Social Desirability in Spouse Ratings

Abstract

Whether or not socially desirable responding is a cause for concern in personality assessment has long been debated. For many researchers, McCrae and Costa laid the issue to rest when they showed that correcting for socially desirable responding in self-reports did not improve the agreement with spouse ratings on the Neuroticism, Extraversion, and Openness to Experience Personality Inventory. However, their findings rest on the assumption that observer ratings in general, and spouse ratings in particular, are an unbiased external criterion. If spouse ratings are also susceptible to socially desirable responding, correcting for the bias in self-rated measures cannot be assumed to increase agreement between self-reports and spouse ratings, and thus failure to do so should not be taken as evidence for the ineffectiveness of measuring and correcting for socially desirable responding. In the present study, McCrae and Costa’s influential study was replicated with the exception of measuring socially desirable responding with the Marlowe–Crowne Social Desirability Scale, in both self-reports and spouse ratings. Analyses were based on responses from 70 couples who had lived together for at least one year. The results showed that both self-reports and spouse ratings are susceptible to socially desirable responding and thus McCrae and Costa’s conclusion is drawn into question.

Full text

Article
Social Desirability
in Spouse Ratings
Vaka Ve
´steinsdo
´ttir and
Eva D. Steingrimsdottir
Department of Psychology, University of Iceland,
Iceland
Adam Joinson
School of Management, University of Bath, UK
Ulf-Dietrich Reips
Department of Psychology, University of Konstanz, Germany
Fanney Thorsdottir
Department of Psychology, University of Iceland, Iceland
Abstract
Whether or not socially desirable responding is a cause for concern in personality
assessment has long been debated. For many researchers, McCrae and Costa laid the
issue to rest when they showed that correcting for socially desirable responding in
self-reports did not improve the agreement with spouse ratings on the Neuroticism,
Extraversion, and Openness to Experience Personality Inventory. However, their
findings rest on the assumption that observer ratings in general, and spouse ratings
in particular, are an unbiased external criterion. If spouse ratings are also susceptible
to socially desirable responding, correcting for the bias in self-rated measures cannot
be assumed to increase agreement between self-reports and spouse ratings, and thus
failure to do so should not be taken as evidence for the ineffectiveness of measuring
and correcting for socially desirable responding. In the present study, McCrae and
Costa’s influential study was replicated with the exception of measuring socially
desirable responding with the Marlowe–Crowne Social Desirability Scale, in both
self-reports and spouse ratings. Analyses were based on responses from 70 couples
who had lived together for at least one year. The results showed that both self-
reports and spouse ratings are susceptible to socially desirable responding and thus
McCrae and Costa’s conclusion is drawn into question.
Psychological Reports
0(0) 1–16
!The Author(s) 2018
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0033294118767815
journals.sagepub.com/home/prx
Corresponding Author:
Vaka Ve
´steinsdo
´ttir, Haskoli Islands, Aragata 14, Reykjavik 101, Iceland.
Email: vakav@hi.is

Keywords
Socially desirable responding, personality, response bias, spouse ratings, content
overlap
Introduction
Socially desirable responding (SDR) generally refers to ‘‘the tendency to choose
items that reflect socially approved behaviors’’ (Nunnally & Bernstein, 1994,
p. 382). It is a socially shared belief of what will be approved of by others
(e.g., Crowne & Marlowe, 1960) that is used to guide responding. SDR can
therefore inflate scores on desirable items, and deflate scores on undesirable
items. One approach to measuring SDR has been through the use of individually
administered scales designed to capture this tendency (see Paulhus, 1991).
When measured with such scales, SDR is seen as an individual difference
variable—that is, respondents are presumed to have varying levels of the ten-
dency to give socially desirable responses.
Researchers have raised concerns about SDR, in particular its nature and its
effects, since the 1930s (e.g., Meehl & Hathaway, 1946; Rosenzweig, 1933;
Vernon, 1934; Wiggins, 1968). For instance, there has been considerable debate
on whether SDR is a cause for concern in personality assessment (e.g., Holden,
2007; Nevid, 1983; Rorer, 1965). SDR has traditionally been regarded as a
response bias that poses a threat to the validity of personality scales (Cronbach,
1946; Edwards, 1953, 1957; Meehl & Hathaway, 1946). Under this assumption, an
observed correlation between a personality scale and a scale designed to measure
SDR indicates that the personality measures are confounded with error variance
(Paulhus, 1991). Those who argue that SDR scales measure a substantive trait
have challenged this interpretation and maintained that any observed correlation
is the result of content overlap between the instruments. Taking this view, correl-
ations between SDR scale scores and personality measures should thus not be
regarded as indicators of response bias (Furnham, 1986; McCrae & Costa, 1983;
Nicholson & Hogan, 1990; Smith & Ellingson, 2002).
More specifically, the debate centers on whether SDR scales measure a
response style or a substantive personality trait, and whether SDR variance
should be removed from personality scales. Because, in self-report studies, sub-
stance cannot be disentangled from response style, one strategy to resolve the
debate is to compare self-reports with external criteria (Furnham, 1986), such as
spouse ratings of a targets’ personality. If SDR scales are indicators of a
response style, then correcting for SDR should improve the agreement between
self-reports and an external criterion, showing that SDR acted as a suppressor
variable (e.g., Ganster, Hennessey, & Luthans, 1983).
In an influential study, McCrae and Costa (1983) used spouse ratings as an
external criterion of personality traits in the domains of Neuroticism,
2Psychological Reports 0(0)

