Social Desirability Scale (Marlowe-Crowne) (MCSD)

Abstract

The Marlowe-Crowne Social Desirability Scale (MCSD), developed by Douglas P. Crowne and David Marlowe in 1960, is one of the most widely utilized psychometric instruments in psychological, sociological, and behavioral science research. Designed to evaluate an individual’s propensity to engage in socially desirable responding (SDR)—the tendency to present oneself in an unrealistically favorable light in self-report evaluations—the instrument was intentionally engineered to measure need for social approval independently of severe psychopathology. Comprising 33 dichotomous items answered in a True/False format, the scale includes behaviors that are culturally sanctioned yet possess a low base-rate of true occurrence, alongside universally prevalent minor transgressions that are socially disapproved. Psychometrically, the original validation demonstrated high internal consistency, yielding a Kuder-Richardson formula 20 (KR-20) coefficient of .88 and a one-month test-retest reliability coefficient of .89. While historically conceptualized as a unidimensional measure of response distortion, contemporary factor-analytic and latent structural investigations substantiate a multidimensional framework, primarily distinguishing between the attribution of culturally sanctioned, improbable virtues and the denial of common human shortcomings. The MCSD remains an indispensable methodological control variable in survey methodologies, organizational assessments, and clinical research, while functioning substantively as an operationalization of the approval motive and self-defensive regulatory strategies.

Keywords

Marlowe-Crowne Social Desirability Scale, MCSD, social desirability bias, response distortion, impression management, self-deceptive enhancement, approval motive, psychometrics, survey methodology, construct validity, Kuder-Richardson 20, self-report bias, attribution, denial, factor analysis

Authors

The Marlowe-Crowne Social Desirability Scale was conceptualized and validated by Douglas P. Crowne and David Marlowe during their tenure at The Ohio State University in the late 1950s.

• Douglas P. Crowne, Ph.D.: Was a prominent clinical psychologist and psychometrician who served as Professor Emeritus of Psychology at the University of Waterloo in Ontario, Canada. His empirical work focused heavily on personality assessment, the mechanics of approval motivation, and social learning dynamics within clinical paradigms.
• David Marlowe, Ph.D.: Served as a leading behavioral research scientist and senior social psychologist, later holding prestigious research appointments including Chief of the Department of Military Psychiatry at the Walter Reed Army Institute of Research (WRAIR). His scholarly trajectory addressed organizational psychology, human performance under environmental and psychological stress, and interpersonal influence processes.

Their foundational collaboration yielded the seminal 1960 publication in the Journal of Consulting Psychology, followed by their definitive theoretical monograph, The Approval Motive: Studies in Evaluative Dependence (Crowne & Marlowe, 1964), which positioned the construct not merely as an artifactual response style, but as a central personality disposition characterized by evaluative dependency and conflict avoidance.

Purpose

The primary purpose of the Marlowe-Crowne Social Desirability Scale is to assess, quantify, and control for the confounding influence of response bias in self-report inventories. In behavioral and social science methodologies, participants often systematically misrepresent their true behavioral patterns, internal affects, and attitudes to align with perceived cultural norms, interpersonal expectations, or experimental demands. This systematically distorts observed covariance structures, inflates or attenuates effect sizes, and compromises both internal and external validity.

The immediate historical catalyst for the development of the MCSD was the fundamental limitation of existing social desirability indices in the 1950s, most notably Allen L. Edwards’ Social Desirability Scale (1957). Edwards derived his scale items largely from the Minnesota Multiphasic Personality Inventory (MMPI). Consequently, Edwards’ scale was heavily saturated with manifest psychopathology and emotional maladjustment; an individual scoring low on Edwards’ scale was often reporting severe psychiatric distress rather than low social conformity. Crowne and Marlowe recognized the imperative for a measurement tool that could detect the need to seek approval and avoid disapproval completely independent of psychiatric symptomatology and emotional pathology.

To realize this objective, the authors devised a set of items where the socially desirable response alternative carries an exceptionally low baseline probability of occurrence in regular human conduct (e.g., claiming to never dislike anyone, or always being willing to admit a mistake), while the socially undesirable alternative represents pervasive, normative human frailty with negligible psychiatric severity (e.g., gossiping occasionally, or feeling resentful when not getting one’s way). As a consequence, high scores on the MCSD reflect a defensive, unrealistically pristine self-portrayal.

