Marlowe Crowne Social Desirability Scale and Short Form C: Forensic Norms

Transcription

1 Marlowe Crowne Social Desirability Scale and Short Form C: Forensic Norms Paul Andrews Private Practice, Tyler, Texas Robert G. Meyer University of Louisville Marlowe Crowne Social Desirability Scale (MC) (Crowne & Marlowe, 1960) scores were collected on 1096 individuals involved in forensic evaluations. No prior publication of forensic norms was found for this instrument, which provides a measure of biased self-presentation (dissimulation). MC mean score was for the sample. Also calculated was the score on Form C (MC-C) (Reynolds, 1982), and the mean for this 13-item scale was The scores for the current sample generally are higher than those published for non-forensic groups, and statistical analysis indicated the difference was significant for both the MC and MC-C (d.75 and.70, respectively, p.001). Neither gender nor educational level proved to be significant factors in accounting for variance, and age did not appear to be correlated with scores. Group membership of subjects based on referral reason (family violence, abuse, neglect, competency, disability) was significant for both the MC and MC-C scores. Results suggest the MC or MC-C can be useful as part of a forensic-assessment battery to measure biased self-presentation Wiley Periodicals, Inc. J Clin Psychol 59: , Keywords: deception; social desirability; forensic evaluation; Marlowe Crowne; dissimulation Despite the fact that the Marlowe Crowne Social Desirability Scale (MC) (Crowne & Marlowe, 1960) was developed to measure a person s tendency to distort self-presentation toward a socially desirable bias, relatively little use has been made of it in clinical practice. It has been used in test development to measure the association between validity The authors wish to thank Kara Marsiony and Chris Fletcher for help with data entry and Lisa Reynolds and Chris Weaver for assistance with statistical analyses. Correspondence concerning this article should be addressed to: Paul Andrews, P.O. Box 6691, Tyler, TX 75711; pand75711@tyler.net. JOURNAL OF CLINICAL PSYCHOLOGY, Vol. 59(4), (2003) 2003 Wiley Periodicals, Inc. Published online in Wiley InterScience ( DOI: /jclp.10136

2 484 Journal of Clinical Psychology, April 2003 scales and social desirability as a response set (Milner, 1986; Morey, 1991), and it has found a place as a measure of defensiveness in medical research (Deshields, Tait, Gfeller, & Chibnall, 1995; Helmers, Krantz, Merz, Klein, et al., 1995; Mann & James, 1998). Some authors are suggesting use of the MC as part of a battery to detect positive impression management (Boyd & Meyer, in preparation; Edens, Buffington, Tomicic, & Riley, 2001). However, a literature search (PsychINFO Database) revealed little information on clinical use other than for research purposes and no forensic norms for this instrument. Information about a subject s response bias is an important component of any clinical assessment and becomes essential in a forensic evaluation. Heilbrun (1995) noted the need to assess the individual s response style in making interpretation of assessment information in child-custody cases, and studies have found that parents in custody litigation typically score.5 to 1 standard deviation above the mean on MMPI-2 L and K scales in efforts to present themselves well (Bathurst, Gottfried, & Gottfried, 1997; Miller, Lape-Brinkman, Roche, Cox, & Motheral, 1999). Favorable self-presentation can take many forms, running a gamut from outright lying to putting one s best foot forward, from being ingratiating with the examiner to simply showing polite manners. The tendency toward presenting self in a socially desirable manner is one aspect of a complex matrix of factors, motives, and behavior patterns affecting choices about behavior and self-revelation. One task of the examiner is then to attempt to determine the extent and direction of altered self-presentation and to take into account its occurrence when interpreting assessment results. Whereas the MC is well suited to assist in this task, an absence of clinical norms has limited its use. Originally developed as an alternative to the Edwards Social Desirability Scale, which was oriented toward admission or denial of symptoms of maladjustment that had been selected from items on the MMPI or the Manifest Anxiety Scale, the MC was designed to consist of items deemed by psychology faculty and graduate students to have minimal pathological implications, whether items were endorsed or denied by a subject. A pool of 50 items thought to be sanctioned culturally, but improbable to occur, was collected by Crowne and Marlowe (1960), who asked ten psychology faculty and graduate students to score the items True or False in the socially desirable direction. There was unanimous agreement on 36 items and 90% agreement on another 11 items. This pool of 47 items was submitted to a second group of faculty and graduate students who rated the degree of exceptional adjustment or maladjustment implied by answering in the socially desirable direction. Results