Extraversion, and Openness to Experience (NEO). Correcting for SDR in
self-reports with the Marlowe–Crowne Social Desirability Scale (MCSDS;
Crowne & Marlowe, 1960) did not improve the agreement with spouse ratings,
and for many traits lowered the agreement. Therefore, the authors concluded
that SDR scales should be given substantive rather than artifactual
interpretations.
Because correcting for SDR failed to improve agreement between self-reports
and spouse ratings in the study of McCrae and Costa (1983), and numerous others
who used observer ratings as external criteria (Borkenau & Ostendorf, 1992;
Dicken, 1963; Kozma & Stones, 1987; Kurtz, Tarquini, & Iobst, 2008; McCrae
et al., 1989; Pauls & Stemmler, 2003; Piedmont, McCrae, Riemann, & Angleitner,
2000), some researchers have concluded that SDR scales are not useful for enhan-
cing the validity of personality measurements (e.g., Borkenau & Ostendorf, 1992)
or even that researchers should not make an effort to detect SDR in personality
measurements at all (e.g., Piedmont et al., 2000). Costa and McCrae (1997) inten-
tionally exclude validity scales to identify SDR to the Revised NEO Personality
Inventory (NEO-PI-R), claiming that they remain unconvinced of the utility of
such scales, and, in support of that opinion, cite their own 1983 study among
other studies using observer ratings. As Ones, Viswesvaran, and Reiss (1996) note,
for many researchers the SDR issue is a ‘‘methodological dead horse’’ (Nevid,
1983, p. 139), laid to rest by McCrae and Costa (1983).
These conclusions are an overgeneralization from McCrae and Costa’s (1983)
results and it should be noted that McCrae and Costa did not rule out the possi-
bility of SDR. They acknowledged that their respondents were disinterested vol-
unteers who answered anonymously and were not motivated to distort their
answers. Therefore, SDR might still be a problem in situations where motivation
to distort is high, for example, in personnel assessments (as has been demon-
strated; e.g., Rosse, Stecher, Miller, & Levin, 1998). In addition, a number of
other studies have shown that SDR degrades the validity of personality measures
(e.g., Ellingson, Sackett, & Hough, 1999; Holden, 2007; Viswesvaran & Ones,
1999). Topping and O’Gorman (1997), for example, reported lowered self-other
agreement on the NEO Five Factor Inventory scales under instructions to ‘‘fake
good’’ in the self-reports. Furthermore, SDR scales can detect faking (Holden,
2007; Lambert, Arbuckle, & Holden, 2016).
Previous objections notwithstanding, McCrae and Costa’s (1983) findings rest
on the assumption that observer ratings in general, and spouse ratings in par-
ticular, are an unbiased external criterion. More specifically, McCrae and
Costa note that although both self-reports and spouse ratings of personality
may contain social desirability (abbreviated SD in their discussion) ...
There is no reason to suspect that the SD of the subject would influence his or her
spouse’s ratings. The only variance common to the two sources would be real trait
variance, and this alone could account for the correlation between them. (p. 884)
Ve
´steinsdo
´ttir et al. 3

This assumption is questionable because if both partners are responding in a
socially desirable manner, their responses are bound to have more in common
than if one of them did and not the other, and therefore the common variance of
thetwosourcesmaynotjustberealtraitvariance.Itseemsintuitivethatifspouses
have a tendency to respond in a socially desirable manner, they will see it desirable
to have a partner with socially desirable qualities and rate them accordingly.
For instance, Funder and Colvin (1997, p. 625) suggest that the ‘‘self’’-
enhancement bias may be poorly named because enhancement effects have
also been found when comparing acquaintance ratings to those made by stran-
gers. These findings may be explained by the Self-Evaluation Maintenance
Model, which assumes that peoples’ self-evaluations are partly determined by
a process whereby the good qualities of close others are perceived as reflecting
something about oneself (Tesser, Pilkington, & McIntosh, 1989).
There are further reasons to believe that spouse ratings may be influenced by
SDR. SDR is based on shared beliefs about what will be approved of by others
(e.g., Crowne & Marlowe, 1960). Responding in a socially desirable manner is
thus responding in accordance with social norms, or more specifically injunctive
norms, which specify what people ought to do, that is, what will be approved or
disapproved of by others, and the anticipated social sanctions for not acting in
accordance with the norm (Cialdini, Reno, & Kallgren, 1990; Reno, Cialdini,
& Kallgren, 1993). Hence, social desirability is very different from personal
preferences because the direction of the resulting bias (the bias produced by
presenting an image that will be viewed favorable by others) is predicable
through the knowledge of injunctive norms. In fact, peoples’ judgments of
how desirable or undesirable a scale item is are highly consistent, even when
obtained from different groups of people (Edwards, 1957).
Because of these shared beliefs, SDR produces similarity between measure-
ments contaminated with SDR, not only when produced by the same individual
but also between individuals. McCrae and Costa (1983) express the same logic in
their discussion of the expected consequences of SDR, but for some reason they
implicitly take it to apply only to measurements obtained from the same person:
‘‘ ...such systematic distortion would be seen in the correlations between scales:
Measures of traits that are socially desirable may covary, not because they are
substantively related but because they are both susceptible to the operation of
SD biases’’ (p. 883). The social aspect of SDR is, however, what makes SDR
function in the same way for people in general. It follows that, if both
self-reports and spouse ratings are contaminated with SDR, the correlation
between the two would be strengthened, and thus ‘‘partialing out the variance
in self-reports that is due to SD’’ (p. 884), as was done by McCrae and Costa,
could not be expected to increase the correlation. Therefore, the failure of this
method to increase the correlation between self-reports and spouse ratings
cannot be taken as evidence for the ineffectiveness of measuring and correcting
for SDR.
4Psychological Reports 0(0)