In contemporary research and applied contexts, the MCSD serves two complementary methodological functions:

• Methodological Control and Screening: It is widely integrated as a covariate or statistical suppressor variable in multivariate modeling. Researchers utilize MCSD scores to detect participants who engage in systematic impression management, enabling statistical partialling of desirability variance or identifying suspect protocols for exclusion in clinical trials, organizational hiring screenings, and high-stakes forensic evaluations.
• Substantive Personality Operationalization: Beyond its role as a validity screen, the scale operationalizes the approval motive. Individuals exhibiting elevated MCSD scores systematically demonstrate greater behavioral conformity, persuasibility, social compliance, vulnerability to peer pressure, and avoidance of interpersonal confrontation in experimental paradigms.

Psychological Construct

The psychological construct evaluated by the MCSD is socially desirable responding (SDR), conceptualized fundamentally by Crowne and Marlowe as the overt reflection of an underlying need for social approval. Rather than viewing the scale strictly as an index of conscious mendacity or faking, the authors conceptualized high scores as an index of generalized evaluative dependency—a personality pattern wherein the individual relies upon defensive self-enhancement to protect a vulnerable self-esteem from perceived social disapproval.

Over decades of psychometric evolution, the construct of socially desirable responding has been theoretically partitioned into two core, interrelated domains (often termed the dual-component framework, formalized comprehensively by Delroy L. Paulhus):

• Impression Management (IM): A deliberate, conscious, and strategic attempt to tailor one’s responses to project a socially sanctioned persona to external audiences. This component represents tactical dissimulation, where the respondent intentionally conceals norm-violating attitudes to secure tangible social rewards or avoid interpersonal sanctions.
• Self-Deceptive Enhancement (SDE): An unconscious, intrapsychic propensity to hold genuinely favorable, overconfident, and idealized self-perceptions. In this mode, the individual is not actively lying to the investigator, but rather displays a pervasive cognitive filter that denies personal vulnerabilities, doubts, and aggressive impulses to preserve internal psychological equilibrium.

The Marlowe-Crowne scale captures elements of both dynamics, though psychometric analyses suggest it leans heavily toward impression management and regulatory avoidance of disapproval. At the item-content level, the construct manifests across two primary operational dimensions:

1. Attribution of Improbable Virtues

This operational dimension involves endorsing statements that attribute culturally revered, morally pristine behaviors to the self, despite the extreme statistical unlikelihood of their uninterrupted occurrence in routine daily life. For instance, endorsing statements such as “I never hesitate to go out of my way to help someone in trouble” or “I am always courteous, even to people who are disagreeable” requires the respondent to assert absolute altruism and faultless interpersonal grace without human exception.

2. Denial of Common Transgressions

Conversely, this dimension encompasses rejecting statements that describe universal, benign human failings, petty resentments, and ubiquitous behavioral lapses. Disconfirming statements such as “I like to gossip at times” or “There have been times when I felt like rebelling against people in authority even though I knew they were right” demonstrates an unwillingness to acknowledge standard psychological impulses, choosing instead to present an artificially flawless moral and psychological profile.

Theoretical Framework

The theoretical bedrock of the Marlowe-Crowne Social Desirability Scale is rooted in Julian Rotter’s Social Learning Theory, developed extensively at The Ohio State University during the era of Crowne and Marlowe’s collaboration. Rotter postulated that the potential for a given behavior to occur in a specific situation is a function of the individual’s expectancy that the behavior will lead to a particular reinforcement, multiplied by the subjective value of that reinforcement to the individual.

Crowne and Marlowe expanded this formula by postulating that the need for approval functions as a generalized secondary drive or motivational state. When individuals possess an exceptionally high need for approval combined with a low generalized expectancy of attaining social reinforcement through normal competence, they manifest evaluative defensiveness. In this framework, socially desirable responding on self-report questionnaires is not a meaningless testing nuisance; it is an active, learned, operant coping behavior designed to elicit external validation and eliminate the threat of rejection.

