2013 - vertical and lateral workplace bullying in nursing - development of the hospital aggressive...

8/10/2019 2013 - Vertical and Lateral Workplace Bullying in Nursing - Development of the Hospital Aggressive Behaviour Scale

1/25

http://jiv.sagepub.com/Violence

Journal of Interpersonal

http://jiv.sagepub.com/content/28/12/2389The online version of this article can be foundat:

DOI: 10.1177/0886260513479027

2013 28: 2389 originally published online 28 March 2013J Interpers ViolenceJos Antonio Jimnez-Barbero

Kathrin Waschgler, Jos Antonio Ruiz-Hernndez, Bartolom Llor-Esteban andthe Hospital Aggressive Behaviour Scale

Vertical and Lateral Workplace Bullying in Nursing: Development of

Published by:

http://www.sagepublications.com

On behalf of:

American Professional Society on the Abuse of Children

can be found at:Journal of Interpersonal ViolenceAdditional services and information for

http://jiv.sagepub.com/cgi/alertsEmail Alerts:

http://jiv.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://jiv.sagepub.com/content/28/12/2389.refs.htmlCitations:

Whatis This?

- Mar 28, 2013OnlineFirst Version of Record

- Jul 29, 2013Version of Record>>

at Universidad de Murcia on August 5, 2013jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/content/28/12/2389http://jiv.sagepub.com/content/28/12/2389http://jiv.sagepub.com/content/28/12/2389http://www.sagepublications.com/http://www.apsac.org/http://jiv.sagepub.com/cgi/alertshttp://jiv.sagepub.com/cgi/alertshttp://jiv.sagepub.com/subscriptionshttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsPermissions.navhttp://jiv.sagepub.com/content/28/12/2389.refs.htmlhttp://jiv.sagepub.com/content/28/12/2389.refs.htmlhttp://jiv.sagepub.com/http://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://jiv.sagepub.com/content/early/2013/03/20/0886260513479027.full.pdfhttp://jiv.sagepub.com/content/28/12/2389.full.pdfhttp://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://jiv.sagepub.com/content/early/2013/03/20/0886260513479027.full.pdfhttp://jiv.sagepub.com/content/28/12/2389.full.pdfhttp://jiv.sagepub.com/content/28/12/2389.refs.htmlhttp://www.sagepub.com/journalsPermissions.navhttp://www.sagepub.com/journalsReprints.navhttp://jiv.sagepub.com/subscriptionshttp://jiv.sagepub.com/cgi/alertshttp://www.apsac.org/http://www.sagepublications.com/http://jiv.sagepub.com/content/28/12/2389http://jiv.sagepub.com/


2/25

Journal of Interpersonal Violence

28(12) 23892412

The Author(s) 2013

Reprints and permissions:

sagepub.com/journalsPermissions.navDOI: 10.1177/0886260513479027

jiv.sagepub.com

Article

Vertical and LateralWorkplace Bullying in

Nursing: Developmentof the HospitalAggressiveBehaviour Scale

Kathrin Waschgler,1Jos Antonio

Ruiz-Hernndez, PhD,1

Bartolom Llor-Esteban, PhD,1and

Jos Antonio Jimnez-Barbero1

AbstractHealthcare staff is one of the professional groups that suffers the highestexposure to sources of occupational stress such as hostility from coworkersand superiors.

In order to contribute to the assessment of bullying behaviors in thehealthcare sector and to obtain a brief and manageable instrument for theassessment of this psychosocial risk, we developed the Hospital AggressiveBehaviour Scaleversion Co-workers-Superiors (HABS-CS).

By means of thorough qualitative analysis, an initial pool of 166 items wasobtained, which were reviewed according to precise criteria until concludingwith a total of 57 items, which were administered to a sample of 1,484healthcare professionals from 11 public hospitals. The analyses concludedwith the selection of 17 items distributed in two subscales. The internal5-factor structure is the result of exploratory and confirmatory factoranalysis conducted in two samples. Both the resulting questionnaire and thefactors identified present adequate psychometric properties: high-internal

1University of Murcia, Murcia, Spain

Corresponding Author:

Kathrin Waschgler, Universidad de Murcia- SERPA, C/Actor Isidoro Maquez 9, Edif. Saavedra

Fajardo, 30007 Murcia- Spain.

