Available online at www.sciencedirect.

com Procedia
Social and
Behavioral
Procedia
Procedia - Social
- Social and Behavioral
and Behavioral Sciences
Sciences 00 (2011)
33 (2012) 707 –000–000
711 Sciences
www.elsevier.com/locate/procedia

PSIWORLD 2011

Psychometric properties of the Marlowe-Crowne Social

Desirability Scale in a Romanian sample
Paul Sârbescuª*, Iuliana Costeaª, Silvia Rusuª
ªWest University, Bld. V. Pârvan nr. 4, Timiúoara, 300233, România

Abstract

This research investigated the psychometric properties and the reliability of the Romanian version of the Marlowe-
Crowne Social Desirability Scale (MCSDS; Crowne & Marlowe, 1960), as well as the fit of two short versions of the
scale. The non-experimental research was performed on a sample of 215 participants, aged between 20 and 35 years,
using the MCSDS. Along with the original 33 item scale, a 21 item version and a short 13 item version were tested
using Confirmatory Factor Analysis (CFA). The results showed that both short versions achieved better fit than the
original scale, the 13 item one being the most adequate. All three versions proved satisfactory reliability. This
research supports the use of the MCSDS short forms in future research requiring the measurement of social
desirability.
©
© 2012 PublishedbybyElsevier
2011 Published Elsevier B.V.
Ltd. Selection
Selection and/or peer-review
and peer-review under responsibility
under responsibility of PSIWORLD2011
of PSIWORLD 2011

Keywords: Social Desirability, Marlowe-Crowne Social Desirability Scale, psychometric properties, reliability.

1. Introduction

Socially desirable responding is the tendency of individuals to make themselves look good according
to current cultural norms (Mick, 1996). People who score high on this trait tend to present a favorable
image of themselves by not giving an honest answer, especially on controversial or sensitive issues (e.g.
race attitudes). An important distinction has been made between two factors of social desirable
responding: self-deception (when the respondent believes his positive self-reports) and impression
management (when the respondent intentionally provides a more positive image of himself) (Paulhus,
1984).

*
Corresponding author. Tel.: +40 726 863 806.
E-mail address: paul.sarbescu@gmail.com.

1877-0428 © 2012 Published by Elsevier B.V. Selection and/or peer-review under responsibility of PSIWORLD2011
doi:10.1016/j.sbspro.2012.01.213
708 P. Paul Sârbescu
Sârbescu et al.et/ Procedia
al. / Procedia - Social
- Social and Behavioral
and Behavioral Sciences
Sciences +40 33
726(2012) 707 – 711
863 806

But, either we talk about impression management, or self-deception, social desirable responding can
affect the validity of our research in domains like nursing and health (Neeley and Cronley, 2004),
consumer behavior (Mick, 1996) or personnel selection (Furnham, 1990). The review made by King and
Bruner (2000) highlights that social desirability bias plays an important role in suppressing or obscuring
relationships among variables, as well as in producing artificial relationships among independent and
dependent variables. Thus, it has been suggested that a measure of social desirability should be applied
whenever the research situation requires it (van de Mortel, 2008; Sullman & Taylor, 2010).
One of the most used instruments for measuring social desirability is the Marlowe-Crowne Social
Desirability Scale (MCSDS; Crowne & Marlowe, 1960). The authors built this instrument by selecting
specific items from some of the most used personality inventories. In order for an item to be included in
the Social Desirability Scale, it had to meet two criteria: (1) to express a behavior that is culturally
sanctioned or approved, (2) to have minimal pathological or abnormal implications. Thus, the scale
contains items on behaviors that are subject of cultural approval and are somehow improbable of
occurrence, but have the minimum of maladjustment implied (Crowne & Marlowe, 1960).
Since 1960, there have been a lot of studies that analyzed the validity of the MCSDS. A meta-analysis
(Beretvas, Meyer & Leite, 2002) found that the reliability of the MCSDS is higher for women (.79) than
for men (.70), and had an unsatisfactory level for adolescents (.53).
Over the years, several short forms of MCSDS have been developed in order to shorten the time
needed to administer the instrument, and because some data show that some of the items contribute
relatively little to the overall scale (Strahan & Carrese Gerbasi, 1972). Research studying the validity of
the short forms of MCSDS have shown contradicting results. Some of the research suggests that the short
forms can be used with as much success as the original version (Reynolds, 1982; Silverstein, 1983; Ii &
Sipps, 1985; Andrews & Meyer, 2003; Loo & Thorpe, 2000), while other studies did not find consistent
model fit for short forms (Barger, 2002). When it comes to cross-cultural use of the short forms of the
MCSDS, a recent study by Verardi et al (2010) suggests that a specific short form of MCSDS (form C)
should not be used to assess social desirability in cross-cultural research and does not reach scalar
equivalence. The same study implies that the development of a social desirability scale valid across
cultures might be possible (Verardi et al, 2010).
In Romania, although the original scale has been used in research for several years, no study (to our
knowledge) verified its validity and reliability. This research investigated the psychometric properties and
the reliability of the Romanian version of the MCSDS, as well as the fit of two short versions of the scale.