showed that the mean rating of the inventory was almost neutral on a 5-point Likert scale. With the scale being pathologically neutral, the resulting score would reflect tendency toward positive self-presentation rather than extent of denial of pathological symptoms. The preliminary scale then was administered to students in two introductory psychology courses where 33 items were found to discriminate between high and low total scores at.05 level or better. These 33 items became the MC Social Desirability Scale. Psychometric research on the MC has shown that it can be used with confidence by evaluators in a forensic context. It is a solid instrument that has good to excellent statistical properties and has been critiqued in psychological literature for more than forty years. Studies have reported internal consistency/reliability scores ranging from.72 to.96 (Ballard, 1992; Crowne & Marlowe, 1960; Fischer & Fick, 1993; Loo & Thorpe, 2000; Reynolds, 1982) and one month test retest correlation of.89 (Crowne & Marlowe, 1960). In addition to the original 33-item MC, several short forms have been developed and researched. Form C (MC-C), a 13-item alternate version of the MC proposed by Reynolds (1982), has been found to have acceptable to good psychometric properties, with

3 Marlowe Crowne and Short Form C: Forensic Norms 485 internal consistency estimates ranging from.62 to.76 (Ballard, 1992; Loo & Thorpe, 2000; Reynolds, 1982; Zook & Sipps, 1985) and six-week test retest correlation of.74 (Zook & Sipps, 1985). Scores on this form have been found to correlate highly with the scores on the original MC with values of.91 to.965 reported (Fischer & Fick, 1993; Loo & Thorpe, 2000; Reynolds, 1982). Criterion-related validity information about the MC reported by Edens et al. (2001) indicated the scale had an Area-Under-Curve (AUC) value of.83. This value represents the probability using all possible cut scores that a randomly selected protocol in the Fake Good condition would have a higher score than one randomly selected from the nondissimulation condition. The MC mean for the Fake Good condition was 24, whereas it was only 15 in the honest condition (t(185) 15.34, p.001). Although the scale is accepted widely as a measure of social desirability, McCrae and Costa (1983) criticized the MC on the basis of construct validity and suggested that it is actually not measuring defensive style of self-presentation, but a lack of neuroticism and perhaps a tendency toward extraversion. Using correlations between the MC and Neuroticism Extraversion Openness (NEO) Inventory scores produced by 215 community participants in the Baltimore Longitudinal Study, the authors argued that individuals who obtained high scores on the Marlowe Crowne were in fact better adjusted, friendlier, and more open to experience than those who scored low (p. 886). Others have found very low scores to be associated with self-critical responding, and Milner (1986) pointed out that low scores on the MC were effective in discriminating Fake Bad profiles on the Child-Abuse Potential Inventory. A limitation of the McCrae and Costa study is that it does not report actual MC scores (range or mean), so the reader does not know what scores were considered high and low or how these scores related to reported norms. Without such information, the authors criticism is not well supported because high scores in the sample might be within the average range for general norms. The authors also pointed out that their subjects were disinterested volunteers, a sample that might behave differently from individuals being assessed under conditions that pull for a biased self-presentation. In summary, the MC is well suited to provide an easy, brief measure of defensiveness (an essential factor to be assessed in forensic evaluations), but the present authors could find no normative data for use with forensic subjects. The purpose of this study is to present such norms and to show how these norms differ from reported means obtained from general population samples. Method Participants Participants were individuals involved in psychological evaluations, mostly for court purposes. Males (n 793) comprised most of the subjects (72%), while females (n 303) accounted for a little more than a quarter of the group (28%). All individuals were at least seventeen years old, and most were young to middle-aged adults. Age information was not available on all subjects in this archival study, but a subset of 313 subjects on whom age data was available showed a range of 17 to 83 years old, with a mean of 33.7 (SD 12.0). By far, most of the individuals in this subset (74%) were between 18 and 40 years old. Participants lived in Kentucky or Texas, where the evaluations were conducted. Education was coded as years of formal education. GED (general equivalency diploma) was coded as 10 years of education. Special-education backgrounds were evaluated for number of years of regular class years; when determination was unclear, an assignment of a value usually between 6 and 9 years was made based on reported school experience.