Email: [email protected]

JIV

28

12

10.1177/0886260513479027Journal of Interpersonal ViolenceWaschgler etal.research-article

2013

http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/


3/25

2390 Journal of Interpersonal Violence 28(12)

consistency (Cronbachs of .86) and adequate criterion validity, analyzed bymeans of significant correlations between the HABS-CS and job satisfaction,burnout components, and psychological well-being. This instrument may be

of great utility for the assessment and prevention of psychosocial risks.

Keywords

lateral workplace bullying, vertical workplace bullying, nursing, scale development

Diverse studies identify nursing as a risk group for workplace bullying andconfirm that the problem of hostility at the workplace is very common in thehealthcare sector (Anderson & Parish, 2003; Di Martino, 2002; Estryn-Behar et al., 2008; Kivimki, Elovainio, & Vahtera, 2000; Leymann &Gustafsson, 1996; Niedl, 1995; Quine, 2001; Rippon, 2000; Vartia, 1996;Waschgler, Ruiz-Hernndez, Llor-Esteban, & Garca-Izquierdo, 2012; Zapf,1999). Zapf, Einarsen Hoel, and Vartia (2003) explain this fact as follows:Working in the social and health sector requires a high level of personalinvolvement, which means sensing and expressing emotions and building up

personal relationships. The higher the level of personal involvement, themore personal information is available, and the more possibilities for beingattacked exist (p. 119).

A stressful working environment such as a hospital could lead to anincrease of the risk of interpersonal conflict, with bullying as a possibleoutcome. Appelberg, Romanov, Honkasalo, and Koskenvuo (as cited inSalin & Hoel, 2011) identified time pressure and a hectic work environ-ment as sources of interpersonal conflict, and Zapf, Knorz, and Kulla

(1996) conclude that time pressure may indirectly affect bullying becauseof fewer opportunities to resolve conflicts. Bullying is related to poorlyorganized work environments, where roles and command structures areunclear, such as nurses in hospital settings. Nursing personnel often workunder diverse authorities and face controversial demands from doctors onthe one hand and supervisors on the other. The consequences of thisdilemma are increased pressure and conflicts (Leymann, 1996). Low team-work quality correlates with lateral workplace bullying, which is negative

behavior perpetrated by coworkers (Estryn-Behar et al., 2008). Anotherfactor related to workplace bullying behaviors is the size of the organiza-tion. In large organizationssuch as hospitalsit is easier to accept hos-tile behaviors and there is less chance of social condemnation (Einarsen,Raknes, & Matthiesen, 1994).



4/25

Waschgler et al. 2391

Bullying at the Workplace: Definition, Actors, and Categories

The increase of studies in this field in the recent years has not led to an agree-

ment on how to define and operationalize the phenomenon. For the purposeof the present study, we adopt a similar definition to the one cited in Einarsen,Hoel, Zapf, and Cooper (2003): Bullying at work means harassing, offend-ing, socially excluding someone or negatively affecting someones worktasks . . . to be applied to a particular activity, interaction or process, it has tooccur repeatedly and regularly (e.g., weekly) and over a period of time (e.g.,6 months) (p. 15).

The measures of hostility used in the literature do not permit the differ-entiation between hostile behaviors by actors or multiple behaviors by asingle actor. Most research using behavioral checklists does not specificallyfocus on hostile behaviors by particular actors. This fact makes it difficultto operationalize facets of pattern and escalation, where it is important totake into account the relationship between actors and targets. A variety ofbehaviors coming from one actor may be experienced differently thanbehaviors coming from a variety of actors (Keashly & Jagatic, 2003).Analyzing studies on the frequency of workplace bullying, the findingssuggest that most victims were bullied by supervisors, followed by col-

