vol. 2) argue that IATs have a number ofadvantages over traditional questionnaire meth-ods for assessing personality. As described bySchnabel et al., IAT measures are designed to‘assess automatic associations between a con-trasted pair of target (such as “me” versus“others”) and attribute (such as “anxious”versus “confident”) concepts through a seriesof discrimination tasks that require fastresponding’. A basic premise of these meas-ures is that such motivational distortion fac-tors as faking good/bad, and so on, are lesslikely to confound or yield misleadingresults. However, while social cognitiveresearch has provided a foundation on whichto build IATs, explanations for IAT effects arestill less than fully understood, and the psy-chometric properties, especially reliability, ofstandard IAT measures is somewhat lowerthan considered desirable. At the same time,there is growing evidence from validity stud-ies that IATs may provide another ‘method’for assessing personality that would allowresearchers and clinicians to potentiallyaddress the issue of contaminating methodvariance that likely occurs when exclusivelyrelying on self-report questionnaires.However at this time, there is not sufficientevidence that IAPs should be used in clinicaldecision making related to diagnosis andtreatment planning and/or selection.

Schuerger (Vol. 2) provides yet a furtheralternative to assessing personality based onthe efforts of Cattell and Warburton (1967) tocreate actual performance tests (T-data) ofpersonality, a careful selection of which havebeen included in the Objective-Analytic TestBattery or OAB. While Schuerger concedesthat that original versions were very


9781412946520-Ch01 5/28/08 12:21 PM

cumbersome and not widely adopted, theidea underlying the OAB is quite contempo-rary and one that clearly supports a multi-trait, multimethod, multimodal approach toassessing human characteristics. A consensusis lacking regarding the factor structure ofthe OAB with Schuerger stating that only sixof the factors originally reported by Cattellhave been replicated in research conductedoutside Cattell’s laboratory. However,Schuerger also contends that the OA testsstill hold remarkable promise as demon-strated in both educational and clinical set-tings, and there may be even greateruntapped potential in individual OA vari-ables. In contrast to the time when Cattellwas developing the OA tests, the advent ofmodern computer technology and wide-spread computer use may yet be the formatfor reviving interest in such performancetests of personality structure. It would be sur-prising not to find more performance meas-ures being presented by computer in the verynear future. Our clinical laboratories alreadyhave this capability but microcomputers willalso make this is a reality for the ‘travellingclinician’ such as the school psychologist.

Standardized personality instruments,such as the Eysenck and Cattell measuresand projective techniques including theRorschach inkblots and TAT, have certainlydominated the field of personality assess-ment for much of the twentieth century.While personality scales are still a mainstayin contemporary psychology, including thosedescribed in the chapters of Vol. 2, bothresearch and practicing psychologists arealso interested in assessing the manifesta-tions and related behaviours of underlyingpersonality dimensions such as anxiety,depression and aggression. Thus, while thetendency towards aggressive behaviour canbe plotted on a three-dimensional matrixdefined by E, N, and P, of greater clinicalutility to psychologists is to have more‘direct’ measures of the level and type ofaggressive behaviour. For example, the earlyBuss–Durkee scale was more focused onassessing the direct expression of aggression

in its own right just as were the depressionand anxiety scales developed by Beck (e.g. Beck and Steer, 1993). Not all scales arefocused on the ‘negative’ or pathological side,and in particular, we now see scales tappinghappiness, life satisfaction and subjectivewellbeing.

