Personality and Individual Differences 70 (2014) 85–91

Contents lists available at ScienceDirect

Personality and Individual Differences

journal homepage: www.elsevier .com/locate /paid

Psychological reactance in psychiatric patients: Examining thedimensionality and correlates of the Hong Psychological Reactance Scalein a large clinical sample

http://dx.doi.org/10.1016/j.paid.2014.06.0270191-8869/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Address: Department of Internal Medicine, Dermatologyand Psychiatry, School of Medicine, Campus de Ofra s/n, 38071 San Cristóbal de LaLaguna, Spain. Tel.: +34 609 521 405; fax: +34 922 319 353.

E-mail addresses: cdelascuevas@gmail.com (C. De las Cuevas), wpenate@ull.es(W. Peñate), moibemo@ull.es (M. Betancort), luisderivera@gmail.com (L. de Rivera).

Carlos De las Cuevas a,⇑, Wenceslao Peñate b, Moisés Betancort c, Luis de Rivera d

a Department of Internal Medicine, Dermatology and Psychiatry, University of La Laguna, San Cristóbal de La Laguna, Canary Islands, Spainb Department of Personality, Assessment and Psychological Treatments, University of La Laguna, San Cristóbal de La Laguna, Canary Islands, Spainc Department of Clinical Psychology, Psychobiology, and Methodology, University of La Laguna, San Cristóbal de La Laguna, Canary Islands, Spaind Department of Psychiatry, University Autonoma de Madrid, Madrid, Spain

a r t i c l e i n f o a b s t r a c t

Article history:Received 4 April 2014Received in revised form 11 June 2014Accepted 16 June 2014

Keywords:Factor structureHong Psychological Reactance ScalePsychiatric outpatients

This study investigated the factor structure and psychometric properties of the Spanish version of theHong Psychological Reactance Scale (HPRS) in psychiatric outpatient care, and how socio-demographicand clinical variables are related to this measure of trait reactance proneness. We carried out a cross-sectional survey involving seven hundred and ten consecutive psychiatric outpatients that completed theHPRS, health locus of control, self-efficacy and drug attitude scales, in addition to a questionnaire includingsocio-demographic and clinical variables. A confirmatory factor analysis to test the dimensionality ofthe HPRS was performed. Results supported that the best-fitting model of reactance processes was atwo-factor structure including affective and cognitive dimensions whose understanding and interactionappear essential to develop effective persuasive clinical messages. Further analyses yielded significantresults with age, educational level, number of drugs prescribed, health locus of control dimensions and atti-tudes toward psychiatric treatment but not with sex, self-efficacy or psychiatric diagnoses. Psychologicalreactance is a longstanding but still promising construct. Our results confirm that a two-factor structureis reasonable for assessing psychological reactance in psychiatric patients and provides an opportunity tounderstand patients’ health behavior.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Patients actively involved in their health and healthcare tend tohave better outcomes and, some evidence suggests, lower costs(Hibbard & Greene, 2013). Nevertheless, some factors attributableto patients and their characteristics and proclivities need to beovercome to carry out effective patient engagement.

In this sense, psychological reactance is an aversivemotivational state that functions to restore an individual’s percep-tions of autonomy in response to regulations or impositions thatimpinge on freedom and autonomy (Brehm, 1966, 1972; Brehm& Brehm, 1981), particularly when individuals feel obliged to adopta particular opinion or engage in a specific behavior. Althoughinitially investigated as a state phenomenon, it has become evident

that individuals are likely to vary as regards their trait propensityto experience reactance (Shen & Dillard, 2005).

The Hong’s Psychological Reactance Scale (HPRS; Hong &Faedda, 1996) was devised for use in the general population tomeasure the individual difference in reactance proneness, that is,a person’s trait propensity to experience psychological reactance.Although psychometric properties of the scale have been subjectto extensive study and the scale has been used in several studies(Dillard & Shen, 2005; Hellman & McMillin, 1997; Hong,Giannakopoulos, Laing, & Williams, 1994; Joubert, 1990, 1992),there is little agreement in terms of the factor structure of thismeasure, which has ranged from an initial four-factor structureto a one-dimensional solution (Hong, 1992; Donnell, Thomas, &Buboltz, 2001; Jonason, 2007; Shen & Dillard, 2005).