Because of the great influence McCrae and Costa’s (1983) study had, and still
has, on the interpretation of SDR, the purpose of the present study is to replicate
it using the NEO-PI-R, with the exceptions of obtaining measures of spouses’
SDR, to test McCrae and Costa’s assumption that spouse ratings of personality
are an unbiased external criterion of their partners’ personality. A measure of
spouses’ SDR cannot be said to share content with a rating of their partner’s
personality as these are not measures of the same individual. Thus, if spouses’
SDR correlates with spouse ratings of personality this can be taken as an indi-
cation of response distortion.
Furthermore, because SDR is a social construct, based on shared beliefs of
what is desirable, spouses’ SDR is hypothesized to correlate with spouse ratings
of personality in a similar way as subjects’ SDR correlates with self-reports of
personality. McCrae and Costa (1983) used the Neuroticism–Extraversion–
Openness (NEO) Inventory as a measure of personality. In their study, the
correlation of these three traits with the measure of SDR (MCSDS) were:
r¼.49 (p<.001) with Neuroticism (N), r¼.15 (p<.05) with Extraversion
(E) and r¼.13 (insignificant) with Openness (O). A similar pattern as observed
by McCrae and Costa is expected in the present study for both partners.
N is an undesirable trait and thus a significant negative correlation is expected.
E is a somewhat desirable trait, though introversion (the other end of the con-
tinuum) is not necessarily undesirable. This might explain why some studies have
not found an association between SDR and E (see, e.g., Ones et al., 1996 for a
meta-analysis of the association between personality and SDR), whereas McCrae
and Costa (1983) and other more recent studies have found a mild positive cor-
relation between SDR (MCSDS) and E (e.g., Holden & Passey, 2010; Kurtz et al.,
2008). It is therefore not entirely clear whether to expect SDR to be uncorrelated,
or have a mild positive correlation, with E, although the latter is more likely since
we intend to use the same measures of personality and SDR as McCrae and Costa.
O, however, is neither desirable nor undesirable as reversed items have a reference
to rational thinking, which cannot be conceived of as undesirable, and thus there is
no reason to expect a correlation between O and the MCSDS. In addition, because
McCrae and Costa (1983) suggest that ‘‘individuals high in SD will appear to score
higher on measures of adjustment, conscientiousness, agreeableness and other
socially desirable traits than they actually are’’ (p. 883), the traits
Conscientiousness (C) and Agreeableness (A) were added in the present study,
with the expectation of a positive correlation to the MCSDS for both partners.
Method
Participants
A convenience sample of 70 heterosexual couples that had lived together for
at least one year participated in the study. Couples were recruited by asking
Ve
´steinsdo
´ttir et al. 5

people (most of whom worked at the same firm) to participate by giving a self-
evaluation of their personality and asking their partner to rate them as well.
An approximately equal number of men and women were recruited for par-
ticipation. The recruitment process was non-random and focused on obtain-
ing a relatively equal number of men and women for completion of self-
reports, and thus observer reports. Those who agreed to participate were
given verbal instructions to give a self-report of their own personality and
ask their partner to fill out a similar questionnaire containing an observer
report of personality. Two of the initially recruited 72 couples were excluded
after initial screening of participants, as they did not meet the requirements of
the study (returning a blank spouse-rating questionnaire and neither partner
giving a valid response to the length of cohabitation question). Self-reports
were obtained from 34 men (48.6%) and 36 women (51.4%), with correspond-
ing spouse ratings from 34 women and 36 men. Participants’ age ranged from
21 to 59 (M¼33, SD ¼9.7). All participants were volunteers and did not
receive compensation for study participation.
The average length of cohabitation was nine years and nine months (ranging
from 1 year and two months to 39 and a half years). There was, however, some
inconsistency in the reported length of cohabitation between some of the par-
ticipants and their partners. In three cases, only one of the partners gave a
response to the question and in 14 cases, the responses from the two partners
did not match (and were thus averaged for the above calculation of average
length of cohabitation). In 12 of those 14 cases, the discrepancy can be con-
sidered relatively minor (four months on average). For the remaining two
couples, the difference in reporting was three and five years (both couples
were however married, and both reported the exact same number of years
married). A little under half of the participating couples in the study were
married (41.1%) and over two-thirds had one or more children in their care
(70%).
Measures
Personality measure. The NEO-PI-R inventory is a 240-item measure of the Big
Five personality traits: Neuroticism, Extraversion, Openness to Experience,
Agreeableness, and Conscientiousness. Each dimension has six facets consisting
of eight items rated on a five-point scale ranging from ‘‘strongly disagree’’ to
‘‘strongly agree’’ (Costa and McCrae, 1992). The Icelandic version of the self-
report form (Form S) and the observer rating form (Form R) (Jo
´nsson &
Bergo
´rsson, 2004) were used.
Neuroticism (N) contains 48 items that measure the predisposition to experi-
ence psychological distress. For the N domain, reported internal consistency of
the Icelandic self-report form is 0.91 (Jo
´nsson & Bergo
´rsson, 2004), which is
almost the same as for the original version (0.92; Costa & McCrae, 1992).
6Psychological Reports 0(0)