The past several decades have witnessedthe development of multiscale measures that tap a wide range of psychologicallyimportant behaviours. Scales developed byAchenbach and Connors for multidimen-sional measures paved the way for many of the new scales that tap a number of behavioural factors of relevance to clinicaldiagnosis and intervention planning. The first and now recently revised BehaviourAssessment System for Children developedby Reynolds and Kamphaus (1992, 2004) isnot a personality measure in the strict senseof tapping those traits thought to underliebehaviour, but rather a more direct assess-ment of behaviour itself. The advantages theyoffer to the practicing psychologist aredescribed by Rowe et al. (Vol. 2) who alsoremind us that this, or any other scale, shouldnever ‘stand alone’ as the sole basis for diag-nosis or prescription. The current BASC-IIprovides statements that the respondent(child, parent, teacher) answers using a four-point (or true–false) Likert-type formatyielding composite, primary and contentscores. For example, the mixture of items onthe primary self-report scales range fromanxiety to attention problems, self-esteem tosense of inadequacy, and locus of control tosensation seeking, thus reflecting a veryeclectic mix of scales all focusing on behav-iour and the behaviours argued to describe,say, locus of control. Of interest is that theBASC-II has become the most often usedbehavioural measure by school and child-adolescent clinical psychologists, in partbecause of its solid psychometric propertiesand time–cost benefits, but more so becauseit provides a ‘direct’ method of assessment ofboth ‘strengths and problem areas’.

There is clearly consensus among psy-chologists for a multimethod approach to


9781412946520-Ch01 5/28/08 12:21 PM

personality assessment based on empiricallysupported models and methods. Blais andBaity (Vol. 2) have critically assessed theposition of the two most well known projec-tive measures, the Rorshach Ink Blot Testand the TAT in the context of clinical diagno-sis by examining the DSM-IV Axis-II per-sonality disorders and diagnostic efficacy ofcurrent scoring methods for both measures.While there is certainly controversy and dis-agreement about the use of projective instru-ments in assessing not just personality, butalso psychopathology (e.g. Gacono et al.,2002), psychologists must remember thatthese are empirical questions that remain tobe decided by the evidence. More to the point,Blais and Baity also remind us that there is nota direct correspondence between the variousRorschach or TAT scoring systems and anactual DSM-IV diagnosis. Rather the contri-bution of these ‘performance’ measures canbest be realized when they are integrated intosystems and perspectives describing personal-ity and psychopathology. In a recent paper thatthe reader may also wish to consult, Hughes et al. (2007) focused attention on the use ofthe Rorschach by school psychologists andafter an extensive review of the Rorschach andExner’s Comprehensive System for adminis-tration and interpretation, concluded that they‘meet current ethical and legal standards fortests’ (2007: 288).

ABNORMAL PERSONALITY TRAITINSTRUMENTS

The last section of Vol. 2 examines severalvery specific measures for assessing abnor-mal personality traits either through anexamination of those personality characteris-tics known or believed to underlie psy-chopathological behaviour, or by a more‘behavioural’ examination of particular clini-cal conditions and syndromes that have beendescribed in DSM and ICD classifications.Just as the Wechsler Intelligence Scales weredeemed the most often used measures for

assessing cognitive ability in the twentiethcentury, so too, the MMPI and more recentlythe MMPI-2 have been among the most pop-ular and often used self-report measures forassessing psychopathology. In fact, theMMPI was so widely used that it found itsway from primarily psychiatric and forensicsettings, to personnel selection and universitycounselling settings.

Again, much like intelligence tests of theearlier part of the twentieth century thatattempted to assess the full complement ofcognitive abilities, the MMPI was intendedas a comprehensive measure of the gamut ofpsychopathological conditions. Thus, insome ways, as the chapter by Helmes (Vol. 2) points out, the MMPI does stand outas compared with the shorter and morespecifically focused measures of abnormalpersonality of more recent years. As both ascreening instrument and for distinguishingbroad types of psychopathology (e.g. depres-sion vs. psychopathy), both versions of theMMPI have served us well. However, theMMPI is more limited for differential diagno-sis (e.g. anxiety vs. depression), but then thatis an unrealistic expectation for any measure,even one as lengthy as the MMPI, since thediagnosis of psychological disorders requiresthe convergence of clinical data from a multi-method approach. A review of the literature(Helmes and Reddon, 1993) does not providea great deal of evidence to support the use ofthe clinical scales for differential diagnosis.With any instrument that has survived as longas the MMPI, there is the tendency for somemyth or beliefs to ‘trump’ what the evidenceactually tells us about the MMPI’s clinicalefficacy and empirically validated best prac-tices use. However, Helmes states that ‘thereis promise that the new RC scales will bebetter able to make such distinctions than thetraditional clinical scales, but the relevantstudies have yet to appear in print’.