Prior research has revealed a considerable amount of negativeconsequences resulting from psychological reactance (Steindl &Jonas, 2012). Within the field of mental health, patients’ perceptionsof limiting or threatening freedoms or control may induce nonad-herence with prescribed treatments so that recommendationsto follow a drug regimen have the potential to elicit reactance

86 C. De las Cuevas et al. / Personality and Individual Differences 70 (2014) 85–91

and, as a result, lead individuals to ignore the recommended treat-ment (Fogarty & Youngs, 2000; Hong, 1992; Moore, Sellwood, &Stirling, 2000), as well as play a role in boosting the efficacy ofpsychotherapy and dealing with client resistance to this (Carver,1991; Dowd, 1990, 1993; Horvath & Goheen, 1990). Althoughpsychological reactance has been present in the field of psychologyfor over 50 years, this potentially useful construct is rarely used inpsychiatric clinical practice and is not even cited in manytextbooks.

The aim of this study was to assess the factor structure andpsychometric properties of the Spanish version of the HongPsychological Reactance Scale in psychiatric outpatient care, andto investigate how socio-demographic and clinical variables arerelated to this measure of trait reactance proneness.

2. Material and methods

2.1. Sample

From October 2013 to January 2014, nine hundred and ten con-secutive psychiatric outpatients seen in the Community MentalHealth Services on Tenerife Island (Canary Islands, Spain) wereinvited to participate in a cross-sectional study; a total of 710accepted. Patients were eligible for inclusion in the study if theywere aged 18 and over and were diagnosed by their psychiatristswith psychiatric disorders using the International Classificationof Diseases, Tenth Edition (ICD-10) codes F20 (schizophrenia),F31 (bipolar affective disorder), F32–33 (depressive episode andrecurrent depressive disorder), F40–48 (obsessive–compulsive dis-order and other neurotic, stress-related and somatoform disor-ders), and F60–69 (Disorders of adult personality and behavior).Each participant received a full explanation of the study, afterwhich they signed an informed consent document approved bythe clinical research ethics committee of Nuestra Señora de Cande-laria University Hospital in Santa Cruz de Tenerife. Each participantthen filled out a brief socio-demographic survey and the remainingquestionnaires.

2.2. Measures

2.2.1. Socio-demographic characteristics and clinical variablesAge, sex, educational level, psychiatric patient history, and class

of psychoactive drugs currently taken were assessed. For assess-ment purposes the drugs were split into the common groups ofpsychotropic drugs: antidepressants, benzodiazepines, antipsy-chotics and mood stabilizers. For statistical analysis purposes, anew variable (number of different drugs) was drawn up as anindirect measure of treatment complexity. We also recorded howlong patients had been under psychiatric treatment (in months),the number of different psychiatrists treating them during thattime, and the number of psychiatric admissions specifying theirvoluntary or involuntary character. Psychiatrists responsible forpatient mental health care were asked about patient diagnosis.

2.2.2. Hong Psychological Reactance Scale (HPRS)The Hong Psychological Reactance Scale (Hong & Faedda, 1996;

Pérez García, 1993) is a 14-item self-report questionnaire designedto measure the individual difference in reactance proneness, thatis, a person’s trait propensity to experience psychologicalreactance. Psychological Reactance (Wallston, 1992) assumes that,when an individual’s freedom is threatened, the individual will bemotivated to restore the perceived loss of freedom. Participantsindicated the extent to which they endorsed each statement on afive-point Likert scale (ranging from 1 = strongly disagree to5 = strongly agree).

2.2.3. Multidimensional Health Locus of Control (MHLC) Form C ScaleForm C of the multidimensional health locus of control (MHLC)

scale (Wallston, Stein, & Smith, 1994) is an 18-item, generalpurpose, condition-specific locus of control scale that could easilybe adapted for use with any medical or health-related condition.There are four subscales of the form C of the MHLC: (1) internalhealth locus of control (IHLC), which is the belief that one’s ownbehaviors affect one’s health status; (2) chance health locus ofcontrol (CHLC), which is the belief that one’s health condition isa matter of fate, luck or chance; (3) doctors (DHCL) health locusof control, which is the belief that it is doctors who determinethe outcomes of patient health; and (4) other people health locusof control (PHLC), which is the belief that other people, such asfamily and friends have control over one’s health status. Internaland chance subscales are comprised of six items, while doctorsand other people subscales are comprised of three items, totaling18 items on the questionnaire. Patients are asked to rate, on asix-point Likert scale, the degree to which they agree or disagreewith each statement. Higher scores on each subscale indicate astronger belief in that kind of control.