2. Methodology

2.1. Participants

The total sample (n = 215) consisted of 195 males (90.7%) and 20 females (9.3%). The age of the
respondents ranged from 20 to 35 years (M = 26.63, SD = 4.20). The sample consisted of students, their
friends and relatives. All subjects volunteered to take part in the study.

2.2. Measures and procedures

The participants completed the MCSDS (Crowne & Marlowe, 1960), which contains 33 true-false
items. The internal consistency alpha showed satisfactory level (.78).
 PaulP.Sârbescu
Sârbescuetetal.
al./ /Procedia
Procedia- -Social
Socialand
andBehavioral
BehavioralSciences
Sciences33+40
(2012)
726 707
863 –806
711 709

3. Results

Along with the original 33 item scale, we decided to test two potentially viable models: a 21 item
version1 formed by removing items with loadings lower than .30, and a 13 item version 2 representing a
short form of the original scale. Table 1 presents means, standard deviations, skewness, kurtosis, and
Cronbach’s Į coefficients of the total sample:

Table 1. Means, standard deviations, skewness, kurtosis and Į coefficients of the MCSDS

Model M SD Skewness Kurtosis Cronbach’s Į

33 items 15.48 5.14 0.38 -0.10 .78
21 items 7.29 4.02 0.61 0.09 .79
13 items 4.61 2.90 0.72 0.09 .75

The skewness and kurtosis values indicate that distributions do not deviate substantially from
normality for any of the three models. The internal consistencies alphas ranged from .75 (for the 13 item
version) to .79 (for the 21 item version), thus showing satisfactory reliability for all three models. The 21
item version correlated .94 with the original scale, while the 13 item version correlated .87 with it.
Confirmatory factor analysis (CFA) was used to test the appropriateness of the three models. The
results of the CFA are presented in Table 2:

Table 2. Primary goodness of fit, comparative and parsimony indices for the three models of MCSDS

Model Ȥ² df GFI AGFI RMSEA NFI CFI PNFI PCFI

33 items 710.66** 495 .83 .81 .045 [.037 - .052] .42 .69 .39 .65
21 items 298.88** 189 .88 .86 .052 [.041 - .063] .59 .79 .54 .71
13 items 72.79 65 .95 .93 .024 [.000 - .049] .81 .94 .67 .81
Notes: ** p < .01

The original 33 item model obtained poor fit; although the RMSEA value was adequate, all the other
indices showed inadequate values. The 21 item model showed improved fit, with some indices reaching
acceptable values (GFI, AGFI), but others showing poor values (NFI, CFI). The 13 item model proved to
be the best fit, with all indices showing optimal or acceptable values, and with the parsimony indices
(PNFI, PCFI) reaching the highest values of all three models. Thus, the short form of the MCSDS seems
to be the most adequate (Fig. 1).

1
Items 2, 3, 4, 5, 6, 10, 11, 12, 14, 15, 19, 21, 22, 23, 26, 28, 29, 30, 31, 32, 33.
2
Items 4, 6, 11, 12, 14, 15, 19, 21, 23, 26, 28, 32, 33.
710 P. Paul Sârbescu
Sârbescu et al.et/ Procedia
al. / Procedia - Social
- Social and Behavioral
and Behavioral Sciences
Sciences +40 33
726(2012) 707 – 711
863 806

e4 s4

e6 s6 .38

e11 s11 .56

.49
e12 s12
.42
e14 s14 .40

e15 s15 .42

.50
e19 s19 SD
.33

e21 s21 .38

.36
e23 s23 .61
.42
e26 s26 .39

e28 s28

e32 s32

e33 s33

Fig. 1. Factor loadings resulting from the CFA on the 13 item model.