The debate surrounding the MMPI/MMPI-2 scales raises an interesting issueabout the current role of such measures inclinical assessment and the issue of effi-ciency. The tendency now seems to be


9781412946520-Ch01 5/28/08 12:21 PM

towards use of shorter, more time-efficientand more focused measures, although themore recently published BASC/BASC-II, forexample, are not so brief. The advantage ofthe MMPI/MMPI-2 is that it does serve as abroad screening measure for evaluating vari-ous broad types of psychopathology (or thelack thereof). In contrast, the use of say abrief depression inventory, would only bemore useful if the psychologist was eitherattempting to rule out depressive sympto-matology or had formed the hypothesis ofdepression, based on other indicators (e.g.interview and presenting symptoms, familyand previous clinical history) and was addingconfirmatory evidence.

What is in store for this ‘battleship’ meas-ure? Helmes summarizes:

The MMPI-2 does not represent a highly sophisti-cated approach to assessment that is based uponthe state-of-the-art in diagnosis and conceptual-izations of psychopathology. Successive introduc-tions of new scales have modernized aspects ofthe interpretation of the test, at the cost of provid-ing increased opportunities for conflicting scoresthat need to be reconciled during the overly com-plex interpretive process. The escalating collectionof scales for the MMPI-2, with each successive setproviding at best modest increases in incrementalvalidity for some applications, simply multiplythe number of potential sources of interpretiveconflict. … the future of the MMPI/MMPI-2 thusremains difficult to predict.

While new measures, including personal-ity, behaviour and psychopathology meas-ures continue to abound in psychology andcertainly present a challenge to formerly well-established measures, Krug (Vol. 2) reviewsthe Cattellian Clinical Analysis Questionnaire(CAQ) and its revised version, the PEPQ, ascomprehensive measures of both normal per-sonality and psychopathology. In spite ofsome support for the psychometric strengthsof the separate and composite scales, they donot appear to be often used in either researchor in clinical settings, although in militarycontexts the CAQ has received considerableuse (e.g. the Australian Army PsychologyCorps has used the CAQ extensively in itspsychological research, assessment andselection procedures – see Boyle, 1989).

This chapter raises the interesting issue ofwhether we have been too quick to abandonthe theoretical, research and measurementcontributions of such key figures in psychol-ogy as Cattell and Eysenck. As noted above,Eysenck’s E and N scales included in theEPQ and EPQ-R instruments are psychome-trically sound and central to a trait descrip-tion of personality. Similarly, a thoroughstudy of the Cattellian instruments is neededbefore we too quickly engage in an ‘out withthe old and in with the new’ attitude and laterdiscover that we may have simply ‘rein-vented the wheel’ (e.g. see the number of dif-ferent scales that assess risk taking, sensationseeking, thrill seeking, etc.)

In contrast to the MMPI and MMPI-2 thatare not grounded in a contemporary model of either personality or psychopathology, the measures described in the remainingchapters provide reassuring alternatives. The Dimensional Assessment of PersonalityPathology (DAPP) measures employ a con-struct validation approach to arrive at a clas-sification of personality disorder. Empiricalevaluations leading to revisions in the initialtheoretical description of personality disor-der are supported by increases in the validityof the classification scheme. The DAPP hasalso evolved from a somewhat differentapproach than many of the current personal-ity and clinical scales that start with adescription of the personality trait (e.g. extra-version) or disorder (e.g. anxiety). Ratherthan beginning with this a priori view,Livesley and Larstone (Vol. 2) state that theDAPP ‘incorporates a bottom-up approach inwhich diagnostic constructs evolve based onempirical evidence of the way the features ofpersonality disorder are organized’. Theadvantage of this approach is that it providesfor a dynamic rather than static view of per-sonality criteria and categories that is forcedto modify or change with new evidence fromboth research studies and clinical use (e.g. Cattell et al., 2002; Roberts et al., 2006a,2006b). Even more compelling is the recon-ciliation between normal and abnormal per-sonality that were treated quite separately


9781412946520-Ch01 5/28/08 12:21 PM

even into the latter part of the twentieth cen-tury. This may well be why measures such asthe EPQ and NEO-FFI and NEO-PI appearedin stark contrast to the many separate pathol-ogy scales (e.g. MMPI), with the latter notgrounded in underlying personality factors,but rather collections of psychiatric symp-toms based on archaic psychiatric nosology(Cronbach, 1990, p. 539). The DAPP and itscounterparts have provided the foundationfor the much-needed reconciliation betweenbasic descriptions of personality models andtraits on the one hand, and personality disor-ders and psychopathology on the other.