2.2.4. General Perceived Self-Efficacy ScaleThe General Perceived Self-Efficacy Scale (GSE) (Schwarzer &

Jerusalem, 1995) is a 10-item self-report scale that measuresgeneral self-efficacy as a prospective and operative construct. Incontrast to other scales designed to assess optimism, this scaleexplicitly refers to personal agency, i.e., the belief that one’s actionsare responsible for successful outcomes. Each item is scored from 1(not at all true) to 4 (completely true). The summary score rangesfrom 10 to 40, with the highest score indicating high self-efficacy.

2.2.5. Drug Attitude InventoryThe Drug Attitude Inventory (DAI-10) (Hogan, Awad, &

Eastwood, 1983) is a 10-item self-report scale devised to measurethe subjective responses and attitudes of psychiatric patientstowards their treatment by revealing whether patients are satisfiedwith their medications and assessing their understanding of howthe treatment is affecting them. Items represent subjectiveexperience presented as self-report statements with which thepatient agrees or disagrees. These are based on actual recordedand transcribed accounts of patients, and response options aretrue/false only. Each response is scored as +1 if correct or �1 ifincorrect. The final score is the grand total of the positive andnegative points and ranges in value from �10 to 10, with higherscores indicating a more positive attitude towards medication. Apositive total score means a positive subjective response; anegative total score means a negative subjective response.

2.3. Data analysis

Frequency was analyzed to describe the sample. An initialexploratory factor analysis (EFA) was performed to identify theempirical structure of the Hong Psychological Reactance Scale. Toperform this analysis, a random sample (approximately 50% ofthose participants without missing data) was used. To obviatethe confounding variable of related factors, an oblique rotationwas carried out. Factors with an eigenvalue equal or higher than1.0 were considered. To assign items to factors, loading coefficientsequal or higher than .30 were taken into account. When one itemloaded in more than one factor, the item was assigned to a factorwith a higher loading coefficient. Confirmatory factor analysis(CFA) was performed to test this empirical structure and otherstructures derived from previous studies. Once again, to avoidtautologic errors with EFA, a random sample representing 50% ofparticipants was generated. Convergent validity and the relationshipwith socio-demographic variables were analyzed with the Pearson

C. De las Cuevas et al. / Personality and Individual Differences 70 (2014) 85–91 87

correlation procedure. To test gender differences, a student’s t testwas performed. Finally, to test whether there were any differencesin psychological reactance among different mental disorders, acovariance analysis was performed (taking into account possiblecovariate variables). To perform CFAs, R statistical software wasused (R Development Core Team, 2008). For the remaining analy-ses SPSS 19 (IBM Corp. Released, 2010) was used.

3. Results

We recorded a high response rate of 78%, which resulted in asample of 710 psychiatric outpatients. Table 1 outlines the sample

Table 1Socio-demographic and clinical characteristics of the sample studied.

Variable Category

Age 18–30 years30–45 years

Mean Age 49.4 ± 13.9 45–60 yearsRank 18–88 60–75 years

>75 years

Sex MaleFemale

Educational level Can read and writePrimarySecondaryUniversity

ICD-10 diagnosis⁄ SchizophreniaBipolar disorderRecurrent depressive disordeDepressive episodeObsessive–compulsive disordAnxiety disordersPersonality disordersOther diagnoses

History of psychiatric admissions No12

59.9% Involuntary 3P4

No. of psychiatrists 12

Mean 2.7 ± 2.1 3Rank 1–12 4

P5

Psychotropic drugs No drugsOne drug

Mean 2.9 ± 1.4 drugs Two drugsRank 0–10 Three drugsPolypharmacy 84.9% Four drugs

Five or more drugs

Treatment AntidepressantsTricyclicsSSRIsSNSRIs

BenzodiazepinesAntipsychotics

ConventionalAtypical

Mood stabilizersAnticholinergics

Form C MHLC scales

InternalChanceDoctorsOther people

General Self-Efficacy Scale

Abbreviations: ICD, International Classification of Diseases; SNRIs, Selective Noradrenalintidimensional Health Locus of Control.

distribution according to socio-demographic and clinical variablesincluded in the study in addition to the scores recorded on Form Cof the MHLC and GSE Scales.