Key foundational assumptions of this framework include:

• Defensive Self-Protection: High MCSD scorers are primarily motivated by the avoidance of disapproval rather than merely the proactive pursuit of praise. Their test-taking behavior represents an avoidant strategy aimed at mitigating threat and social vulnerability.
• Cross-Situational Generalization: The propensity to respond in a socially desirable manner on a paper-and-pencil inventory generalizes to overt interpersonal situations. Individuals scoring high on the MCSD exhibit higher conformity in Asch-type perceptual conformity tasks, display greater susceptibility to verbal conditioning, show heightened compliance with authoritarian directives, and inhibit the expression of hostile or aggressive impulses in face-to-face interactions.
• Independence from Structural Pathology: In contrast to psychodynamic views that equate extensive repression or denial solely with neurosis, Crowne and Marlowe established that the approval motive operates independently of clinical psychiatric categorization. An individual can maintain high social defensiveness without displaying the severe functional impairment characteristic of formal psychiatric syndromes.

Validity

The Marlowe-Crowne Social Desirability Scale has undergone rigorous empirical validation across thousands of studies, establishing robust evidence for its construct, convergent, discriminant, and criterion-related validity.

Construct and Experimental Validity

In their original monograph and subsequent validation studies, Crowne and Marlowe (1964) demonstrated construct validity by linking high MCSD scores directly to observable behavioral tendencies in controlled laboratory settings. Specifically, high scorers exhibited:

• Significantly greater susceptibility to verbal conditioning paradigms, readily adopting experimenter-reinforced sentence completions.
• Higher compliance and persistence on monotonous, intrinsically unrewarding tasks (such as turning spools on a board) when such performance was framed as assisting the experimenter.
• Markedly greater conformity to peer group consensus when judging visual stimuli, mirroring findings from classic social influence literature.

Convergent Validity

The MCSD demonstrates moderate to high correlations with alternative measures of response distortion, impression management, and defensiveness. It correlates positively ($r = .55$ to $.75$) with the Lie (L) and Correction (K) validity scales of the Minnesota Multiphasic Personality Inventory (MMPI-2), as well as the Lie scale of the Eysenck Personality Questionnaire (EPQ). When evaluated against modern multi-component instruments, such as the Balanced Inventory of Desirable Responding (BIDR), the MCSD correlates substantially higher with the Impression Management (IM) dimension ($r \approx .65 – .73$) than with the Self-Deceptive Enhancement (SDE) subscale ($r \approx .30 – .42$), underscoring its pronounced sensitivity to interpersonal self-presentation.

Discriminant Validity

Crucially, Crowne and Marlowe achieved their goal of establishing discriminant validity against psychopathology. While Edwards’ Social Desirability Scale correlated exceptionally high with the Taylor Manifest Anxiety Scale ($r = -.84$) and various MMPI clinical scales, indicating profound confounding with psychiatric distress, the MCSD exhibited statistically negligible or modest correlations with these same psychiatric indices ($r = -.25$ to $-.35$ with trait anxiety; $r < .15$ with measures of cognitive ability and IQ). This empirical separation confirmed that the MCSD detects evaluative defensiveness rather than general psychological health or the absence of psychiatric distress.

Criterion and Predictive Validity

The scale possesses extensive predictive validity in applied settings. In organizational and high-stakes personnel selection, elevated MCSD scores predict systematic over-reporting of organizational citizenship behaviors and under-reporting of counterproductive work behaviors. In epidemiological and health research, high MCSD scores correlate with significant under-reporting of alcohol consumption, illicit substance use, sexual risk behaviors, and dietary non-compliance.

Reliability

The Marlowe-Crowne scale demonstrates sound and stable psychometric reliability across diverse populations, clinical cohorts, and cultural contexts.

Internal Consistency

In the seminal 1960 investigation involving a sample of 76 undergraduate students, Crowne and Marlowe calculated the internal consistency of the 33-item scale using the Kuder-Richardson formula 20 (KR-20), yielding an internal consistency estimate of .88. Subsequent large-scale investigations utilizing Cronbach’s alpha ($lpha$) in collegiate, community, and clinical cohorts have consistently documented reliability coefficients ranging from .73 to .88, indicating a satisfactory level of item homogeneity despite the broad behavioral coverage of the item inventory.

Test-Retest Reliability

Temporal stability was originally evaluated by re-administering the scale over a 28-day (one month) test-retest interval to a sub-sample of 31 participants. The resulting Pearson correlation coefficient was $r = .89$ ($p < .001$), demonstrating that the scale measures a stable, enduring personality disposition rather than a fluctuating emotional state. Longer-term test-retest studies over intervals of 6 to 12 months have continued to show high stability coefficients, generally ranging between $r = .72$ and $r = .82$.

Split-Half Reliability

Calculations utilizing the Spearman-Brown corrected split-half technique in early normative studies revealed reliability coefficients consistently exceeding .86, further confirming the equivalence of the scale halves and overall structural stability.