4 486 Journal of Clinical Psychology, April 2003 For purposes of analysis, participants were coded as belonging to one of nine groups; when a person qualified for membership in more than one group, categorization was determined by the referral issue. The groups and characteristics of the individuals are detailed below: 1. Physical-Abuse Offenders: Parents or parent figures with admitted, proven, or strongly alleged moderate or major physical abuse of a child (injuries evident); Child Protective Services (CPS) involved; court ordered/referred for testing and/or treatment in most cases. 2. Child-Neglect Offenders: Parents or parent figures with admitted, proven, or strongly alleged moderate or major neglect of a child, usually resulting in removal of child from home by CPS; includes neglectful supervision, medical neglect, physical neglect, failure to protect child from danger, or refusal to allow child to live in home (refusal to accept parental responsibility); CPS involved; court ordered/ referred for testing and/or treatment in most cases. 3. Domestic-Violence Offenders: Participants in court-ordered domestic-violence psychoeducational-treatment programs following court appearance on domesticviolence charge. 4. Child Sexual-Abuse Offenders: Adult males with admitted, proven, or strongly alleged sexual abuse of a child; CPS involved in most situations; court referred or ordered for testing and/or treatment. 5. Pretrial Competency Defendants: Felony defendants referred for court-ordered Competency-to-Stand-Trial evaluations; incarcerated in jail or forensic hospital. 6. Domestic-Violence Victims: Court or self-referred adults seeking psychological treatment because of at least one incident of domestic violence involving police; individual not considered instigator of violent incident. (Note: Many subjects in other groups also were current or prior victims of domestic violence, but participants in this group were referred specifically due to experiencing physical abuse in a domestic setting.) 7. Non-Offending Family Members in Child-Abuse Cases: Parents or parent figures in CPS cases being referred for evaluation for prospective placement of child or for assessment of parenting skills/deficits in cases where moderate to major abuse and/or neglect had occurred; either no allegations of abuse/neglect for the subject or no substantiation of abuse/neglect involving participant. 8. Disability Examinees: Adults referred as part of initial evaluation or re-examination connected with application for Social Security disability benefits. 9. Other: Diverse group of participants referred for psychological evaluation due to civil court proceedings (child custody or visitation, personal injury suit), criminal court matters (risk assessment for dangerousness), medical conditions (usually connected with injury claim), or professional board/administrative organizational issue (suitability for practice, personnel matters); a few participants were referred by treating therapists only for treatment diagnosis, but most involved forensic or quasi-forensic systems and issues. Procedure Each subject was given the MC as part of a more comprehensive evaluation. Except for Group 3 (Domestic-violence Offenders who were interviewed and given written inven-

5 Marlowe Crowne and Short Form C: Forensic Norms 487 tories by clinicians in charge of a psychoeducational group), participants were evaluated by the first author using a variety of written and examiner-administered psychological instruments. A few subjects unable to read the MC were presented the items by audiotape or orally by the examiner and asked to mark their own responses to statements. The inventories were hand scored to obtain scores for both the original version and the short form (MC-C). Data were analyzed using SPSS-PC software program. For this study, scores on the MC-C were derived from the responses subjects made to the