An exploratory factor analysis was initially performed to iden-tify the empirical structure of the HPRS. As previous studies havecorroborated, HPRS items are related; consequently, factor analysiswas carried out according to an oblique rotation on principal com-ponents extraction. To prevent bias with subsequent confirmatoryanalyses, a subsample was obtained at random, which representedapproximately 50% of participants (without missing data), asgenerated by the SPSS statistical package. The KMO coefficientobtained was .86, and Barlett’s test led to obtaining X2(91) =1040.74 (P = .000). These data revealed the feasibility of factor

Number of cases % Of the sample

76 10.7192 27.0300 42.3117 16.524 3.4

263 37.1446 62.9

57 8.0237 33.4286 40.3130 18.3

128 18.088 12.4

r 163 23.0173 24.4

er 13 1.8112 15.822 3.111 1.5

468 65.988 12.449 6.941 5.864 9.0

244 34.4185 26.1108 15.259 8.3114 16.0

15 2.192 13.0200 28.2175 24.6127 17.9101 14.2

492 69.330 4.2366 51.5252 35.5567 79.9235 33.137 5.2227 32.0202 28.533 4.6

Mean ± SD

24.4 ± 7.315.0 ± 6.815.3 ± 3.410.7 ± 3.7

29.2 ± 6.7

e Reuptake Inhibitors; SSRIs, Selective Serotonin Reuptake Inhibitors; MHLC, Mul-

88 C. De las Cuevas et al. / Personality and Individual Differences 70 (2014) 85–91

analysis. The rotated factor structure offered a two-factor solution.Table 2 summarizes this structure.

As can be observed, this structure represents the two-factorsolution. There are several relevant secondary loadings, represent-ing the fact that both factors are closely related. Pearson correla-tion was highly statistically significant (rxy = .58, P = .000). Thiscoefficient indicates a second-order one-factor solution. The firstfactor (items 4, 6, 7, 8, 12, 14) represents an affective dimensionof reactance (‘‘I become frustrated. . .’’, ‘‘It irritates me. . .’’, ‘‘Ibecome angry. . .’’), and the second factor (items 1, 2, 3, 5, 9, 10,11, 13) represents a cognitive dimension of reactance (‘‘I find itcontradicting. . .’’, ‘‘. . . I usually think. . .’’, ‘‘The thought of beingdependent. . .’’). The internal consistency for the affective andcognitive dimensions was .75, and .70, respectively.

Together with this solution, a CFA (maximum likelihood proce-dure) was carried out to test the fit of the initial four-factor solu-tion and the corrected four-factor to 11 items (Hong, 1992; Hong& Faedda, 1996; Hong & Page, 1989; Thomas, Donnell, & Buboltz,2001), as well as the two-factor (Brown, Finney, & France, 2009)and one-factor solution (Jonason, 2007; Jonason & Knowles,2006). The independence of error terms was specified, and thefactors were correlated. A random subsample representing 50% ofparticipants was generated again. Table 3 collates the principalcoefficients found: the v2 Goodness of Fit Test, the Root MeanSquare Error of Approximation (RMSEA), Normed Fit Index (NFI),the Goodness of Fit Index (GFI), and the Comparative Fit Index(CFI). To interpret a good fit of the data, we assumed that for theRMSEA, the value must be close to or less than .05 to indicate agood fit, and values as high as .08 indicate a reasonable fit. NFIvalues equal or higher than .80 indicate a good fit. As regards theCFI and GFI, values close to or greater than .95 must be attained.

Observing the fit coefficients, the one-factor model did notattain adequate fit indices. This data does not support the unidi-mensionality of HPRS. Moreover, the two-factor model based ona general factor explaining common variance and specific factorsalso explained the fact that additional variance did not attain agood fit index. However, the two dimensions affective/cognitivefactor model and the four-factor model (both 14-items and11-items) attained relevant fix indices, which were very similarbetween both models. There is only one index where the two-factor is better than the four-factor model: CFI. For the remainingcoefficients both models attained good indices; however, thetwo-factor model attained higher coefficients. With these data,

Table 2Hong Psychological Reactance Scale Items and Factor Structure with ob

HPRS items

1. Regulations trigger a sense of resistance in me2. I find contradicting others stimulating3. When something is prohibited, I usually think, ‘‘That’s exactly w4. The thought of being dependent on others aggravates me5. I consider advice from others to be an intrusion6. I become frustrated when I am unable to make free and indepe7. It irritates me when someone points out things, which are obvio8. I become angry when my freedom of choice is restricted9. Advice and recommendations usually induce me to do just the10. I am content only when I am acting of my own free will11. I resist the attempts of others to influence me12. It makes me angry when another person is held up as a role m13. When someone forces me to do something, I feel like doing th14. It disappoints me to see others submitting to standards and ru

Eigenvalue% of variance

Bold values correspond to the highest value recorded in each item of

the remaining analysis will use the two-factor structure, but alsoconsidering that the four-factor model has a correct fit.