Internal consistency of the Icelandic observer rating form (0.90) is also almost
the same as for the original version (0.91; McCrae & Terracciano, 2005). In the
present study, Cronbach’s alpha values for N were 0.93 (Form S) and 0.95
(Form R).
Extraversion (E) contains 48 items used to measure, for example, friendliness
and preference for the company of others. For the E domain, reported inter-
nal consistency of the Icelandic self-report form is 0.88 (Jo
´nsson &
Bergo
´rsson, 2004), almost the same as the original version (0.89; Costa &
McCrae, 1992). Internal consistency of the Icelandic observer rating form is
equal to the original version (0.91; McCrae & Terracciano, 2005). In the
present study, Cronbach’s alpha values for E were 0.92 (Form S) and 0.93
(Form R).
Openness to Experience (O) contains 48 items used to measure, for example,
aesthetic sensitivity and intellectual curiosity. For the O domain, reported
internal consistency of the Icelandic self-report form is 0.87 (Jo
´nsson &
Bergo
´rsson, 2004), which is equal to the original version (Costa &
McCrae, 1992). Internal consistency of the Icelandic observer rating form
(0.88) is also equal to the original version (McCrae & Terracciano, 2005). In
the present study, Cronbach’s alpha values for O were 0.81 (Form S) and 0.86
(Form R).
Agreeableness (A) contains 48 items used to measure, for example, helpfulness
and cooperation with others. For the A domain, reported internal consistency of
the Icelandic self-report form is 0.82 (Jo
´nsson & Bergo
´rsson, 2004), which is
similar to the original version (0.86; Costa & McCrae, 1992). Internal consist-
ency of the Icelandic observer rating form (0.92) is almost equal to the original
version (0.93; McCrae & Terracciano, 2005). In the present study, Cronbach’s
alpha values for A were 0.85 (Form S) and 0.90 (Form R).
Conscientiousness (C) contains 48 items used to measure, for example, orga-
nizing and carrying out tasks. For the C domain, reported internal consistency
of the Icelandic self-report form is 0.88 (Jo
´nsson & Bergo
´rsson, 2004), which is
similar to the original version (0.90; Costa & McCrae, 1992). Internal consist-
ency of the Icelandic observer rating form (0.95) is almost equal to the original
version (0.94; McCrae & Terracciano, 2005). In the present study, Cronbach’s
alpha values for C were 0.91 (Form S) and 0.94 (Form R).
Social desirability measure. An Icelandic translation of the MCSDS was used
as a measure of SDR (Ve
´steinsdo
´ttir, Reips, Joinson, & Thorsdottir, 2012).
The MCSDS is a true–false questionnaire consisting of 33 statements of either
uncommon but socially desirable behavior or common but undesirable behavior
(Crowne & Marlowe, 1960). All participants (both subjects and spouses) were
instructed to rate themselves on the MCSDS. In the present study, Cronbach’s
alpha values were 0.77 and 0.75 for the subjects’ self-reports and for the spouses’
self-reports, respectively.
Ve
´steinsdo
´ttir et al. 7

Procedure
Each couple received an envelope containing the self-report (Form S) and obser-
ver rating form (Form R) of the NEO-PI-R inventory. Attached to both Form S
and Form R was the MCSDS, which both partners were specifically asked to
give self-reports on. Participants filled out the questionnaires in their own homes
under the instructions to answer one of the two forms independently of their
spouses. Upon completion, participants returned the questionnaires in an enve-
lope labeled only with identification numbers to ensure complete anonymity.
Results and Discussion
Descriptive statistics for scales used in this study are shown in Table 1.
The mean scores and standard deviations (SD) of the MCSDS were in line
with those reported in previous studies (for an overview of descriptive statistics
from previous studies, see Ve
´steinsdo
´ttir, Reips, Joinson, & Thorsdottir, 2015).
No difference was observed between mean MCSDS scores for subjects and
spouses (t(130) ¼.056, p¼.95) and no gender differences were observed. For
the male subjects (n¼31), the mean score (M¼15.7, SD ¼6.0) on the
MCSDS did not differ from female subjects’ (n¼34) mean score (M¼16.5,
SD ¼4.1), t(52.52) ¼.608, p¼.55. The male spouses’ (n¼34) mean score
(M¼16.5, SD ¼5.7) also did not differ from the female spouses’ (n¼33)
mean score (M¼15.7, SD ¼4.2), t(60.52) ¼.613, p¼.54.
Table 1. Descriptive statistics for the five dimensions of the NEO-PI-R and the MCSDS.
NMean SD Minimum Maximum
MCSDS Subject 65 16.2 5.1 2 30
Spouse 67 16.1 5.0 7 30
Neuroticism Subject 66 90.9 25.2 52 155
Spouse 66 83.9 29.7 24 154
Extraversion Subject 64 118.0 22.0 60 161
Spouse 62 118.5 25.1 57 168
Openness Subject 63 109.6 14.9 76 150
Spouse 67 105.0 17.8 68 161
Agreeableness Subject 64 122.6 16.0 76 152
Spouse 62 126.7 20.8 89 172
Conscientiousness Subject 66 116.0 21.4 60 165
Spouse 64 123.4 26.9 55 170
MCSDS: Marlowe–Crowne Social Desirability Scale; NEO-PI-R: Revised NEO Personality Inventory.
8Psychological Reports 0(0)