As noted above, while there has been a ten-dency towards constructing more specificallyfocused and brief measures of both normaland abnormal personality traits, in contrast,the Personality Assessment Inventory (PAI)devised by Morey (1991, 2007), has beenconstructed on the basis of contemporarydiagnostic classifications, and as an attemptto overcome the well-documented limitationsof the MMPI/MMPI-2 instruments (Helmesand Reddon, 1993). The PAI comprises nofewer than 344 items with a mix of validity,clinical, treatment consideration and interper-sonal scales. Morey and Ambwani (Vol. 2)summarize findings showing that this multi-dimensional inventory has received increas-ing attention in both research and clinicalpractice settings. Studies supporting thevalidity of the PAI subscales in the assess-ment of a wide range of psychological prob-lem areas, ranging from eating disorders toemotional injury, will ensure its continueduse in both applied and research settings.

This section ends with a summary of theMillon inventories and view of personalityassessment. All would agree that no singletheory or measurement instrument, no matterhow robust or broad, can ever give a com-plete description of an individual’s personal-ity structure. As Millon (Vol. 2) argues, ourefforts to measure human personality with apredefined set of traits that are reflected inour assessment tools is complicated by thevery nature of examining a breakdown ofthese traits for each individual, and thenreconstructing a description of personality;

the ‘loop’ from idiographic individuality tonomothetic commonality to nomothetic indi-viduality is brought to closure. Millon’schapter provides a detailed overview of theinventories that he has developed over thepast several decades, highlighting the linkswith both the DSM and ICD taxonomies ofpersonality disorders, as well as the theoreti-cal basis for conceptualizing both personalityand abnormal behaviour. The critical questionthat has so often been posed regarding thedirect association between assessment, diag-nosis and treatment planning or therapy isaddressed by Millon who argues that hisinventories provide a necessary basis forassociating polarity schemes and clinicaldomains with corresponding therapies: ‘Any discussion of personalized psychother-apy … must take place at a level of abstrac-tion or integration commensurate with that ofpersonality itself. Personality disorders andclinical syndromes cannot be remedied if theperson is thoroughly integrated while the therapy is not. Therapy must be as individualized as the person’.

SUMMARY COMMENTS

The chapters included in this volume are tes-timony to the incredible progress that hasbeen made in the measurement and assess-ment of personality, particularly in morerecent years. Guided by various theoreticalmodels and research findings as well asextensive interdisciplinary collaboration, thesophistication of psychological measurementwill continue to provide the necessary assess-ment tools to further our basic and appliedanalysis of human personality. Each chapterin this volume is a celebration of the researchcontributions and clinical knowledge of lead-ing experts in personality measurement andassessment. We thank all of the authors forsharing with us their critical analyses of themodels and methods for measuring personal-ity and especially their insights and creativitythat will serve well the measurement andassessment of human personality.


9781412946520-Ch01 5/28/08 12:21 PM

REFERENCES

Aiken, L.R. (2000) Rating Scales and Checklists:Evaluating Behavior, Personality, andAttitude. New York: Wiley.

American Educational Research Association,American Psychological Association,National Council on Measurement inEducation (1999) Standards for Educationaland Psychological Testing. Washington, DC:American Educational Research Association.

Ashton, M.C., Lee, K., de Vries, R.E., Perugini,M., Gnisci, A. and Sergi, I. (2006) ‘TheHEXACO model of personality structure andindigenous lexical personality dimensions inItalian, Dutch, and English’, Journal ofResearch in Personality, 40(6): 851–75.