This affective/cognitive structure attained a negative statisti-cally significant correlation with age (affective factor, rxy = �.19,P = .000; cognitive factor, rxy = �.14, P = .000), which indicates lessreactance as age increases. This relationship changes if we considerthe treatment duration. Now only the affective factor attained anegative, but less statistically significant correlation (�.09,P = .019). An inverse pattern was found for number of differentdrugs prescribed (as a measure of treatment complexity): thecognitive factor attained a positive and statistically significantcorrelation (.11, P = .004). Educational level attained a low butpositive and statistically significant correlation (.08, P = .038).According to gender, no differences were found [affective,t(607) = �.83; cognitive, t(607) = 1.73].

As convergent validity, the two-factor structure was correlatedwith locus of control dimensions and self-efficacy. Table 4 collatesthe correlation coefficients. Self-efficacy is independent regardingpsychological reactance (both affective and cognitive dimension),the correlation coefficients are close to zero. However, the relation-ship patterns with locus of control are of significance. In general,there are positive relationships, indicating more reactance as locusof control increases, with independent locus nature, with theexception of relying on doctors, as external locus of control dimen-sion. In this case, patients are less reactant as they have more con-fidence in their psychiatrists. Attitudes to medicines also revealstatistically significant correlations. In this case, the negativeresults indicate that there are more negative attitudes to drugsas patients are more reactant.

A final group of analysis was performed to contrast reactancedimensions among the participants’ diagnosis (schizophrenia,bipolar disorder, major depressive disorder, anxiety disorder, andpersonality disorder). No differences were found (affective factor,F[4,697] = .95; cognitive factor, F[4,697] = 1.10). Since, aspreviously noted, age, educational level, treatment duration, andnumber of different drugs can all influence the results, they wereconsidered as covariate variables. In this sense an ANCOVA wasperformed. A statistically significant difference was found for theaffective reactance dimension (F[4,697] = 2.7, P = .030; g2 = .015),but not for the cognitive dimension (F[4,697] = .54). By performingpost hoc analyses for affective reactance and diagnosis groups(Bonferroni correction), no specific differences were found amongthe five groups. With respect to the average scores of each patient

lique rotation on principal component extraction (n = 354).

Factors loadings

Affective Cognitive

.29 .52

.07 .71hat I am going to do’’ .11 .74

.71 .23

.40 .47ndent decisions .71 .31

us to me .68 .27.68 .39

opposite .31 .60.32 .47.39 .49

odel for me to follow .64 .14e opposite .50 .55les .52 .47

4.21 1.5130.08 10.78

the questionnaire in factor loading.

Table 3Fit Indexes and Model Comparisons of confirmatory factor analysis for the HongPsychological Reactance Scale.

Model X2 DF RMSEA NFI GFI CFI

One factor 299.59 77 .08 .80 .97 .84Two factors (general/specific) 173.37 73 .05 .89 .93 .91Two factors (affective/cognitive) 158.15 71 .05 .93 .99 .96Four factors 77.10 36 .05 .90 .98 .94Four factors – 11 items 77.10 36 .05 .93 .98 .94

Abbreviations: X2 = Chi square; DF = Degree of freedom; RMSEA = Root Mean SquareError of Approximation; NFI = Normed Fit Index; GFI = Goodness of Fit Index;CFI = Comparative Fit Index.

Table 4Correlations coefficients among psychological reactance factors, Self-efficacy, andLocus of control dimensions (n = 699).

Affective factor Cognitive factor

Self-efficacy .05 P = .185 .01 P = .781MHLC internal .14 P = .000 .12 P = .002MHLC chance .13 P = .000 .17 P = .000MHLC doctors �.10 P = .006 �.14 P = .000MHLC other people .18 P = .000 .04 P = .260Drug Attitude Inventory �.20 P = .000 �.18 P = .000

Abbreviations: MHLC = Multidimensional Health Locus of Control.

C. De las Cuevas et al. / Personality and Individual Differences 70 (2014) 85–91 89

group, Fig. 1 represents the distribution of affective and cognitivedimensions. As can be observed, all diagnostic categories obtainedaverage scores above the theoretical mean score (2.5) for theaffective dimension; personality disorders and anxiety disordersregistered higher scores and schizophrenia and bipolar disorderslower scores. Different data were recorded for the cognitivedimension because no diagnostic group scored higher than 2.22.

4. Discussion

Psychological reactance may help us understand the involve-ment of patients in health behaviors that promote coping withsymptoms, adherence with treatment, and prevention of relapses.This is the first large-scale, community psychiatry-based survey toexplore the relationship of socio-demographic, psychological andclinical variables with psychological reactance in psychiatricoutpatients.