4. Discussion

This research investigated the psychometric properties and the reliability of the Romanian version of
the MCSDS, as well as the fit of two short versions of the scale. The results showed that both short
versions achieved better fit than the original scale, the 13 item one achieving optimal fit. These results are
similar to those found by other researches (Silverstein, 1983; Ii & Sipps, 1985; Andrews & Meyer, 2003),
which tested short forms of the MCSDS and found them to be more adequate than the original one.
Although the 21 item version proved a bit higher reliability than the original scale, while the 13 item
version showed a bit lower reliability, all three versions achieved satisfactory reliability. Also, the 21 item
version correlated strongly with the original scale than the 13 item version, but the magnitude for both
correlations was very high.
The present research has one major limitation: the overwhelming majority of male participants.
Although there doesn’t seem to exist major differences on social desirability between men and women
(Reynolds, 1982; Loo & Thorpe, 2000), it appears that reliability differences might exist (Beretvas et al.,
2002). Thus, future research could verify the reliability of the original MCSDS, as well as for the two
short forms presented, on more balanced samples, or even on a predominantly female sample.
In conclusion, this research shows that the Marlowe-Crowne Social Desirability Scale is a reliable
instrument in the Romanian culture. Also, it provides two short forms which seem to be more adequate
than the original scale. Any of this two short versions can be successfully applied to the Romanian
population in future research requiring the measurement of social desirability.

References

Andrews, P. & Meyers, R.G. (2003). Marlowe-Crowne Social Desirability Scale and Short Form C: Forensic Norms. Journal of
Clinical Psychology, 59(4), 483-492.
 Paul
P.Sârbescu
Sârbescuetetal.
al.//Procedia
Procedia--Social
Socialand
andBehavioral
BehavioralSciences
Sciences33 (2012)
+40 707 806
726 863 – 711 711

Barger, S. D. (2002). Journal of Personality The Marlowe-Crowne Affair: Short Forms, Psychometric Structure and Social
Desirability. Journal of Personality Assessment, 79(2), 286–305.
Beretvas, S. N., Meyers, J. L., & Leite, W. L. (2002). A Reliability Generalization Study of the Marlowe-Crowne Social Desirability
Scale. Educational and Psychological Measurement, 62(4), 570-589.
Crowne, D. P., & Marlowe, D. (1960). A new scale of social desirability independent of psychopathology. Journal of consulting
psychology, 24(4), 349-354.
Furnham, A. (1990). Faking personality questionnaires: Fabricating different profiles for different purposes. Current psychology,
9(1).
Ii, A. Z.& Sipps, G. J. (1985), Cross-validation of a short form of the Marlowe-Crowne Social Desirability Scale. Journal of Clinical
Psychology, 41: 236–238.
King, M. F., & Bruner, G. C. (2000). Social desirability bias: A neglected aspect of validity testing. Psychology and Marketing,
17(2), 79-103.
Loo, R. & Thorpe, K. (2000). Confirmatory factor analyses of the full and short versions of the Marlowe-Crowne Social Desirability
Scale. The Journal of Social Psychology, 140(5), 628-635.
Mick, D.G. (1996). Are studies of dark side variables confounded by socially desirable responding? The case of materialism. The
Journal of Consumer Research, 23(2), 106-119.
Neeley, S.M. & Cronley, M.L. (2004). When research participants don’t tell it like it is: pinpointing the effects of social desirability
bias using self vs. indirect-questioning. Advances in Consumer Research, 31.
Paulhus, D.L. (1984). Two component models of socially desirable responding. Journal of Personality and Social Psychology,
46(3), 598-609.
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.
Silverstein, A.B. (1983). Validity of random short forms: II . The Marlowe-Crowne Social Desirability Scale. Journal of Clinical
Psychology, 39(4), 582-585.
Sullman, M.J.M., & Taylor, J.E. (2010). Social desirability and self-reported driving avoidance: Should we be worried?
Transportation Research Part F, 13, 215-221.
Strahan, R., & Gerbasi, K. C. (1972). Short, homogeneous versions of the Marlow-Crowne Social Desirability Scale. Journal of
Clinical Psychology, 28(2), 191-193.
Van de Mortel, T.F. (2008). Faking it: social desirability response bias in self-report research. Australian Journal of advanced
nursing, 25(4), 40-48.
Verardi, S., Dahourou, D., Ah-Kion, J., Bhowon, U., Tseung, C. N., Amoussou-Yeye, D., Adjahouisso, M., et al. (2009).
Psychometric Properties of the Marlowe-Crowne Social Desirability Scale in Eight African Countries and Switzerland. Journal
of Cross-Cultural Psychology, 41(1), 19-34.