Austin, E.J. and Saklofske, D.H. (2005) ‘Far toomany intelligences? On the communalitiesand differences between social, practical, andemotional intelligence’, in R. Schulze andR.D. Roberts (eds), Emotional Intelligence: AnInternational Handbook. Cambridge, MA:Hogrefe and Huber, pp. 107–128.

Bandura, A. (1983) ‘Psychological mechanismsof aggression’, in R.G. Geen and E.I.Donnerstein (eds), Aggression: Theoreticaland Empirical Reviews (Vol. 1). New York:Academic, pp. 1–40.

Bar-On, R. (1997) BarOn Emotional QuotientInventory (EQ-i): Technical Manual. Toronto:Multi-Health Systems.

Barrett, P. (2005) ‘What if there were no psy-chometrics?: Constructs, complexity, andmeasurement’, Journal of PersonalityAssessment, 85(2): 134–40.

Beck, A.T. and Steer, R.A. (1993) ‘Manual forthe Beck Depression Inventory ’, San Antonio,TX: The Psychological Corporation.

Bernreuter R.G. (1931) The Personality Inventory.Stanford, CA: Stanford University Press.

Beutler, L.E. and Groth-Marnat, G. (2003) (eds),Integrative Assessment of Adult Personality(2nd edn). New York: Guilford.

Block, J. (1995) ‘A contrarian view of the five-factor approach to personality description’,Psychological Bulletin, 117(2): 187–229.

Boyle, G.J. (1985) ‘Self-report measures ofdepression: Some psychometric considera-tions’, British Journal of Clinical Psychology,24(1): 45–59.

Boyle, G.J. (1989a) ‘A Guide to Use of theSixteen Personality Factor Questionnaire and

Clinical Analysis Questionnaire in MilitarySelection’, Canberra: Australian ArmyDirectorate of Psychology, 1st PsychologicalResearch Unit.

Boyle, G.J. (1989b) ‘Re-examination of themajor personality-type factors in the Cattell,Comrey and Eysenck scales: Were the factorsolutions by Noller et al. optimal?’, Personalityand Individual Differences, 10(12): 1289–99.

Boyle, G.J. (2006) ‘Scientific analysis of person-ality and individual differences’, DSc thesis,University of Queensland.

Boyle, G.J. and Saklofske, D.H. (2004) (eds),Sage Benchmarks in Psychology: ThePsychology of Individual Differences (Vols.1–4). London: Sage.

Boyle, G.J., Stankov, L. and Cattell, R.B. (1995)‘Measurement and statistical models in thestudy of personality and intelligence’, in D.H.Saklofske and M. Zeidner (eds), InternationalHandbook of Personality and Intelligence.New York: Plenum.

Buss, A. and Durkee, A. (1957) ‘An inventoryfor assessing different kinds of hostility’,Journal of Consulting Psychology, 21(4):343–9.

Cattell, R.B. (1973) Personality and Mood byQuestionnaire. New York: Jossey Bass.

Cattell, R.B. (1995) ‘The fallacy of five factors inthe personality sphere’, The Psychologist,8(5): 207–8.

Cattell, R.B., Boyle, G.J. and Chant, D. (2002)‘The enriched behavioral prediction equationand its impact on structured learning and thedynamic calculus’, Psychological Review,109(1), 202–5.

Cattell, R.B., Cattell, A.K., Cattell, H.E., Russell,M.T. and Bedwell, S. (2003) ‘The PsychEvalPersonality Questionnaire’, Champaign, IL:Institute for Personality & Ability Testing, Inc.

Cattell, R.B., Eber, H.W. and Tatsuoka, M.M.(1970) Handbook for the Sixteen Perso-nality Factor Questionnaire. Champaign, IL: Institute for Personality and AbilityTesting, Inc.

Cattell, R.B. and Johnson, R.C. (1986) (eds),Functional Psychological Testing: Principlesand Instruments. New York: Brunner/Mazel.

Cattell, R.B. and Warburton, F.W. (1967)Objective Personality and Motivation Tests: ATheoretical Introduction and PracticalCompendium. Champaign, IL: University ofIllinois Press.