Mental health professionals usually rely on their professionalstatus to convince patients to make necessary behavior changes(Elder, Ayala, & Harris, 1999). However, every persuasive attempt

3.17 3.22 3.34

2.17 2

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

Schizophrenia Bipolar Disorder DepreDisor

Affec�ve

Fig. 1. Psychological reactance dimensions me

has the potential to succeed and produce the desired result, or tofail, yielding no discernible effect or causing changes in attitudeor behavior at odds with the persuader’s intention (Guttman,Kegler, & McLeroy, 1996; Stewart & Martin, 1994). Becausedetermining the factor structure of psychological reactance wouldhelp clarify its nature (Jonason, Bryan, & Herrera, 2010), we havetried to identify this structure in a sample of psychiatric patientscompleting the Hong Psychological Reactance Scale. The one-factorsolution could not be verified. Whereas the four-factor solutionreached adequate fit coefficients, our results strongly support thetwo-factor solution. Nevertheless, the nature of our sample, withmost affective disorders, may influence this affective/cognitivesolution. The emotional method (versus a cognitive-rationalmethod) of processing information may be over-represented inour sample. The strength observed in the four-factor modelsuggests the need to test the HPRS in different populations.

The two-factor structure corresponds to two basic ways ofprocessing information: emotional (feelings about influences),and cognitive (thoughts against influences). Cognition and emotion‘compete’ in directing human behavior (Eder, Hommel, & DeHouwer, 2007) and the instrument studied introduces these twomental functions into the reactance construct, mediated by twoseparate but interacting brain systems (LeDoux, 1998, 2000).

As for socio-demographic characteristics, we have not foundany statistically significant sex-related differences in HPRSdimensions, in line with other studies carried out in the generalpopulation (Hong et al., 1994; Shen & Dillard, 2005). However,both affective and cognitive psychological reactance yielded astatistically significant negative age effect: the level of reactancetends to decrease with increasing age. Conformity of psychiatricpatients appears to increase with age with attitudes, beliefs, andbehaviors matched to group norms (Cialdini & Goldstein, 2004).The attitudes of older people change in response to personalexperience, which supports the lifelong openness model of attitudechange (Tyler & Schuller, 1991). However, when taking relatedvariables into consideration, the pure age effect attenuates whenadjusted for the influence of length of treatment. This effect maybe accounted for on three separate counts: one for its obviousconcomitant variation with age, the second by the gradualdecrease in reactance as time of actual compliance increases, andthird, by a possible progressive abandonment of follow-up by themore high-reactance patients.

In relation to locus of control, it has been known for some timethat patients differ in terms of desired control (Fogarty, 1997).Psychological reactance, as could be expected, shows an inversecorrelation with doctors’ health locus of control: the more a patientbelieves that their mental health depends on their psychiatrist, the

3.49 3.58

2.02 2.08 2.22

ssive der

Anxiety Disorder Personality Disorder

Cogni�ve

an scores according psychiatric diagnoses.

90 C. De las Cuevas et al. / Personality and Individual Differences 70 (2014) 85–91

less likely they are to be reactant to the psychiatrist’s advice, or inother words, the less psychological reactance, the more likely apatient is to believe that their health depends on following theirdoctor’s advice. Along the same lines, reactance also shows anegative correlation with drug attitude. The DAI-10 may be, in fact,an indicator of ‘‘medication’s health locus of control’’. The highcorrelation between Doctors’ Health Locus of Control and DrugAttitude supports this suggestion. This positive correlation mayalso show that belief in the doctor’s influence on personal healthis mainly related to belief in the doctor’s drug managementabilities. We may expect a less clear correlation in different patientgroups, i.e., patients following or demanding psychotherapeutictreatments.

Conversely, the more a patient considers themself responsiblefor their health, the more likely they are to be reactant to outsideinfluences, thus the positive correlation between internal healthlocus of control and psychological reactance.

The positive correlation between reactance and chance locus ofcontrol suggests that patients who attribute their health to chancefactors consider that rejection of advice (high reactance) is a logicalbehavior. The same reasoning may be applied to the positivecorrelation between reactance and others’ health locus of control.In this case, however, the effect of others’ locus of control onreactance is somewhat attenuated: only the affective componentshows a positive correlation, suggesting that, in this case, patientsmay accept cognitively external advice and influence, but continuerejecting affectively those influences.