leagues, and, finally, by subordinates (Zapf, Escartn, Einarsen, Hoel, &Vartia, 2011). Leymann (1993) introduced mobbingas the definition of alasting conflict among colleagues. In his study, there were only marginallymore colleagues than supervisors among the bullies. In a study of Rutherfordand Rissel (2004), they found that, in most cases, the aggressors werecoworkers, followed by users and supervisors or center directors. Otherstudies have confirmed high rates of lateral violence (Farrell, Bobrowski, &Bobrowski, 2006; Rippon, 2000).Vartia and Hyyti (1999, as cited in Zapf et

al., 2003) found that women were more often bullied by coworkers, whereasmen were more often bullied by immediate supervisors or managers. Thelateral aggressive behaviors perpetrated by coworkers habitually describedin the studies are tone of voice or facial expressions that make you feelbad and intimidating behaviors such as contempt, disparaging comments,or shouting. The hostile behaviors by superiors in this study were mostlytone of voice or inappropriate facial expressions, pressure due to excessiveand impossible demands, intimidating behaviors such as contempt, dispar-

aging comments, or shouting.Taking the literature as a whole, Einarsen, Hoel, Zapf, and Cooper (2011)and Zapf et al. (2011) concluded that most researchers differentiate betweentwo factors of bullying behaviors: person-related bullying (verbal aggression,spreading gossip or rumors, persistent criticism, playing practical jokes,

http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/


5/25


intimidation, emotional abuse, attacking the victims private life, personalderogation, humiliation, and isolation or social exclusion) and work-relatedbullying (giving unreasonable deadlines or unmanageable workloads, exces-

sive monitoring of work, assigning meaningless tasks or even no tasks, con-trol and manipulation of information, and control and abuse of workingconditions).

Diverse studies conclude that organizational measures affecting victimstasks and competences, social isolation, attacking the private person, verbalaggression, and spreading rumors are typical categories of bullying (Niedl,1995; Zapf et al., 1996). When bullying comes from superiors, work-relatedbullying seems to be the most frequently applied behavior (Hoel & Cooper,

2000, as cited in Zapf et al., 2011), whereas coworkers use social isolationand attacking the private sphere more often (Zapf, 1999; Zapf et al., 1996). Itseems clear that the context in which workplace bullying occurs will influ-ence which behavioral category is most frequent. Although some researchers(Leymann, 1996; Niedl, 1995; Zapf et al., 1996) include physical abuse intheir categorization of bullying, they all agree that the behaviors involved inworkplace bullying are mainly of a psychological rather than a physicalnature. It should also be acknowledged that there are some cultural differ-

ences in the use of bullying behaviors (Escartn, Zapf, Arrieta, & Rodrguez-Carballeira, 2011; Moreno-Jimnez, Muoz, Salin, & Morante-Benadero,2008).

Measuring Workplace Bullying and Epidemiology

Diverse strategies for measuring bullying behaviors have been employed:subjective ones, such as the Self-labeling Method (Einarsen & Skogstad,1996; Vartia, 1996); operational or objective ones, such as the BehavioralExperience Method (LIPT, Leymann, 1990; NAQ, Einarsen & Raknes,1997), and the Latent Class Cluster analysis (Notelaers, Einarsen, de Witte,& Vermunt, 2006). A strong point of the objective method is its reliability,because respondents do not need to evaluate whether or not they are victimsof workplace bullying. Such lists can therefore be seen as valuable instru-ments when studying the nature of bullying behavior in different occupations(Salin, 2001).

The results of these studies indicated that using lists of negative acts pro-

vides a considerable higher victimization rate than self-judgments, and thatthe two strategies do not even necessarily identify the same victims. Notelaerset al. (2006) verified that, in their study, the Operational Classification Methodestimated almost seven times more victims of workplace bullying than theLatent Class Cluster method (20% vs. 3%). It is important to consider that the



6/25


frequency of bullying depends very much on how it is measured, and the mea-sures, again, are influenced by the general understanding of what bullying is(Zapf et al., 2003).