Table 2 shows the correlation between subjects’ measured SDR
(MCSDSsubject) and self-ratings of the five dimensions of the NEO-PI-R:
Neuroticism, Extraversion, Openness to Experience, Conscientiousness, and
Agreeableness, together with spouse’s SDR (MCSDSspouse) and spouse ratings
of their partners’ personality traits.
The first line in Table 2 shows the correlations between subjects’ self-ratings
of personality and SDR. The results are mostly in line with what previous
studies on the relationship between self-rated personality and SDR have
found (Holden & Passey, 2010; Kurtz et al., 2008; McCrae & Costa, 1983).
However, contrary to our predictions, the correlation between E and the SDR
did not reach significance in the current study. This may be attributed to a
smaller sample than used in previous studies as the correlation coefficient for
E and MCSDS is in line with that reported in previous studies (r¼.22, p¼.05;
r¼.10, p<.01; r¼.15, p<.05, respectively). Overall the pattern of results is very
similar between the studies. Kurtz et al. (2008) and McCrae and Costa (1983)
interpreted this same pattern of correlations between the SDR and personality
traits as content overlap, and Holden and Passey (2010) as method variance
attributable to self-report.
However, when looking at the observed correlations between SDR and
spouse ratings (shown in the second line of Table 2), it can be seen that spouses’
SDR correlates with their ratings of partners’ personality in the expected direc-
tion and in a similar pattern as their partners’ SDR and self-reported personal-
ity. Although the correlation between subjects’ SDR and A is somewhat stronger
than between spouses’ SDR and A, the difference between the two correlation
coefficients for SDR and A is insignificant, p¼.14, (however, due to the small
sample size of the current study and the resulting lack of power in significance
testing, this finding should be interpreted with caution). On the whole, similar
results were obtained for both partners. As observed for subjects’, a negative
correlation was found for spouses’ between SDR and the undesirable trait N,
and positive correlations between SDR and the desirable traits A and C (but not
for the somewhat desirable trait E).
Table 2. Correlation between self-reports of personality and subjects’ SDR, together with
the correlation between spouse ratings of personality and spouses’ SDR.
Neuroticism Extraversion Openness Agreeableness Conscientiousness
rprprpr p r p
MCSDSsubject .38** <.01 .19 .07 .02 .43 .48** <.01 .29* .01
MCSDSspouse .27* .02 .12 .19 .16 .10 .24* .04 .33** .01
Note: Due to the small sample size, a pairwise deletion was used in all calculations to maintain power.
*Correlation significant at the .05 level (one-tailed). **Significant at the .01 level (one-tailed).
Ve
´steinsdo
´ttir et al. 9

This finding can neither be interpreted as content overlap nor method variance
attributed to self-report. Spouses’ SDR scores cannot be taken to indicate content
of their partners’ personality. Nor can method variance, attributed to self-report,
account for the correlation between self-reports (the MCSDS) and rater-reports
(spouse ratings of personality). Both subject and spouse ratings of personality are
associated with SDR in a similar way, and because the correlation between
spouses’ SDR and spouse ratings cannot be written off as content overlap, a
more likely interpretation is that both measures are contaminated by SDR.
These findings mean that spouse ratings cannot be taken to be an unbiased
external criterion. Partialing out the variance due to subjects’ SDR cannot be
expected to increase the correlation between self-reports and spouse ratings of
personality, and thus the failure of this method to do so does not indicate that
the MCSDS is an invalid measure of SDR.
However, it does not follow that partialing out the subjects’ SDR should
decrease the overall correlation between self-reports and spouse ratings, because
SDR is seen here as an individual difference variable, that is, a tendency of the
respondent, not necessarily exhibited by both partners. Thus, only in the situ-
ation where both respond in a socially desirable manner would partialing out the
subjects’ SDR decrease the correlation between self-reports and spouse ratings.
If, however, one partner responds in a socially desirable manner and the other
does not, partialing out the subjects’ SDR could either have no effect or increase
the correlation, depending on which partner exhibits the tendency. The third
possibility is that neither partner displays this tendency, in which case partial
correlation would have no effect.
As obvious as the result of the third possibility is, McCrae and Costa (1983)
make no mention of this in relation to the insignificant correlation between O
and the MCSDS in their study. Because the correlation was insignificant to
begin with, correcting for MCSDS is a pointless procedure that adds nothing
to the argument that MCSDS is an invalid measure of SDR, even if their
assumption of spouse ratings being an unbiased criterion was correct. Much
the same holds for E, as the correlation between E and the MCSDS was low
(r¼.15) and therefore correcting for the MCSDS cannot be assumed to have
much of an effect. When correcting the correlation between self-reported and
spouse rated N for the MCSDS, the correlation dropped from .58 to .49, which
McCrae and Costa took to indicate content overlap. As explained above,
this interpretation is unwarranted because spouses’ SDR also correlates with
N and thus correcting for MCSDS lowers the correlation for couples where
both respond in a socially desirable manner. It is only when the subject responds
in a socially desirable manner and the spouse does not that an increase in cor-
relation can be expected when correcting for the subject’s SDR.
A limitation to the present study is the small sample size and consequently the
above-explained correction patterns cannot be demonstrated with the current
data. However, with a sufficient sample size, future research could demonstrate
10 Psychological Reports 0(0)