As for the relationship between psychiatric diagnosis andreactance, there is no influence of diagnosis on the cognitive com-ponent of reactance. The affective component is higher in patientswith personality disorders, followed by those with anxiety anddepressive disorders, and is lower in psychotic patients. Depres-sive, anxious and personality disorders may share an emotionalinstability that makes them more reactant, in line with the con-cept of general neurotic syndrome (Tyrer, Seivewright, Ferguson,& Tyrer, 1992; Tyrer, Seivewright, & Johnson, 2003). All patientstend to think that it is appropriate to follow advice, but those withpersonality and ‘‘emotional’’ disorders keep an intimate affectivereactance against it. In contrast, psychotic patients are moreprone to feel a need for advice and regulations. We may hypoth-esize that external regulations serve a supportive function in psy-chotic patients, while they are more likely experienced asundesirable impositions by those with personality and emotionaldisorders.

4.1. Clinical implications

This paper is exploratory in nature, and should be read withcaution; generalizations should be avoided in the absence of directempirical evidence of a link between reactance and healthbehavior. When those links are confirmed, psychological reactancecould be used to select newly diagnosed patients, and to ensurethat they receive a type of psychiatric treatment provision thatreduces the likelihood of inducing reactance, and increases theprobability of perceiving treatment as beneficial.

References

Brehm, J. W., & Brehm, S. S. (1981). Psychological reactance: A theory of freedom andcontrol. San Diego, CA: Academic Press.

Brehm, J. W. (1966). A theory of psychological reactance. New York: Academic Press.Brehm, J. W. (1972). Responses to loss of freedom: A theory of psychological reactance.

Morristown, NJ: General Learning Press.Brown, A. R., Finney, S. J., & France, M. K. (2009). Examining the dimensionality of

the Hong Psychological Reactance Scale. Paper presented at the annual meetingof the Northeastern Educational Research Association, Rocky Hill, CT.

Carver, C. S. (1991). Self-consciousness and reactance. Journal of Research inPersonality, 15, 16–29.

Cialdini, R. B., & Goldstein, N. J. (2004). Social influence: Compliance andconformity. Annual Review of Psychology, 55, 591–621.

Dillard, J. P., & Shen, L. (2005). On the nature of reactance and its role in persuasivehealth communication. Communication Monographs, 72, 144–168.

Donnell, A. J., Thomas, A., & Buboltz, W. C. (2001). Psychological reactance:Factor structure and internal consistency of the Questionnaire for theMeasurement of Psychological Reactance. Journal of Social Psychology,141(5), 679–687.

Dowd, E. T. (1990). A cognitive theory of resistance and reactance: Implications fortreatment. Journal of Mental Health Counseling, 12, 458–469.

Dowd, E. T. (1993). Motivational components of client reactance. Journal ofCounseling and Development, 71, 533–538.

Eder, A. B., Hommel, B., & De Houwer, J. (2007). How distinctive is affectiveprocessing? On the implications of using cognitive paradigms to study affectand emotion. Cognition and Emotion, 21(6), 1137–1154.

Elder, J. P., Ayala, G. X., & Harris, S. (1999). Theories and intervention approaches tohealth–behavior change in primary care. American Journal of PreventiveMedicine, 17(4), 275–284.

Fogarty, J., & Youngs, G. A. (2000). Psychological reactance as a factor in patientnoncompliance with medication taking: A field experiment. Journal of AppliedSocial Psychology, 30, 2365–2391.

Fogarty, J. S. (1997). Reactance theory and patient noncompliance. Social Science &Medicine, 45(8), 1277–1288.

Guttman, N., Kegler, M., & McLeroy, K. R. (1996). Health promotion paradoxes,antimonies, and conundrums. Health Education Research, Theory, and Practice,11, 1–13.

Hellman, C. M., & McMillin, W. L. (1997). The relationship between psychologicalreactance and self-esteem. The Journal of Social Psychology, 137(1),135–139.

Hibbard, J. H., & Greene, J. (2013). What the evidence shows about patientactivation: Better health outcomes and care experiences; fewer data on costs.Health Affairs, 32(2), 207–214.

Hogan, T. P., Awad, A. G., & Eastwood, R. (1983). A self-report scale predictive ofdrug compliance in schizophrenics: Reliability and discriminative validity.Psychological Medicine, 13, 177–183.

Hong, S. M., & Faedda, S. (1996). Refinement of the Hong psychological reactancescale. Educational and Psychological Measurement, 56, 173–182.

Hong, S. M., Giannakopoulos, E., Laing, D., & Williams, N. A. (1994). Psychologicalreactance: Effects of age and gender. Journal of Social Psychology, 134, 223–228.