MC. This procedure of extracting short-form scores from the long form has been used widely (Ballard, 1992; Fischer & Fick, 1993; Loo & Thorpe, 2000; Reynolds, 1982; Strahan & Gerbasi, 1972). Although Zook and Sipps (1985) criticized such an approach, they found that separate presentation of a short form led to psychometric results that compared favorably with results obtained by extracting short-form scores from the long form. Results Mean educational level for the sample was determined to be years, with a range from 3 to 27 years of formal education. Both median and mode were 12 years. Primarily because of the referrals from professional boards, Group 9 had the highest mean educational level by more than two years (14.78 years). Pretrial detainees had the lowest level of formal education (10.13 years) (see Table 1). Univariate analysis of variance was conducted with factors of group membership, sex, and educational level; only group was significant in accounting for variance of scores on MC (F(8,866) 2.122, p.031) or MC-C (F(8,864) 1.989, p.045). No interactive effects or any other main effects attained a level of statistical significance, although the mean for females in the sample was higher than that for males: (6.56) and (6.35), respectively. Because a single group (Group 5) comprised almost half of the total sample and was over 93% male (as compared to the entire sample being 72% male), data from Group 5 were removed so that analysis could check for possible gender differences in a more gender-balanced sample (which then had only 53.9% males). In this analysis of the remaining 581 subjects, sex still was not a significantly contributing variable; in this Table 1 Group Characteristics Characteristics Group n Male Female Mean Years Education Total (72.4%) 303 (27.6%) 11.56

6 488 Journal of Clinical Psychology, April 2003 Table 2 Group Means, Standard Deviations, and Correlations Between Marlowe Crowne and Form C Group n MC mean (SD) MC-C mean (SD) r r (6.24) 8.21 (3.23) (6.67) 8.74 (3.53) (5.97) 7.35 (3.16) (7.41) 7.03 (3.45) (7.48) 7.04 (3.50) (5.25) 6.75 (2.79) (6.56) 9.18 (3.44) (5.15) 7.06 (2.78) (6.82) 6.90 (3.69) Total (6.44) 7.61 (3.32) Note. All correlations significant at level p.001, two-tailed test. smaller sample, the effect of group fell slightly beyond the level of significance for the MC (F(7,407) 1.943, p.06) but remained significant for the MC-C (F(7,405) 2.105, p.04). Because age data was not available on all subjects, the information available on the subset of 313 individuals was used to explore any correlation between age and MC scores. The results of Pearson correlation analysis showed age to have no significant connection with scores on the MC or the MC-C (r.001, p.99 and r.005, p.93, respectively). Results seen in Table 2 show the MC mean score for the sample was (SD 6.44), with a range of 2 33 and a median of 20. MC-C mean score was 7.61 (SD 3.32). Correlation of MC and MC-C for the entire sample was r.913 ( p.001 level, two-tailed test). The pooled mean of the non-forensic samples from eleven cited studies shown in Table 3 is obtained simply by calculating the sum of [n mean] for each study and dividing Table 3 Marlowe Crowne Mean and Standard Deviation Scores from Other Studies Source n Mean (SD) Crowne & Marlowe (1960) 120 students (5.78) Crowne & Marlowe (1964) a 300 students 15.5 (4.4) Strahan & Gerbasi (1972) 272 students and community sample 14.5 (5.4) Reynolds (1982) 608 students (5.91) Fraboni & Cooper (1989) 231 students and community sample (6.01) Ballard (1992) 399 students (5.10) Dalton (1994) 14 male adoption applicants 19.9 (7.0) Dalton (1994) 14 female adoption applicants 22.6 (5.6) Vella-Brodrick & White (1997) 178 (Australian) community sample 15.8 (5.8) Loo & Thorpe (2000) 232 (Canadian) students and community (4.88) Edens et al. ( students 15.0 (6.1) Total 2554 students (5.5) a Reported in Vella-Brodrick & White (1997).