Factor Analysis

The latent structural properties of the MCSD have been the subject of extensive psychometric investigation, shifting from early assumptions of unidimensionality to more nuanced multidimensional models.

Exploratory Factor Analysis (EFA)

Crowne and Marlowe initially constructed the scale under the premise of a single, overarching factor: the general need for social approval. However, subsequent principal components and exploratory factor analyses (e.g., Millham, 1974; Ramanaiah, Schill, & Leung, 1977) demonstrated that the 33 items systematically load across two distinct, robust factors:

• Attribution (Enhancement of Virtues): Primarily composed of the 14 positively keyed (True) items. These items involve the active, self-aggrandizing claim of culturally valued behaviors that are exceedingly rare (e.g., Item 1: investigating all candidates before voting; Item 13: always being a good listener).
• Denial (Repudiation of Common Faults): Composed primarily of the 19 negatively keyed (False) items. These items involve the active disavowal of common, universally shared minor faults, impulses, and irritations (e.g., Item 11: liking to gossip; Item 6: feeling resentful when not getting one’s way).

Confirmatory Factor Analysis (CFA)

Modern confirmatory factor analyses comparing competing structural models have consistently verified that a two-factor correlated model (Attribution vs. Denial) or a bifactor model (consisting of a general social desirability factor along with specific Attribution and Denial group factors) provides a superior fit to empirical data compared to a strict unidimensional model.

Typical fit indices reported across contemporary CFA studies (such as Loo & Thorpe, 2000; Barger, 2002; Leite & Beretvas, 2005) indicate:

• Strict Unidimensional Model: Frequently exhibits borderline or inadequate fit metrics, characterized by Root Mean Square Error of Approximation (RMSEA) values between .065 and .085, and Comparative Fit Index (CFI) values generally below .85, reflecting significant item residual covariance driven by keying direction.
• Two-Factor Correlated Model: Yields significantly improved fit indices, with RMSEA values dropping between .038 and .049, CFI reaching .90 to .94, and Standardized Root Mean Square Residual (SRMR) remaining below .055. The latent correlation between the Attribution and Denial factors typically ranges between $r = .40$ and $r = .65$, demonstrating related yet distinct psychological dynamics.

These factor-analytic findings have led many psychometricians to recommend examining Attribution and Denial subscale scores separately, as denial of negative attributes often reflects defensive avoidance, whereas attribution of improbable virtues reflects active social self-promotion.

Instrument / Measurement Tool

The Marlowe-Crowne Social Desirability Scale is structured as an objective, self-report psychometric inventory designed for rapid group or individual administration.

• Test Type: Standardized self-report personality and response bias inventory.
• Administration Format: Paper-and-pencil questionnaire, computer-assisted self-interview (CASI), or web-based survey instrument.
• Target Population: Adolescents and adults (typically ages 16 and older), requiring an approximate 6th-grade reading level.
• Item Count: 33 dichotomous items.
• Administration Time: Approximately 8 to 12 minutes.
• Response Scale: Dichotomous (True / False).
• Scoring System:
  • Each item is scored dichotomously: 1 point is awarded for each response in the socially desirable direction; 0 points are awarded for responses indicating a departure from social desirability.
  • Items keyed True (14 items): 1, 2, 4, 7, 8, 13, 16, 17, 21, 24, 25, 29, 30, 33. Endorsement of “True” yields 1 point.
  • Items keyed False (19 items): 3, 5, 6, 9, 10, 11, 12, 14, 15, 18, 19, 20, 22, 23, 26, 27, 28, 31, 32. Endorsement of “False” yields 1 point.
  • Total Score Range: 0 to 33.
• Score Interpretation Guidelines:
  • Scores 0–8 (Extremely Low): Unusually candid, highly non-conforming, or self-critical. The respondent willingly acknowledges numerous behavioral deviations and minor moral lapses without concern for social evaluation. In clinical settings, very low scores may suggest active self-depreciation, depressive self-focus, or severe social alienation.
  • Scores 9–19 (Average / Normative): Typical balancing of self-presentation and honest self-appraisal. The respondent possesses normal awareness of social standards without engaging in excessive defensive dissimulation. Survey data from these respondents are generally considered methodologically valid and uncorrupted by major response bias.
  • Scores 20–33 (High to Extremely High): Strong tendency toward socially desirable responding. The respondent portrays an unrealistically virtuous self-image, characterized by high conformity, high need for social approval, and extensive defensive posturing. In assessment contexts, substantive self-reports (e.g., personality, psychopathology, sensitive behaviors) from these respondents should be interpreted with caution or statistically controlled for bias.
• Common Short Forms: Due to the length of the 33-item inventory, several psychometrically validated short versions have been developed for survey contexts, including the 10-item and 20-item forms by Strahan & Gerbasi (1972), and Reynolds’ (1982) popular 13-item (Form C) scale ($lpha \approx .76$).