Regarding prevalence rates, Zapf (1999) showed that employees of thehealthcare and social sector had a sevenfold risk of being bullied. InScandinavian studies, workplace bullying prevalence rates range from 3% to16% (Einarsen & Skogstad, 1996; Kivimki et al., 2000). Niedl (1995)reported a prevalence rate of 26.6% in an Austrian hospital, and Minkel (ascited in Zapf et al., 2011) found 8.7% of workplace bullying victims inGermany. In the UK, percentages range from 18% to 40% (Paice, Aitken,Houghton, & Firth-Cozens, 2004; Quine, 1999; Tehrani, 2004). In Spanish

studies, in a mixed population from the educational and nursing sector, aprevalence rate of 14% in the nursing sector was found (Garca-Izquierdo,Llor, Garca-Izquierdo, & Ruiz, 2006).

Health and Well-Being Correlates

Depending on the reported negative behavior, the impact on the target per-sons health may also differ significantly, although the more frequent the

exposure to hostile workplace behavior, the greater the negative effects(Hoel, Faragher, & Cooper, 2004).Exposure to bullying behaviors can have devastating effects on the health

and well-being of individuals. Strong correlations between perceived bully-ing and lowered psychological well-being have been found: (a) increasedlevels of depressive symptoms, (b) anxiety symptoms, (c) psychosomatichealth complaints, and (d) psychiatric pathologies (Agervold & Mikkelsen,2004; Einarsen, Matthiesen, & Skogstad, 1998; Einarsen & Raknes, 1997;Hogh, Mikkelsen, & Hansen, 2011; Keashly & Jagatic, 2003; Leymann &Gustafsson, 1996; Matthiesen & Einarsen, 2004; Meseguer, Soler, Sez, &Garca, 2008; Mikkelsen & Einarsen, 2002; Niedl, 1995; Quine, 1999;Rippon, 2000; Rospenda, Richman, & Shannon, 2009; Tehrani, 2004; Zapf etal., 1996).

With regard to professional-organizational effects, significant correlationsbetween perceived bullying behaviors and overall job satisfaction have beenfound (Einarsen et al., 1998; Einarsen & Raknes, 1997; Hoel, Sheehan,Cooper, & Einarsen, 2011; McNeese-Smith, 1999; Quine, 1999). Bullying

has been regarded as a severe social stressor at work (Niedl, 1995; Zapf et al.,1996), causing: burnout, intention to leave the organization, higher sick leaveand absenteeism (Agervold & Mikkelsen, 2004; Estryn-Behar et al., 2008;Hoel et al., 2011; Keashly & Jagatic, 2003; Kivimki et al., 2000; Quine,2001). Higher turnover, consequences in productivity, risk of unemployment,



7/25


and earlier retirements are also organizational consequences (Hoel et al.,2011; Hogh et al., 2011; Leymann, 1996).

The diverse sources of conflicts found in the research literature suggest

that negative behaviors by coworkers are associated with depression, andnegative relations with supervisors are linked to global physical health (Israel,House, Schurman, Heaney, & Mero, 1989). Victims bullied by their superiorshave been shown to suffer more than the victims of coworker bullying(Einarsen & Raknes, 1997). Of the different strategies used, attacks on thevictims private life and personal derogation have been shown to have thestrongest correlation with psychological ill health (Einarsen & Raknes, 1997;Zapf et al., 1996).

Aim of the Study

Taking into account the abovementioned prevalence rates and health corre-lates of workplace bullying in nursing, it can be considered as an emergingglobal problem that has serious effects on victims, organizations, healthcaresystem, and quality of life of the population. Although workplace bullying innursing has been assessed using instruments such as the NAQ (Einarsen &

Raknes, 1997), these instruments do not permit differentiation betweendiverse actors of workplace bullying. Therefore, in our study, we want tohighlight the importance of the different sources of hostility in the healthcaresector. So, we extend the definition of workplace bullying in the healthcaresector to include, as a whole, hostile behaviors received by a healthcareworker from colleagues (lateral workplace bullying) and superiors (verticalworkplace bullying). The aim of the present study is, therefore, to develop aninstrument that permits the evaluation of workplace bullying behaviors fromdifferent sources in the nursing staff by exploring this phenomenon throughthese two dimensions: lateral and vertical workplace bullying. The instru-ment should be easy to use, short, and helpful to explore the different originsof such behaviors and, thus, it should be useful for risk prevention in thehealthcare sector.