7 Marlowe Crowne and Short Form C: Forensic Norms 489 Table 4 Marlowe Crowne Form C Mean and Standard Deviation Scores from Other Studies?6? Source n Mean (SD) Reynolds (1982) 608 students 5.67 (3.20) Zook & Sipps (1985) 176 male students 4.02 (2.81) Zook & Sipps (1985) 253 female students 4.19 (2.97) Robinette (1991) 307 male military trainees 5.70 (3.57) Robinette (1991) 174 female military trainees 6.30 (3.67) Ballard (1992) 399 students 5.39 (2.85) Loo & Thorpe (2000) 232 Canadian students 5.76 (2.68) Total 2149 subjects 5.37 (3.13) total sum by the total N. The pooled standard deviation of the non-forensic samples is calculated by working backward to get the squares for each study (calculated as SD 2 multiplied by [n 1]) and taking the square root of the total squares that has been divided by [N number of groups]. With these calculations of pooled values and formulas presented by Zakzanis (2001), it is possible to see if there is actually a significant difference between the mean scores of the non-forensic group (15.08) and the current study s forensic subjects (19.42). Using the process above to calculate the pooled standard deviation (SD) for the entire 3650 subjects (forensic and non-forensic), Cohen s d can be calculated as a test of significance of the difference of the two means. Dividing the difference of the means ( ) by the pooled SD (5.8) yields d.75. Although there would be a distribution overlap of about 55% for the two samples, the value of d approaches the 0.8 level, which Cohen suggested as being a large effect size and far exceeds the 0.5 cut-off for medium effect size. Using the same process (see Table 4), the MC-C pooled standard deviation is found to be 3.19 (non-forensic and forensic data). Cohen s d is.70, which indicates that the distribution overlap of the two groups (non-forensic and forensic samples) is about 57%. As with the MC, the value is not a large effect size, but far exceeds the cut-off for medium effect size. The statistical significance of d for these studies can be calculated using a t-test formula presented by Rosenthal (1984): t d=df02. Using this formula, the resulting value of t for the MC is 31.99, p.001. The value of t for MC-C is 28.16, p.001. With both the MC and MC-C, the non-forensic scores and the forensic scores appear to come from different populations. Discussion As with other studies, no gender effects were found to be statistically significant in this study. Only the research by Fraboni and Cooper (1989) was found in a literature review to show males having higher mean scores than females on the MC, whereas Robinette (1991), Zook and Sipps (1985), Dalton (1994), and the current study found that females produced higher mean scores on the instrument. One may speculate as to why this statistically non-significant trend seems to show up across studies; it is interesting that the domestic-violence victims in this study (90% female) had the lowest mean score, whereas the other two predominantly female groups (family members: 62% female; neglect offenders: 86% female) had the two highest group mean scores. It may well be that the experi-

8 490 Journal of Clinical Psychology, April 2003 ences bringing females into these evaluations had a greater effect on their self-presentation than precipitating events had on the men s approach to the task. Certainly, anyone familiar with the personality characteristics of domestic-violence victims is aware of the psychological problems and negative self-image that usually typify these individuals. On the other extreme, it is conceivable that females accused of neglect or under scrutiny (if not accusation) of having contributed to abuse of a child would be very defensive; societal expectations for good parenting are generally higher for mothers than for fathers. Gender differences of defensiveness within a forensic context might be explored more directly. Neither education level nor age was found to be a significantly correlated variable in regard to MC or MC-C scores in this study. For the entire sample, education failed to be identified as making a significant contribution to the variance of scores. In the limited subset for which data on age was available (n 313), there was almost no relationship between age and scores. Although Fraboni and Cooper (1989) found a small