this by splitting the sample into three groups, one with couples where both have
high scores on the MCSDS (HH group), a second group where one partner
would have a high score on the MCSDS and the other would have a low
score (HL group), and a third group where both would score low on the
MCSDS (LL group). Correcting for the MCSDS in the HH group should
reduce the correlation between self-reported and spouse-rated personality,
whether the correction is made only for the subjects’ SDR or both the subjects’
and the spouses’ SDR. In the second group correcting for the subjects’ SDR
could bring the correlation up or down depending on the composition of the
sample, but correcting for both partners’ SDR should increase the correlation.
Corrections for the MCSDS should have no effect on correlation between
self-reports and spouse ratings in the LL group; however, the correlation
observed for this group could be taken to be unbiased by SDR and therefore
the corrected correlation obtained for the other two groups should approximate
the correlation in the LL group, given that the correlation is corrected for both
partners’ SDR.
Conclusion
The debate over the validity of measures of SDR is essentially a debate over the
interpretation of correlational relationships between measures of SDR and per-
sonality, which centers on the content overlap argument proposed by McCrae
and Costa (1983), among others. The argument follows that SDR scales are not
a measure of SDR but a measure of substantive traits, and thence that any
observed correlation between SDR scales and personality measures is therefore
the result of content overlap between measures. McCrae and Costa used spouse
ratings as an external criterion, claiming it to be unbiased. As the current study
shows, spouse ratings are affected by SDR and are thus an invalid criterion.
Furthermore, the change in correlation due to correction for SDR is dependent
on the composition of the sample, that is, to what extent neither, just one, or
both partners display SDR and to what degree, and cannot therefore be pre-
dicted with accuracy. In any case, spouse ratings of their partners’ personality
cannot be claimed to have content overlap with spouses’ SDR, that is, spouses’
SDR cannot be taken to be a measure of their partners’ N, A, and C. With
content overlap ruled out, the correlation between spouses’ SDR and spouse
ratings of personality can hardly be interpreted as displays of a substantive trait,
unless giving favorable descriptions can be called a trait, which would just be a
relabeling of the tendency to give desirable answers—otherwise known as
socially desirable response style.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research,
authorship, and/or publication of this article.
Ve
´steinsdo
´ttir et al. 11

Funding
The author(s) disclosed receipt of the following financial support for the research, author-
ship, and/or publication of this article: This work was part funded by the Centre for
Research and Evidence on Security Threats (ESRC Award: ES/N009614/1) and The
Eimskip Fund of The University of Iceland (Ha
´sko
´lasjo
´ôur Eimskipafe
´lags I
´slands).
The funding sources had no role in study design; in the collection, analysis, and inter-
pretation of data; in the writing of the report; or in the decision to submit the article for
publication.
References
Borkenau, P., & Ostendorf, F. (1992). Social desirability scales as moderator and sup-
pressor variables. European Journal of Personality,6(3), 199–214. doi: 10.1002/
per.2410060303
Cialdini, R. B., Reno, R. R., & Kallgren, C. A. (1990). A focus theory of normative conduct:
Recycling the concept of norms to reduce littering in public places. Journal of Personality
and Social Psychology,58(6), 1015–1026. doi: 10.1037/0022-3514.58.6.1015
Costa, P. T., & McCrae, R. R. (1992). Revised NEO personality inventory (NEO-PI-R)
and NEO five factor inventory (NEO-FFI). Professional manual. Odessa, FL:
Psychological Assessment Resources.
Costa, P. T., & McCrae, R. R. (1997). Stability and change in personality assessment: The
revised NEO Personality Inventory in the year 2000. Journal of Personality
Assessment,68(1), 86–94. doi: 10.1207/s15327752jpa6801_7
Cronbach, L. J. (1946). Response sets and test validity. Educational and Psychological
Measurement,6(4), 475–494. doi: 10.1177/001316444600600405
Crowne, D. P., & Marlowe, D. (1960). A new scale of social desirability independent of
psychopathology. Journal of Consulting Psychology,24(4), 349–354. doi: 10.1037/
h0047358
Dicken, C. (1963). Good impression, social desirability, and acquiescence as suppressor
variables. Educational and Psychological Measurement,23, 699–720. doi: 10.1177/
001316446302300406
Edwards, A. L. (1953). The relationship between the judged desirability of a trait and the
probability that the trait will be endorsed. Journal of Applied Psychology,37(2), 90–93.
doi: 10.1037/h0058073
Edwards, A. L. (1957). The social desirability variable in personality assessment and
research. New York, NY: Dryden.
Ellingson, J. E., Sackett, P. R., & Hough, L. M. (1999). Social desirability corrections in
personality measurement: Issues of applicant comparison and construct validity.
Journal of Applied Psychology,84, 155–166. doi: 10.1037/0021-9010.84.2.155
Funder, D. C., & Colvin, C. R. (1997). Congruence of other’s and self-judgments of
personality. In R. R. Hogan, J. Johnson, & S. Briggs (Eds), Handbook of personality
psychology (pp. 617–647). San Diego, CA: Academic Press.
Furnham, A. (1986). Response bias, social desirability and dissimulation. Personality and
Individual Differences,7, 385–400. doi: 10.1016/0191-8869(86)90014-0
Ganster, D. C., Hennessey, H. W., & Luthans, F. (1983). Social desirability response
effects: Three alternative models. Academy of Management Journal,26(2), 321–331.
doi: 10.2307/255979
12 Psychological Reports 0(0)