Method

Participants

In a first, qualitative, phase of the study, three focus groups were carried outwith a total of 21 participants (16 women and 5 men; 14 were healthcareprofessionals, and 7 were technical personnel for prevention of occupationalhazards of the healthcare service), and six in-depth interviews of nursesbelonging to diverse hospitals were performed.



8/25


In a second, quantitative, phase aimed at the psychometric assessment ofthe instrument, the research questionnaire was applied to a total sampleof 1,484 healthcare professionals from 11 public hospitals of the Region of

Murcia. The sample represents 28.98% of the total nursing staff and 18.96%of the total of auxiliary nursing staff of the Region of Murcia. The globalresponse rate of the protocols administered was 70.48%.

The sociodemographic and sociolabor characteristics are presented inTable 1. The mean age of the sample was 42 years (SD= 9.75), and 83.4%were women. Of the sample, 62.8% were nurses and 37.2% were auxiliarynursing staff. Of them, 64.3% had a permanent contract and 51.1% worked inrotating shifts. Mean job tenure in the current post was 7 years and 5 months,

and mean job tenure in the profession was 15 years and 4 months. One thirdof the sample reported one sick leave in the past 12 months.

Procedure

Figure 1 shows an outline of the procedure employed. In the first qualitativephase, following the guidelines of Krueger (1991), two trained interviewersand two observers directed three discussion groups with a previously estab-

lished script. Through this procedure, we collected information about therelation between coworkers and superiors as sources of hostility in the health-care staffs work. Within the framework of qualitative analysis of the focusgroups, after transcription, we identified categories of aggressive behaviors(verbal, nonverbal, and physical), which were the basis to draft the items foreach potential area of conflict, respecting the content of the focus groups tothe utmost. The items were (a) reviewed by a group of experts, by consensus,and following explicit and previously established criteria (De Vellis, 1991;Dunn-Rankin, 1983; Togerson, 1958); (b) submitted to semantic validationby means of in-depth interviews of healthcare professionals; and (c) re-reviewed by the group of experts. This group of experts was made up of threeuniversity professors, two doctoral candidates, the general director of thenursing department of a hospital, and five practicing nurses.

The items were drafted to express potentially hostile or conflictive situ-ations that can emerge at the healthcare staffs workplace. A 6-pointLikert-type response scale was used to rate the frequency of exposure tothe type of hostility at work, ranging from 0 (never) to 5 (daily). Thus, a

high score indicated higher frequency of exposure to hostile situations atthe workplace.

In a second, quantitative, phase, after obtaining authorization for the study,all the public hospitals of the Region of Murcia participated. We randomlyselected a first group of participants from five hospitals (N= 790), which wasused to elaborate the instrument; and a second sample from six hospitals



9/25


Table 1. Sociodemographic and Sociolabor Characteristics.

Variable N %

Age (years) < 30 148 10.7

30-39 440 31.6

40-49 432 31.2

50 368 26.5

Gender

Male 244 16.6

Female 1,227 83.4

Marital status

Single 386 26.2

Married/common-law partner 1,005 68.1

Divorced, separated, widow/er 85 5.8

Did you take sick leave in the past 12 months?

Yes 486 33.3

No 975 66.7

Profession

Nursing 922 62.8

Auxiliary nursing personnel 546 37.2

Provenance-Hospital

University hospitals 1,194 80.4 Associated University Hospitals 290 19.4

Current type of shift

Permanent morning shift 444 30.2

Permanent night shift 37 2.5

Rotating shift 750 51.1

Others (morning/afternoon shift)

237 16.1

Type of contract

Permanent 883 64.3

Temporary 491 35.7

Job tenure (years)

0-2 360 25.3

2-5 417 29.3

5-10 293 20.6

10-15 151 10.6

+ 15 202 14.2

Profession tenure (years)

0-2 19 1.4

2-5 122 8.7 5-10 386 27.7

10-15 278 19.9

15-20 208 14.9

+20 383 27.4



10/25


Figure 1. Outline of the procedure.