9781412946520-Ch01 5/28/08 12:21 PM

Comrey, A.L. (1994) Revised Manual and Handbook of Interpretations for the Comrey Personality Scales. San Diego. CA:Educational and Industrial Testing Service.

Comrey, A.L. and Lee, H.B. (1992) A FirstCourse in Factor Analysis (2nd edn).Hillsdale, NJ: Erlbaum.

Costa, P.T. and McCrae, R.R. (1992) The NEOPersonality Inventory (Revised) Manual. Odessa,FL: Psychological Assessment Resources.

Crick, N.R. and Dodge, K.A. (1996) ‘Socialinformation-processing mechanisms in reactive and proactive aggression’, ChildDevelopment, 67(3): 993–1002.

Cronbach, L.J. (1990) Essentials of Psycho-logical Testing (5th edn). New York: Harper & Row.

Cuttance, P. and Ecob, R. (1987) (eds),Structural Modeling by Example: Appli-cations in Educational, Sociological, andBehavioral Research. Cambridge, UK:Cambridge University Press, pp. 241–79.

Eysenck, H.J. (1984) ‘Cattell and the theory of personality’, Multivariate BehavioralResearch, 19(2–3): 323–36.

Eysenck, H.J. (1991) ‘Dimensions of perso-nality: 16, 5, or 3? Criteria for a taxonomic paradigm’, Personality and IndividualDifferences, 12(8): 773–90.

Eysenck H.J. (1992) ‘Four ways five factors are not basic’, Personality and IndividualDifferences, 13(6): 667–73.

Eysenck, H.J. and Eysenck, S.B.G. (1975)Manual of the Eysenck PersonalityQuestionnaire (Adult and Junior). London,UK: Hodder & Stoughton.

Fernandez, E. (2002) ‘Anxiety, depression, andanger in pain: Research findings and clinicaloptions’, Dallas, TX: Advanced PsychologicalResources.

Fernández-Ballesteros, R. (2002) ‘How shouldpsychological assessment be considered’,American Psychologist, 57(2): 138–9.

Flanagan, R. (2007) ‘Comments on the minis-eries: Personality assessment in school psycho-logy’, Psychology in the Schools, 44(3):311–18.

Gacono, C.B., Evans, F.B. and Viglione, D.J.(2002) ‘The Rorschach in forensic practice’,Journal of Forensic Psychology Practice, 2(3):33–53.

Garb, H.N., Klein, D.F. and Grove, W.M. (2002)‘Comparison of medical and psychologicaltests’, American Psychologist, 57(2): 137–8.

Goodenough, F.L. (1929) Measurement ofIntelligence by Drawings. New York:Harcourt, Brace & World.

Gough, H.G. (1987) California PsychologicalInventory Manual. Palo Alto, CA: ConsultingPsychological Press.

Haggbloom, S.J., Warnick, R., Warnick, J.E.,Jones, V.K., Yarbrough, G.L., Russell, T.M.,Borecky, C.M., McGahhey, R., Powell III, J.L.,Beavers, J. and Monte, E. (2002) ‘The 100most eminent psychologists of the 20th cen-tury’, Review of General Psychology, 6(2):139–52.

Hathaway, S.R. and McKinley, J.C. (1940) ‘A multiphasic personality schedule(Minnesota): 1. Construction of the sched-ule’, Journal of Psychology, 10: 249–54.

Hathaway, S.R. and McKinley, J.C. (1943) ‘TheMinnesota Multiphasic Personality Inventory(rev edn)’, Minneapolis: University of MinnesotaPress.

Helmes, E. and Reddon, J.R. (1993) ‘A perspec-tive on developments in assessing psy-chopathology: A critical review of the MMPIand MMPI-2’, Psychological Bulletin, 113(3):453–71.

Hunsley, J. (2002) ‘Psychological testing andpsychological assessment: A closer examina-tion’, American Psychologist, 57(2): 139–40.

Hunsley, J. and Meyer, G.J. (2003) ‘The incremental validity of psychological testingand assessment: Conceptual, methodo-logical, and statistical issues’, PsychologicalAssessment, 15(4): 446–55.

