172 CHAPTER 5
Social desirability
A bias toward answering questions in a way
that respondents think makes them “look
good” (i.e., that he or she thinks is socially
desirable), even if the responses are not true.
regardless of the content of the statement (Javeline, 1999; Welkenhuysen-Gybels,
Billiet, & Cambré, 2003). This type of response bias, called acquiescence, can be
reduced by wording half the items negatively. Thus, if you were interested in as-
sessing self-reported shyness in a questionnaire, you might include both the item
“I often feel shy when meeting new people” and the item “I don’t usually feel shy
when meeting new people,” which is simply a negative rewording of the fi rst item.
Another type of response bias is social desirability: answering questions in a way
that that the respondent thinks makes him or her “look good” (that is, in a way that he
or she thinks is socially desirable), even if the answer is not true. For instance, some peo-
ple might not agree with the statement “It is better to be honest, even if others don’t like
you for it.” However, they may think that they should agree and respond accordingly.
In contrast to a social desirability bias, some people answer questions in a way that they
think makes them “look bad” or look worse than they actually are. People with malin-
gering and factitious disorder (see Chapter 3) are likely to respond so that they appear
more ill than they are. To compensate for these biases, many personality inventories
have a scale that assesses the participant’s tendency to answer in a socially desirable or
falsely symptomatic manner. This scale is then used to adjust (or, in the language of test-
ing, to “correct”) the scores on the part of the inventory that measure traits.
Another type of response bias can occur because people generally assume that
the center of the response scale is average. If a response scale is set up so that in-
dividuals believe that their responses are above (or below) average, they may shift
their responses so that they are closer to average; see Figure 5.5 (Schwarz, 1999).5.5 • High- and Low-Frequency Scales Depending on where
someone’s response falls on a low- or high-frequency scale, the person may shift
the response closer to the assumed average—the center.
Source: Schwartz, 1999. Copyright © 1999 American Psychological Association. For more
information see the Permissions section.Figure 5.555 Hi h dL F S l d h
Low–Frequency ScaleHigh–Frequency Scale( ) ( ) ( ) (X) ( ) ( )
( ) (X) ( ) ( ) ( ) ( )
Less than
once a
monthAbout once
a monthAbout once
in two
weeksAbout
once
a weekAbout twice
a weekMore
oftenLess than
twice a
weekAbout twice
a weekAbout
four times
a weekAbout
six times
a weekAbout once
every 24
hoursMore
oftenChallenges in Researching Social Factors
Challenges in researching social factors not only affect what can be learned about
such factors themselves, but also affect what can be learned about other sorts of fac-
tors. Information obtained from and about people always has a context. For research
on psychopathology, a crucial part of the context is defi ned by other people. Social
factors arise from the setting (such as a home, hospital, outpatient clinic, or university),