11/25


(N = 694), which was used to validate the instrument. The protocol wasadministered through the service directors, using a randomized procedure to

avoid overrepresentation of the people who were more sensitized to the prob-lem. Envelopes were handed out with the protocols for them to be returnedtherein, in order to guarantee the participants anonymity and confidentiality.Participation was voluntary and the participants were identified by numbercodes, thus maintaining strict confidentiality in the treatment and use of thedata.

Instruments

Besides the sociodemographic and occupational variables, in order to studythe properties of the instrument, we used the following scales that assessdiverse psychosocial variables (see Table 2):

The Overall Job Satisfactionof Warr, Cook, and Wall (1979), adapted byPrez and Hidalgo (1995). It includes 15 items divided into two subscales:Intrinsic Satisfaction, which addresses aspects such as acknowledgment ofwork, responsibility, promotion, etc.; and Extrinsic Satisfaction with aspectsof work organization such as the schedule, remuneration, the physical condi-

tions of the work, etc. The degree of satisfaction or dissatisfaction with eachone of the items is rated on a 7-point scale, ranging from 1 (very dissatisfied)to 7 (very satisfied).In the present study, a Cronbach of .84 was obtainedfor Intrinsic Satisfaction, an of .70 for Extrinsic Satisfaction, and an of .87for the total scale.

Table 2. Psychosocial Variables (Descriptive Statistics and Reliability of ValidationScales).

Items M SD N

Total satisfaction 15 .87 70.69 13.49 1,461

Intrinsic satisfaction 7 .84 32.58 7.30 1,462

Extrinsic satisfaction 8 .70 38.11 6.98 1,461

MBIEmotional exhaustion 5 .85 9.07 5.39 1,473

MBICynicism 5 .70 6.81 5.09 1,468

MBIProfessional efficacy 6 .86 21.79 8.71 1,442

GHQTotal 28 .92 19.58 10.07 1,455

GHQSomatization 7 .86 6.12 3.87 1,472

GHQAnxiety insomnia 7 .90 5.54 4.27 1,474

GHQSocial dysfunction 7 .74 6.60 1.97 1,466

GHQMajor depression 7 .88 1.33 2.60 1,462

Note: MBI = Maslach Burnout Inventory; GHQ = General Health Questionnaire;



12/25


The Maslach Burnout InventoryGeneral Survey of Schaufeli, Leiter,Maslach, and Jackson (1996) in the Spanish version of Salanova, Schaufeli,Llorens, Peir, and Grau (2000). This scale has 16 items grouped into three

dimensions: Emotional Exhaustion, which refers to the loss of emotionalresources due to work; Cynicism, which reflects indifference and distant atti-tudes toward work; and Professional Efficacy, the efficacy perceived whencarrying out the work. All the items are rated on a 7-point scale, ranging from0 (never) to 6 (always). The coefficients for each one of the subscales were.85, .70, and .86, respectively.

The General Health QuestionnaireGHQ-28 (Goldberg & Hillier,1979) according to the adaptation of Lobo, Prez-Echeverra, and Artal

(1986). This questionnaire is a measure of general psychological health andcomprises four subscales with seven items each: (a) Somatic Symptoms ofpsychological origin, such as feelings of exhaustion, weakness or illness,and bodily discomfort; (b) Anxiety and Insomnia; (c) Social Dysfunction,such as problems to perform and enjoy responsibilities, daily activities; and(d) Depressive Symptomatology, such as thoughts and feelings of personalworthlessness, sadness, hopelessness, and suicide. Table 2 shows the Cronbach values of these subscales (.85, .86, .74 and .82, respectively) and

of the total scale (.91).

Data Analysis

To analyze construct validity, we used both exploratory and confirmatoryfactor analysis, Cronbach and Meehl (1955) confirm that such analyses arethe most important type of validation.