Hughes, T.L., Gacono, C.B. and Owen, P.F.(2007) ‘Current status of Rorschach assess-ment implications for the school psycholo-gist’, Psychology in the Schools, 44(3):281–91.

John, O.P., Angleitner, A. and Ostendorf, F.(1988) ‘The lexical approach to personality: A historical review of trait taxonomicresearch’, European Journal of Personality,2(3): 171–203.

Kaplan, R.M. and Saccuzzo, D.P. (2005)Psychological Testing: Principles, Applications,and Issues (6th edn). Belmont, CA: ThomsonWadsworth.

Krug, S.E. and Johns, E.F. (1986) ‘A large scalecross-validation of second-order personalitystructure defined by the 16PF’, PsychologicalReports, 5(2): 683–93.

Matthews, G. and Gilliland, K. (1999) ‘The personality theories of H.J. Eysenck and


9781412946520-Ch01 5/28/08 12:21 PM

J.A. Gray: A comparative review’, Personalityand Individual Differences, 26(4): 583–626.

Matthews, G., Deary, I.J. and Whiteman, M.C.(2003) Personality Traits (2nd edn).Cambridge. UK: Cambridge University Press.

Matthews, G., Schwean, V.L., Campbell, S.E.,Saklofske, D.H. and Mohamed, A.A.R.(2000) ‘Personality, self-regulation, andadaptation: A cognitive-social framework’, inM. Borksrtyd, P.R. Pintrich and M. Zeidner(eds), Handbook of Self-regulation. SanDiego: Academic, pp. 171–207.

Matthews, G., Zeidner, M. and Roberts, R.D.(2002) Emotional Intelligence: Science andMyth. Cambridge, MA: MIT Press.

Mayer, J.D. and Salovey, P. (1997) ‘What is emotional intelligence?’, in P. Salovey and D.J. Sluyter (eds), Emotional Develop-ment and Emotional Intelligence: Educa-tional Implications. New York: Basic Books,pp. 3–31.

McAdams, D.P. (1992) ‘The five-factor model inpersonality: A critical appraisal’, Journal ofPersonality, 60(2): 329–61.

Mershon, B. and Gorsuch, R.L. (1988) ‘Numberof factors in the personality sphere: Doesincrease in factors increase predictability ofreal-life criteria?’, Journal of Personality andSocial Psychology, 55(4): 675–80.

Meyer, G.J., Finn, S.E., Eyde, L.D., Kay, G.G.,Moreland, K.L., Dies, R.R., Eisman, E.J.,Kubiszyn, T.W. and Reed, G.M. (2001)‘Psychological testing and psychologicalassessment: A review of evidence and issues’,American Psychologist, 56(2): 128–65.

Millon, T. (1977) Manual for the Millon ClinicalMultiaxial Inventory (MCMI). Minneapolis:National Computer Systems.

Millon, T. (1987) Manual for the Millon ClinicalMultiaxial Inventory-II (MCMI-II). Minneapolis:National Computer Systems.

Millon, T. (2006) Millon Clinical MultiaxialInventory-III Manual (3rd edn). Minneapolis:NCS Pearson.

Morey, L.C. (1991) The Personality AssessmentInventory Professional Manual. Odessa, FL:Psychological Assessment Resources.

Morey, L.C. (2007) The Personality AssessmentInventory Professional Manual (2nd edn).Odessa, FL: Psychological AssessmentResources.

Morgan, C.D. and Murray, H.A. (1935) ‘Amethod for investigating phantasies: TheThematic Apperception Test’, Archives

of Neurology and Psychiatry, 34(2):289–306.

Naglieri, J., McNeish, T. and Bardos, A. (1991)Draw-A-Person: Screening Procedure ofEmotional Disturbance. Austin: Proed.

Paulhus, D.L. (2002) ‘Socially desirableresponding: The evolution of a construct’, inH.I. Braun and D.N. Jackson (eds), The Roleof Constructs in Psychological andEducational Measurement. Mahwah, NJ:Erlbaum, pp. 49–69.

Payne, A.F. (1928) Sentence Completions. NewYork: New York Guidance Clinic.