Holden, R. R. (2007). Socially desirable responding does moderate personality scale
validity both in experimental and in nonexperimental contexts. Canadian Journal of
Behavioural Science,39(3), 184–201. doi: 10.1037/cjbs2007015
Holden, R. R., & Passey, J. (2010). Socially desirable responding in personality assess-
ment: Not necessarily faking and not necessarily substance. Personality and Individual
Differences,49(5), 446–450. doi: 10.1016/j.paid.2010.04.015
Jo
´nsson, F. H., & Bergo
´rsson, A. (2004). Fyrstu niôursto
¨ôur u´ r sto
¨ôlun NEO-PI-R
a
´I
´slandi [First results of NEO-PI-R standardization in Iceland]. Sa
´lfræôiritiô,9,
9–16.
Kozma, A., & Stones, M. J. (1987). Social desirability in measures of subjective well-
being: A systematic evaluation. Journal of Gerontology,42, 56–59. doi: 10.1093/geronj/
42.1.56
Kurtz, J. E., Tarquini, S. J., & Iobst, E. A. (2008). Socially desirable responding in
personality assessment: Still more substance than style. Personality and Individual
Differences,45(1), 22–27. doi: 10.1016/j.paid.2008.02.012
Lambert, C. E., Arbuckle, S. A., & Holden, R. R. (2016). The Marlowe–Crowne social
desirability scale outperforms the BIDR impression management scale for identifying
fakers. Journal of Research in Personality,61, 80–86. doi: 10.1016/j.jrp.2016.02.004
McCrae, R. R., & Costa, P. T. (1983). Social desirability scales: More substance than
style. Journal of Consulting and Clinical Psychology,51(6), 882–888. doi: 10.1037/0022-
006X.51.6.882
McCrae, R. R., Costa, P. T., Dahlstrom, W. G., Barefoot, J. C., Siegler, I. C., &
Williams, R. B. (1989). A caution on the use of the MMPI K-correction in research
on psychosomatic medicine. Psychosomatic Medicine,51(1), 58–65. doi: 10.1097/
00006842-198901000-00006
McCrae, R. R., & Terracciano, A. (2005). Universal features of personality traits from
the observer’s perspective: Data from 50 cultures. Journal of Personality and Social
Psychology,88(3), 547–561. doi: 10.1037/0022-3514.88.3.547
Meehl, P. E., & Hathaway, S. R. (1946). The K factor as a suppressor variable in the
Minnesota Multiphasic Personality Inventory. Journal of Applied Psychology,30(5),
525–563. doi: 10.1037/h0053634
Nevid, J. S. (1983). Hopelessness, social desirability, and construct validity. Journal of
Consulting and Clinical Psychology,51, 139–140. doi: 10.1037/0022-006X.51.1.139
Nicholson, R. A., & Hogan, R. T. (1990). The construct validity of social desirability.
American Psychologist,45, 290–292. doi: 10.1037/0003-066X.45.2.290
Nunnally, J. C., & Bernstein, I. H. (1994). Psychological theory (3rd ed.). New York, NY:
McGraw-Hill.
Ones, D. S., Viswesvaran, C., & Reiss, A. D. (1996). Role of social desirability in per-
sonality testing for personnel selection: The red herring. Journal of Applied
Psychology,81(6), 660–679. doi: 10.1037/0021-9010.81.6.660
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). San Diego, CA: Academic Press.
Pauls, C. A., & Stemmler, G. (2003). Substance and bias in social desirability responding.
Personality and Individual Differences,35(2), 263–275. doi: 10.1016/S0191-
8869(02)00187-3
Ve
´steinsdo
´ttir et al. 13