In the stage of elaborating the instrument of the quantitative phase (seeFigure 1), we used a group of 790 participants. To reduce the number of itemsand to refine the scale, we carried out principle component exploratory factoranalysis with varimax rotation, and various criteria were combined: (a) eachfactor should explain at least 5% of the total variance; (b) in the factorsselected, the factor loading of the items should be at least .50; (c) an itemcould not load on two factors with more than .40; and (d) the items containedin each factor should have adequate internal consistency ( > .70; Nunnally& Bernstein, 1994).

To validate the scale obtained in the first stage, we used a second group

with 694 participants. To analyze the fit of the factor structure to the resultingmodel of exploratory factor analysis, we performed confirmatory factor anal-ysis. Following the habitual recommendations for this analysis (Hu & Bentler,1999), diverse fit indexes were employed: Root mean square error of approx-imation (RMSEA), comparative goodness-of-fit index (CFI), and normed fit



13/25


14/25


abovementioned criteria. The application of this methodology led to a finalscale of 17 items made up of five factors. The factors explained 74.17% ofthe variance, and the KaiserMeyerOlkin sample adequacy measurement

value was high (KMO = .865) and adequate (George & Mallery, 1995).Table 3 shows the rotated solution of the analysis, the factor loadings ofeach item, as well as the percentages of variance explained by each factor,and the Cronbach values for the total scale and each factor.

The first factor, called Active Workplace Bullying-Superior explains37.16% of the variance, with an of .95, and includes four items aboutviolent acts by supervisors/directors of nursing departments, on whom theprofessional depends hierarchically (i.e., My superior threatens to attack

me).The second factor, which we called Person-related Bullying-Colleagues,explains 15.51% of the variance, with an of .79, and it has four items (i.e.,Some co-workers spread false rumours about me, Some co-workers criti-cize my private life).

The third factor, Passive Workplace Bullying-Colleagues, explains 8.76%of the variance, with an of .82, and is made up of three items (i.e., Someco-workers have stopped talking to me, Some co-workers ignore me and

do not answer my questions).Factor IV, Passive Workplace Bullying-Superior, explains 6.6% of thevariance, with an of .79, and made up of three items (i.e., My superiorignores me).

The last factor, Work-related Bullying-Colleagues, explains 6.15% of thevariance, with an of .72, and contains three items (i.e., Some co-workersdeliberately accuse me of other peoples mistakes).

We call the instrument thus elaborated the Hospital Aggressive BehaviorScale- Co-Workers and Superiors (HABS-CS), in which factors related tocoworkers are considered to be part of lateral workplace bullying, and factorsrelated to superiors are vertical workplace bullying behaviors.

Confirmatory Factor AnalysisConstruct Validity

of the HABS-CS

Subsequently, we used the validation sample to test the fit of the factor model.The analyses indicated a goodness-of-fit index of .91 (GFI > .90), a compara-

tive goodness-of-fit index of .94 (CFI > .93), and a normed fit index andTucker-Lewis index of .92 (NFI, TLI > .90), and, lastly, an RMSEA index of.079 (RMSEA < .08). Taking into account the reference criteria, we concludedthat the model presents a good fit (see Table 4). The five factors were alsosignificantly correlated with each other at the .001 level (see Table 5).



15/25


Table 3. Exploratory Factor Analysis.

Items FI FII FIII FIV FV

77 My superior threatens toattack me.

0.938

73 My superior makes unwelcomesexual insinuations or gestures

0.931

81 My superior has restrained ortouched me in a hostile way

0.860

78 My superior recommends mycoworkers not to talk to me

0.858

95 Some coworkers spread false

rumors about me

0.805

96 Some coworkers play ironic jokeson me

0.713

115 Some coworkers criticize myprivate life

0.712

105 Some coworkers criticize the wayI work

0.706

121 Some coworkers have stoppedtalking to me

0.847

124 Some coworkers ignore me or donot answer my questions

0.807

122 I get contemptuous looks or angryfaces from my coworkers

0.756

82 My superior ignores me 0.854

94 My superior undervalues myinitiatives

0.834

85 My superior impairs myparticipation in training, teaching,

or research activities

0.746

110 Some coworkers deliberatelyaccuse me of other peoplesmistakes

0.761

113 Some coworkers try to preventme from learning the habitualtechniques of my service

0.754

118 Some coworkers exclude me frominformation about the Service

(changeovers, change of protocol,etc.)