Reynolds, C.R. and Kamphaus, R. (1992)Behavior Assessment System for Children.Circle Pines, MN: American Guidance Service.

Reynolds, C.R. and Kamphaus, R.W. (2004)Behavior Assessment System for Children(2nd edn manual). Circles Pines, MN:American Guidance Service.

Riccio, C.A. and Rodriguez, O.L. (2007)‘Integration of psychological assessmentapproaches in school psychology’, Psychologyin the Schools, 44(3): 243–55.

Roberts, B.W., Walton, K.E. and Viechtbauer,W. (2006a) Patterns of mean-level change inpersonality traits across the life course: A meta-analysis of longitudinal studies.Psychological Bulletin, 132(1): 1–25.

Roberts, B.W., Walton, K.E. and Viechtbauer, W.(2006b) ‘Personality traits change in adulthood: Reply to Costa and McCrae(2006)’, Psychological Bulletin, 132(1): 29–32.

Rogers, C.R. and Dymond, R.F. (1954) (eds),Psychotherapy and Personality Change.Chicago, IL: University of Chicago Press.

Rorschach, H. (1921) Psychodiagnostik. Berne:Birchen.

Rotter, J.B. (1975) ‘Some problems and misconceptions related to the construct ofinternal versus external control of reinforce-ment’, Journal of Consulting and ClinicalPsychology, 43(1): 56–67.

Saklofske, D.H. and Eysenck, H.J. (1994) ‘Extra-version–introversion’, in V.S. Ramachandran(ed.), Encyclopedia of Human Behavior. NewYork: Academic, pp. 321–32.

Saklofske, D.H. and Zeidner, M. (1995)International Handbook of Personality andIntelligence. New York: Plenum.

Sattler, J.M. (2001) Assessment of Children:Cognitive Applications (4th edn). San Diego:CA: Jerome M. Sattler, Publisher, Inc.


9781412946520-Ch01 5/28/08 12:21 PM

Schuerger, J.M. (1986) ‘Personality assessmentby objective tests’, in R.B. Cattell and R.C.Johnson (eds), Functional PsychologicalTesting: Principles and Instruments. NewYork: Brunner/Mazel, pp. 260–87.

Schmitz, P.G. (2004) ‘On the alternative five-factor model: Structure and correlates’, inR.M. Stelmack (ed.), On the Psychobiology ofPersonality: Essays in Honor of MarvinZuckerman. Amsterdam: Elsevier, pp. 65–87.

Schneider, R.J., Hough, L.M. and Dunnette, M.D.(1995) ‘Broadsided by broad traits: How tosink science in five dimensions or less’, Journalof Organizational Behavior, 17(6): 639–55.

Schutte, N.S., Malouff, J.M., Hall, L.E., Haggerty,D.J., Cooper. J.T., Golden, C.J. and Dornheim,L. (1998) ‘Development and validation of ameasure of emotional intelligence’, Personalityand Individual Differences, 25(2): 167–77.

Smith, D.A. (2002) ‘Validity and values:Monetary and otherwise’, AmericanPsychologist, 57(2): 136–7.

Tellegen, A. (1982) ‘Brief manual of theMultidimensional Personality Questionnaire’, Unpublished manuscript, University ofMinnesota.

Tellegen, A. (1995) ‘Multidimensional PersonalityQuestionnaire’, Unpublished document,University of Minnesota.

Thurstone, L.L. and Thurstone, T. (1930) ‘A neurotic inventory’, Journal of SocialPsychology, 1(1): 3–30.

Vaillant, G.E. (1977) Adaptation to Life.Boston, MA: Little Brown.

Woodworth, R.S. (1919) ‘Examination of emo-tional fitness for warfare’, PsychologicalBulletin, 16(2): 59–60.

Zuckerman, M. and Kuhlman, D.M. (2000)‘Personality and risk-taking: Common biosocial factors’, Journal of Personality,68(6): 999–1029.

Zuckerman, M. (2005) Psychobiology ofPersonality (2nd rev edn). New York:Cambridge University Press.


9781412946520-Ch01 5/28/08 12:21 PM

personality measurement and testing: an overview

Documents