Permissions & Fee and Test Year

The Marlowe-Crowne Social Desirability Scale was officially published in 1960 in the Journal of Consulting Psychology (Volume 24, Issue 4, pages 349–354). Under standard academic psychometric conventions and historical copyright frameworks, the full 33-item instrument is widely regarded as residing in the public domain for non-commercial educational, scientific, and academic research purposes. The scale items were published fully in the original 1960 article and in Crowne and Marlowe’s subsequent 1964 volume without proprietary licensing restrictions or per-administration testing fees.

Researchers and clinicians are generally free to administer, reproduce, and adapt the MCSD for empirical studies without seeking formal written permission, provided proper academic citation is given to Crowne and Marlowe (1960). However, commercial distribution, inclusion in fee-for-service testing platforms, or publication within commercial psychological test batteries remains subject to standard fair-use guidelines and applicable intellectual property rights governed by the American Psychological Association (APA).

References

• Barger, S. D. (2002). The Marlowe-Crowne Social Desirability Scale does not measure generalized defense: An item response theory analysis. Journal of Psychopathology and Behavioral Assessment, 24(4), 271–284. https://doi.org/10.1023/A:1020755106208
• Crowne, D. P., & Marlowe, D. (1960). A new scale of social desirability independent of psychopathology. Journal of Consulting Psychology, 24(4), 349–354. https://doi.org/10.1037/h0047358
• Crowne, D. P., & Marlowe, D. (1964). The Approval Motive: Studies in Evaluative Dependence. John Wiley & Sons.
• Edwards, A. L. (1957). The Social Desirability Variable in Personality Assessment and Research. Dryden Press.
• Leite, W. L., & Beretvas, S. N. (2005). Validation of scores on the Marlowe-Crowne Social Desirability Scale and the Balanced Inventory of Desirable Responding. Educational and Psychological Measurement, 65(1), 140–154. https://doi.org/10.1177/0013164404267285
• Loo, R., & Thorpe, K. (2000). Confirmatory factor analyses of the full and short-form versions of the Marlowe-Crowne Social Desirability Scale. The Journal of Social Psychology, 140(5), 628–635. https://doi.org/10.1080/00224540009600503
• Millham, J. (1974). Two components of need for approval: Response styles and defensiveness. Journal of Consulting and Clinical Psychology, 42(3), 378–385. https://doi.org/10.1037/h0036647
• Paulhus, D. L. (1984). Two-component models of socially desirable responding. Journal of Personality and Social Psychology, 46(3), 598–609. https://doi.org/10.1037/0022-3514.46.3.598
• Paulhus, D. L. (1991). Measurement and control of response bias. In J. P. Robinson, P. R. Shaver, & L. S. Wrightsman (Eds.), Measures of Personality and Social Psychological Attitudes (pp. 17–59). Academic Press. https://doi.org/10.1016/B978-0-12-590241-0.50006-X
• Ramanaiah, N. V., Schill, T., & Leung, L. S. (1977). A test of the hypothesis about the two-dimensional nature of the Marlowe-Crowne Social Desirability Scale. Journal of Research in Personality, 11(2), 251–259. https://doi.org/10.1016/0092-6566(77)90036-7
• Reynolds, W. M. (1982). Development of reliable and valid short forms of the Marlowe-Crowne Social Desirability Scale. Journal of Clinical Psychology, 38(1), 119–125. https://doi.org/10.1002/1097-4679(198201)38:1<119::AID-JCLP2270380118>3.0.CO;2-I
• Strahan, R., & Gerbasi, K. C. (1972). Short, homogeneous versions of the Marlow-Crowne Social Desirability Scale. Journal of Clinical Psychology, 28(2), 191–193. https://doi.org/10.1002/1097-4679(197204)28:2<191::AID-JCLP2270280220>3.0.CO;2-G

Items of the Scale