0.671

% Explained variance Total:74.17

37.16 15.51 8.76 6.60 6.15

Total: .863 0.95 0.79 0.82 0.79 0.72



16/25


Reliability of the HABS-CS

Reliability analysis in the elaboration group (N= 790) yielded a Cronbach of .86 for the total scale, and the values of the factors ranged between .95for Factor I and .72 for Factor V (see Table 3). The validation group (N= 694)and the total sample (N= 1,484) obtained Cronbach values that were practi-cally identical. All the values are higher than the common criterion value of

.70 (Nunnally & Bernstein, 1994).

Correlates of the HABS-CSStudy of Criterion Validity

In order to analyze the criterion validity, we calculated the correlationsbetween the factors of the HABS-CS and job satisfaction, burnout compo-nents, and the four health components assessed with the GHQ-28. In thisphase of the study, we used the entire sample of all 11 hospitals (N= 1,484).

In accordance with the bibliographic review carried out, we found signifi-cant correlations (p< .001) between the factors of the HABS-CS and the vali-dation scales. Table 6 shows the correlations found between the diverseHABS-CS factors and job satisfaction, burnout components, and psychologi-cal well-being.

Table 4. Fit Indexes for the Models of the HABS Coworkers and Superiors.

X2 df RMSEA GFI CFI NFI TLI (=NNFI)

Sample 1 550.40 109 .08 .91 .94 .92 .92Sample 2 651.08 109 .08 .91 .93 .92 .92

Note: RMSEA = Root mean square error of approximation; GFI = goodness-of-fit index; CFI =goodness-of-fit index; NFI = normed fit index; TLI = Tucker-Lewis index.

Table 5. Correlations Among the Five Factors of the HABS-CS.

M SD

WB

Superiors -Active

WB

Superiors-Passive

WB

Coworkers-Personal

WB

Coworkers-Passive

WB Superiorsactive 1.02 .21

WB Superiorspassive 1.14 .50 .430***

WB Coworkerspersonal 1.47 .77 .241*** .262***

WB Coworkerspassive 1.25 .65 .277*** .264*** .572***

WB Coworkerswork 1.17 .50 .346*** .361*** .492*** .546***

Note:WB = Workplace Bullying.***The correlation is significant at the .001 level (two-tailed).



17/25

2404

Table6.

Co

rrelationsBetweenFactorsandPsychosocialV

ariables.

P

sychosocialV

ariables

Factors

Satisfaction

E

motional

Exhaustion

Professional

Efficacy

Cynicism

GHQ

TOT

GHQ

SOM

GHQ

ANS

GHQ

DIS

GHQ

DEP

Workplacebullying

Superiors

total

.2

14***

.161***

.009

.208***

.145***

.119

***

.106***

.061*

.160***

Workplacebullying

Superiors

active

.0

45

.052*

.025

.107***

.052*

.036

.038

.0

08

.092***

Workplacebullying

Superiors

passive

.2

63***

.192***

.000

.219***

.171***

.147

***

.125***

.095***

.163***

Workplacebullying

Coworkerstotal

.2

23***

.244***

.0

35

.246***

.289***

.228

***

.255***

.108***

.248***

Workplacebullying

Coworkers

personal

.2

05***

.260***

.0

37

.228***

.268***

.228

***

.251***

.094***

.204***

Workplacebullying

Coworkerswork

.1

88***

.157***

.0

13

.189***

.230***

.145

***

.189***

.118***

.235***

Workplacebullying

Coworkerspassive

.1

81***

.169***

.0

32

.193***

.222***

.182

***

.197***

.068**

.188***

Note:*p

2013 - vertical and lateral workplace bullying in nursing - development of the hospital aggressive...

Documents