but statistically significant amount of variance explained by age and SES in their study using a modified MC (31 items), neither of these factors nor education level has been identified often as a significant variable in MC studies. In fact, some studies do not even report information about age of subjects (Ballard, 1992; Crowne & Marlowe, 1960) or report only a mean age (Loo & Thorpe, 2000; Vella-Brodrick & White, 1997), whereas information on SES or education is seldom reported or explored even though other measures of dissimulation, such as MMPI-2 L and K scales, have been found to be influenced by SES and related factors. While several short forms of the MC have been developed and researched, some have been presented based on single studies and lack replication. Form C (Reynolds, 1982) presented in this study stands up as a solid instrument for clinical and forensic use, although Form X1 (Strahan & Gerbasi, 1972) is advocated as another good choice by others (Fischer & Fick, 1993; Fraboni & Cooper, 1989). Either one provides a reliable, useful instrument that could fit easily into most assessments and offers the examiner insight into the response style of the individual being assessed. Finally, with these data, it can be seen that scores on the MC in a forensic situation are expected to be higher than in a non-forensic context. Examiners using these norms have the option of comparing an individual s score to means from both groups to aid in understanding to what degree a person may be attempting to alter self-presentation. A score that is high for the non-forensic group but average in a forensic context might be interpreted differently (more deceptive than general population but as expected for the assessment context) than one that is elevated for both groups (unexpectedly high even in the context of the evaluation). In addition, a score lower than average should alert the examiner to the possibility of a negative, self-critical response set. This test needs to be studied for its application in unusual circumstances. It has been the experience of one of the authors that very high scores on the MC have been helpful at times in identifying deceptive self-presentation in individuals who were psychologically savvy enough (or who may have been coached on instruments) so that they produced acceptable scores on validity scales of major personality tests but succumbed to the lure of righteous self-presentation this instrument allows and revealed a strongly biased approach. Persons producing such results had MC scores of 29, produced valid profiles on the MMPI-2 and/or the Child-Abuse-Potential Inventory, and had strong evidence against them for child abuse, which they denied. It is not clear what aspects of this instrument led the subjects to produce such extreme scores when they had passed other validity measures. Another area for further research involves the personality dynamics associated with an extremely low score on the MC. Again, clinical experience has found that scores in the

9 Marlowe Crowne and Short Form C: Forensic Norms 491 range of 12 in a forensic setting generally coincide with an individual with low emotional resiliency, a very negative self-evaluation, and a tendency to exaggerate reported symptoms. Often very low scores have been associated with extremely high scores on the F scale of the MMPI-2; and, as noted above, Milner found that low scores on the MC identified Fake Bad profiles on the Child Abuse Potential Inventory. A very low score on the MC should at least caution the examiner to explore the possibility that the client is feeling overwhelmed and defenseless and may be overly critical of self. The Marlowe Crowne Social Desirability Scale is a clinically useful instrument that, along with its short forms, offers examiners a measure of a subject s tendency to present self in a socially desirable light. It correlates moderately well with the MMPI/ MMPI-2 L scale (Crowe & Marlowe, 1960), but is easier to administer and score and less expensive than these inventories. The short forms are even more attractive as brief instruments that can be added to a test battery or stand alone in a screening context. Having multiple measures of deceptive tendencies, especially if one of them is brief and inexpensive as well as reliable and valid, is