Piedmont, R. L., McCrae, R. R., Riemann, R., & Angleitner, A. (2000). On the invalidity
of validity scales: Evidence from self-reports and observer ratings in volunteer sam-
ples. Journal of Personality and Social Psychology,78(3), 582–593. doi: 10.1037/0022-
3514.78.3.582
Reno, R. R., Cialdini, R. B., & Kallgren, C. A. (1993). The transsituational influence of
social norms. Journal of Personality and Social Psychology,64(1), 104–112. doi:
10.1037/0022-3514.64.1.104
Rorer, L. G. (1965). The great response-style myth. Psychological Bulletin,63(3),
129–156. doi: 10.1037/h0021888
Rosenzweig, S. (1933). The experimental situation as a psychological problem.
Psychological Review,40(4), 337–354. doi: 10.1037/h0074916
Rosse, J. G., Stecher, M. D., Miller, J. L., & Levin, R. A. (1998). The impact of response
distortion on preemployment personality testing and hiring decisions. Journal of
Applied Psychology,83(4), 634–644. doi: 10.1037/0021-9010.83.4.634
Smith, D. B., & Ellingson, J. E. (2002). Substance versus style: A new look at social
desirability in motivating contexts. Journal of Applied Psychology,87(2), 211–219. doi:
10.1037/0021-9010.87.2.211
Tesser, A., Pilkington, C. J., & McIntosh, W. D. (1989). Self-evaluation maintenance and
the meditational role of emotion: The perception of friends and strangers. Journal of
Personality and Social Psychology,57, 442–456. doi: 10.1037/0022-3514.57.3.442
Topping, G. D., & O’Gorman, J. G. (1997). Effects of faking set on validity of the NEO-
FFI. Personality and Individual Differences,23, 117–124. doi: 10.1016/S0191-
8869(97)00006-8
Ve
´steinsdo
´ttir, V., Reips, U.-D., Joinson, J., & Thorsdottir, F. (2012, November).
Psychometric properties of an Internet administered version of the Marlowe-Crowne
Social Desirability Scale. Paper presented at the Sixth ISM Workshop on Internet
Survey Methodology, Ljubljana, Slovenia. Supported by COST Action IS1004.
Retrieved from www.webdatanet.eu
Ve
´steinsdo
´ttir, V., Reips, U. D., Joinson, A., & Thorsdottir, F. (2015). Psychometric
properties of measurements obtained with the Marlowe–Crowne Social Desirability
Scale in an Icelandic probability based Internet sample. Computers in Human Behavior,
49, 608–614. doi: 10.1016/j.chb.2015.03.044
Vernon, P. E. (1934). The attitude of the subject in personality testing. Journal of Applied
Psychology,18(2), 165–177. doi: 10.1037/h0074033
Viswesvaran, C., & Ones, D. S. (1999). Meta-analyses of fakability estimates:
Implications for personality assessment. Educational and Psychological
Measurement,59, 197–210. doi: 10.1177/00131649921969802
Wiggins, J. S. (1968). Personality structure. Annual Review of Psychology,19(1), 293–350.
doi: 10.1146/annurev.ps.19.020168.001453.
Author Biographies
Vaka Ve
´steinsdo
´ttir is a PhD student in Methodology at the University of
Iceland, Department of Psychology, but will be joining the Psychological
Methods, Assessment and iScience team at the University of Konstanz as a
postdoc. She obtained her BA and MA degrees at the University of Iceland
14 Psychological Reports 0(0)

on the topics of telephone interviewers’ voice and acceptance rates, and honesty
testing. Her research interests are in survey methodology, particularly in ques-
tionnaire construction and measurement bias, with a focus socially desirable
responding.
Eva D. Steingrimsdottir has a BS degree in Psychology from the University of
Iceland. She also has a BS degree in Computer Science from the University of
Iceland. She currently works as a software developer.
Adam Joinson holds the post of Professor of Information Systems at the
University of Bath, School of Management. His research focuses on the inter-
action between psychology and technology, with a particular focus on how
technology can be to shape behaviour, social relations and attitudes. Recently
this work has taken in privacy attitudes and behaviours, the social impact of
monitoring technology, computer-mediated communication and the human
aspects of cyber security. The EPSRC, ESRC, EU, British Academy and UK
Government have funded this work, and he has published over 80 articles in the
field, as well as editing the Oxford Handbook of Internet Psychology (OUP, 2007)
and authoring two books on psychology and technology. He is principal inves-
tigator for the Cyber-Security Across the LifeSpan (www.csalsa.uk) project and
co-investigator of the Centre for Research and Evidence on Security Threats
(www.crestresearch.ac.uk). His website is www.joinson.com.
Ulf-Dietrich Reips is a full professor in the Faculty of Sciences at the University
of Konstanz, where he holds the Chair for Psychological Methods, Assessment,
and iScience. For more than two decades he has been working on Internet-based
research methodologies (or Internet science), the psychology of the Internet,
measurement, development, the cognition of causality, personality, privacy,
Social Media, crowdsourcing, and Big Data. In 1994, he founded the Web
Experimental Psychology Lab, the first laboratory for conducting real experi-
ments on the World Wide Web. Ulf was elected the first non-North American
president of the Society for Computers in Psychology (SCiP) and he is the
founding editor of the International Journal of Internet Science. Many of his
over 140 publications are among the most highly cited in their journals, see
http://www.uni-konstanz.de/iscience/reips/pubs/publications.html. Ulf has
worked, lived, and studied in California, Colorado, Israel, Germany, Spain,
Switzerland, and the UK. In 2014, he was ranked 7th of ‘‘Top Scientists working
at Spanish Private Universities’’ by the Consejo Superior de Investigaciones
Cientificas, Spain. Recently, he has been asked to direct the Leibniz institute
for Psychology information in Trier, Germany. Ulf and his team develop and
provide free Web tools for researchers, teachers, students, and the public. They
received numerous awards for their Web applications (available from the
Ve
´steinsdo
´ttir et al. 15

iScience Server at http://iscience.eu/) and methodological work serving the
research community.
Fanney Thorsdottir is an associate professor of Psychology at the University of
Iceland, School of Health and a director of the Center of Methodological
Research. She has a BA and a Master (Psychology) from the University of
Iceland, a Diploma in Social Science Data Analysis from the University of
Essex and a PhD (psychology) from the London School of Economics. Her
research interests are in the areas of psychometrics and survey methodology
with a particular focus on measurement bias in self-report measurements.
16 Psychological Reports 0(0)