Hong, S. M., & Page, S. (1989). A psychological reactance scale: Development, factorstructure and reliability. Psychological Reports, 64, 1323–1326.

Hong, S. M. (1992). Hong’s psychological reactance scale: A further factor analyticvalidation. Psychological Reports, 70, 512–514.

Horvath, A. V., & Goheen, M. D. (1990). Factors mediating the success of defiance-and compliance-based interventions. Journal of Counseling Psychology, 37,363–371.

IBM Corp. Released 2010. IBM SPSS Statistics for Windows, Version 19.0. Armonk,NY: IBM Corp.

Jonason, P., & Knowles, H. (2006). A unidimensional measure of Hong’sPsychological Reactance Scale. Psychological Reports, 98, 569–579.

Jonason, P. K., Bryan, A., & Herrera, J. (2010). Trimming the fat reveals a one-factorconstruct of Hong’s Psychological Reactance Scale. Individual DifferencesResearch, 8(4), 220–228.

Jonason, P. K. (2007). Further tests of validity for the one-dimensional Hong’spsychological reactance scale. Psychological Reports, 101(3), 871–874.

Joubert, C. E. (1992). Antecedents of narcissism and psychological reactance asindicated by college students’ retrospective reports of their parents’ behavior.Psychological Reports, 70(3), 1111–1115.

Joubert, C. E. (1990). Relationship among self-esteem, psychological reactance, andother personality variables. Psychological Reports, 66(3), 1147–1151.

LeDoux, J. E. (1998). The emotional brain: The mysterious underpinnings of emotionallife. New York: Touchstone.

LeDoux, J. E. (2000). Emotion circuits in the brain. Annual Review of Neuroscience, 23,155–184.

Moore, A., Sellwood, W., & Stirling, J. (2000). Compliance and psychologicalreactance in schizophrenia. British Journal of Clinical Psychology, 39, 287–295.

Pérez García, A. M. (1993). [Spanish Adaptation of Hong and Page PsychologicalReactance Scale]. Informe del Departamento de Psicología de la Personalidad,Evaluación y Tratamientos Psicológicos. Madrid: UNED.

R Development Core Team (2008). R: A language and environment for statisticalcomputing. R Foundation for Statistical Computing. Vienna, Austria: <http://www.R-project.org>.

Schwarzer, R., & Jerusalem, M. (1995). Generalized Self-Efficacy scale. In J.Weinman, S. Wright, & M. Johnston (Eds.), Measures in health psychology: Auser’s portfolio. Causal and control beliefs (pp. 35–37). Windsor, England: NFER-NELSON.

Shen, L., & Dillard, J. P. (2005). Psychometric properties of the Hong psychologicalreactance scale. Journal of Personality Assessment, 85(1), 74–81.

Steindl, C., & Jonas, E. (2012). What reasons might the other one have? Perspective-talking to reduce psychological reactance in individualists and collectivists.Psychology, 3, 1153–1160.

Stewart, D. W., & Martin, I. M. (1994). Intended and unintended consequences ofwarning messages: A review and synthesis of empirical research. Journal ofPublic Policy and Marketing, 13, 1–19.

C. De las Cuevas et al. / Personality and Individual Differences 70 (2014) 85–91 91

Thomas, A., Donnell, A., & Buboltz, W. C. (2001). The Hong psychological reactancescale: A confirmatory factor analysis. Measurement and Evaluation in Counselingand Development, 34, 2–13.

Tyler, T. R., & Schuller, R. A. (1991). Aging and attitude change. Journal of Personalityand Social Psychology, 61(5), 689–697.

Tyrer, P., Seivewright, N., Ferguson, B., & Tyrer, J. (1992). The general neuroticsyndrome: A coaxial diagnosis of anxiety, depression and personality disorder.Acta Psychiatrica Scandinavica, 85(3), 201–206.

Tyrer, P., Seivewright, H., & Johnson, T. (2003). The core elements of neurosis: Mixedanxiety–depression (cothymia) and personality disorder. Journal of PersonalityDisorders, 17(2), 129–138.

Wallston, K. A. (1992). Hocus-pocus, the focus isn’t strictly on locus: Rotter’s sociallearning theory modified for health. Cognitive Therapy and Research, 16, 183–199.

Wallston, K. A., Stein, M. J., & Smith, K. A. (1994). Form C of the MHLC Scales: Acondition-specific measure of locus of control. Journal of Personality Assessment,63, 534–553.