helpful. In limited evaluations in which instruments such as the MMPI-2 or other long inventories are not used, the MC or MC-C offers a very attractive alternate way to get information similar to what is obtained through the validity scales of the longer instruments. References Ballard, R. (1992). Short forms of the Marlowe Crowne Social Desirability Scale. Psychological Reports, 71, Bathurst, K., Gottfried, A.W., & Gottfried, A.E. (1997). Normative data for the MMPI-2 in child custody litigation. Psychological Assessment, 9, Boyd, A., & Meyer, R. (in preparation). The assessment of deception. Boston, MA: Allyn & Bacon. Crowne, D.P., & Marlowe, D. (1960). A new scale of social desirability independent of psychopathology. Journal of Consulting Psychology, 24, Dalton, J.E. (1994). MMPI-168 and Marlowe Crowne profiles of adoptive applicants. Journal of Clinical Psychology, 50, Deshields, T.L., Tait, R.C., Gfeller, J.D., & Chibnall, J.T. (1995). Relationship between social desirability and self-report in chronic pain patients. Clinical Journal of Pain, 11, Abstract retrieved from PsychINFO database, Item Edens, J.F., Buffington, J.K., Tomicic, T.L., & Riley, B.D. (2001). Effects of positive impression management on the Psychopathic Personality Inventory. Law and Human Behavior, 25, Fischer, D.G., & Fick, C. (1993). Measuring social desirability: Short forms of the Marlowe Crowne Social Desirability Scale. Educational and Psychological Measurement, 53, Fraboni, M., & Cooper, D. (1989). Further validation of three short forms of the Marlowe Crowne scale of social desirability. Psychological Reports, 65, Heilbrun, K. (1995). Child custody evaluation: Critically assessing mental health experts and psychological tests. Family Law Quarterly, 29, Helmers, K.F., Krantz, D.S., Merz, C.N.B., Klein, J., et al. (1995). Defensive hostility: Relationship to multiple markers of cardiac ischemia in patients with coronary disease [On-line]. Health Psychology, 14, Abstract from: PsychINFO Database Item: Loo, R., & Thorpe, K. (2000). Confirmatory factor analyses of the full and short versions of the Marlowe Crowne Social Desirability Scale. Journal of Social Psychology, 140, Mann, S.J., & James, G.D. (1998). Defensiveness and essential hypertension. Journal of Psychomatic Research, 45, Abstract retrieved from PsychINFO database, Item

10 492 Journal of Clinical Psychology, April 2003 McCrae, R.R., & Costa, P.T., Jr. (1983). Social desirability scales: More substance than style. Journal of Consulting and Clinical Psychology, 51, Miller, H.A., Lape-Brinkman, L., Roche, D., Cox, M., & Motheral, L. (1999, August). Child custody evaluations: Normative date for the MMPI-2 and the Rorschach. Poster session presented at the annual meeting of the American Psychological Association, Boston, MA. Milner, J.S. (1986). The Child Abuse Potential Inventory: Manual (2nd ed). Dekalb, IL: Psytech. Morey, L.C. (1991). The Personality Assessment Inventory professional manual. Odessa, FL: Psychological Assessment Resources. Reynolds, W.M. (1982). Development of reliable and valid short forms of the Marlowe Crowne Social Desirability Scale. Journal of Clinical Psychology, 38, Robinette, R.L. (1991). The relationship between the Marlowe Crowne Form C and the validity scales of the MMPI. Journal of Clinical Psychology, 47, Rosenthal, R. (1984). Meta-analytic procedures for social research. Beverly Hills, CA: Sage. Strahan, R., & Gerbasi, K.C. (1972). Short, homogeneous versions of the Marlowe Crowne Social Desirability Scale. Journal of Clinical Psychology, 28, Vella-Brodrick, D.A., & White, V. (1997). Response set of social desirability in relation to the Mental, Physical and Spiritual Well-Being Scale. Psychological Reports, 81, Zakzanis, K.K. (2001). Statistics to tell the truth, the whole truth, and nothing but the truth: Formulae, illustrative numerical examples, and heuristic interpretation of effect size analysis for neuropsychological researchers. Archives of Clinical Neuropsychology, 16, Zook, A., & Sipps, G.J. (1985). Cross-validation of a short form of the Marlowe Crowne Social Desirability Scale. Journal of Clinical Psychology, 41,

11 本文献由 学霸图书馆 - 文献云下载 收集自网络, 仅供学习交流使用 学霸图书馆 ( 是一个 整合众多图书馆数据库资源, 提供一站式文献检索和下载服务 的 24 小时在线不限 IP 图书馆 图书馆致力于便利 促进学习与科研, 提供最强文献下载服务 图书馆导航 : 图书馆首页文献云下载图书馆入口外文数据库大全疑难文献辅助工具