Changes in the need for social approval, 1958–2001

https://doi.org/10.1016/j.jrp.2006.03.006Get rights and content

Abstract

American college students’ and children’s scores on two measures of the need for social approval closely follow changes in the state of the larger society, decreasing significantly from 1958 to 1980 and leveling off between 1980 and 2001 (total n = 36,004 across 203 samples of college students responding to the Marlowe–Crowne Social Desirability scale; total n = 4741 across 38 samples of children responding to the Children’s Social Desirability Questionnaire). Need for social approval correlates with positive social trends such as a low divorce rate, low crime rate, and low unemployment rate. However, need for social approval does not correlate over time with changes in anxiety and self-esteem, suggesting that these birth cohort trends are not due to shifts in response styles.

Introduction

Bob is careful about how he dresses, rarely talks about his bad qualities, and tries to present himself in a good light. In other words, he wants other people to approve of him. Mike, in contrast, dresses poorly, readily admits to his faults, and feels no need to shape his behavior to make a good impression. Thus Mike he does not much care what other people think of him. What is the difference between these two men? Bob is high in the need for social approval, whereas Mike is low in this trait.

Over the past forty years, the Marlowe–Crowne Social Desirability Scale (MCSD; Crowne and Marlowe, 1960, Crowne and Marlowe, 1964) has been cited in over 3600 articles. The authors of the MCSD describe the scale as a measure of the need for social approval, which can be considered a quantifiable personality trait. For example, people who scored high on the MCSD displayed obedience to authority (Marlowe & Crowne, 1961), conformed in an Asch-like judgment paradigm (Strickland & Crowne, 1962), and anchored their social behavior in cultural norms (Horton, Marlowe, & Crowne, 1963). Overall, they were characterized by “conventional, polite, acceptable behavior” (Crowne & Marlowe, 1964, p. 39) and shaped their behavior around “conventional, even stereotyped, cultural norms” (p. 85). Paulhus (1991) describes those who score high on scales like the MCSD as presenting “a socially conventional, dependable persona” (p. 21).

For much of its history, the MCSD has been used as a measure of socially desirable responding (SDR). There has been much debate about what SDR scales actually measure. Many of these scales were originally intended to measure the “response bias” of trying to look good to others (Paulhus, 1991). However, several recent studies have found that using SDR scales to “correct” personality test scores does not improve results (Ellingson et al., 1999, Ones et al., 1996, Piedmont et al., 2000). In many cases, it undermines validity because the tendency toward SDR may be considered a personality trait in and of itself (McCrae & Costa, 1983). Thus this paper will instead define the MCSD as a measure of the need for social approval, just as its authors originally labeled it.

In this paper, we explore birth cohort, or generational, changes in the MCSD and its companion measure for children, the Children’s Social Desirability Questionnaire (CSDQ: Crandall, Crandall, & Katkovsky, 1965). We examine how scores on these measures have changed in the United States between 1958 and 2001, analyzing samples of college students and children from the literature in a cross-temporal meta-analysis (e.g., Twenge & Campbell, 2001). Examining like-aged samples collected at different times allows us to examine birth cohort changes in the measure. In other words, were 1950s young people more concerned with social approval than young people in the 1970s? What about today?

As many theorists have argued, the larger social environment can have a substantial impact on the individual. For example, research finds considerable differences between Eastern and Western selves (e.g., Choi et al., 1999, Heine and Lehman, 1997, Markus and Kitayama, 1991, Suh et al., 1998). In addition, culture affects questionnaire responses and self-presentation (e.g., Heine and Renshaw, 2002, Heine et al., 2000, Kitayama et al., 1997).

Generations and birth cohorts experience varying cultures as well. For example, 1950s America was a very different culture than 1970s America. Thus birth cohort becomes a proxy for the larger cultural environment (Caspi, 1987, Stewart and Healy, 1989, Twenge, 2000, Twenge, 2006). These cultural shifts have an impact on the individual. Recent research has found birth cohort changes in many different traits and attitudes, including anxiety, extraversion, self-esteem, locus of control, and depressive symptoms (Klerman and Weissman, 1989, Twenge, 2000, Twenge, 2001a, Twenge and Campbell, 2001, Twenge et al., 2004). This paper seeks to explore whether cultural trends over time have also affected individuals’ need for social approval.

In psychology, most previous work on environmental effects (especially within countries) focused on family environment. While some of this research has been successful (e.g., Reiss, 1997), much of it has failed to find that growing up in the same house and with the same parents has much measurable effect on personality (e.g., Bergeman et al., 1988, Langinvaionio et al., 1984, Loehlin, 1992, Rowe, 1990, Shields, 1962). However, many of these studies do not consider environmental influences outside the immediate family. Yet as the studies on culture demonstrate, the larger social environment can have a substantial impact on individuals’ personalities. This research seeks to demonstrate that the need for social approval has changed over time in response to shifts in American culture over the last 40 years.

We hypothesize that the need for social approval will follow the trends in American society. By definition, the need for social approval indicates a desire to conform, a concern with others’ opinions, and an urge to be socially acceptable. People high in the need for social approval care what others think. Thus, cultures that value social harmony, integration, and collectivism should have higher MCSD scores. Cross-cultural research supports this assumption, finding that people from collectivistic cultures (such as Asian countries and Mexico) score higher on the MCSD compared to Americans (Middleton and Jones, 2000, Ross and Mirowsky, 1984).

What do these predictions mean for change across historical time in the United States? The 1950s were infused with conformity to social rules and rigid codes of behavior and dress; it was important to fit in and be seen as conventional (Yankelovich, 1981). 1950s society was also high in social integration (i.e., stable social connections), and promoted the idea that it was better to get along with others than express unorthodox ideas. Thus the need for social approval is likely to be high during the 1950s and early 1960s.

During the mid- to late-1960s, however, young people began to flout social norms, rebelling against conformity with unconventional dress and unconventional life choices. By the 1970s, this anti-authoritarian mentality had entered the mainstream, glorified by pop psychology books and the human potential movement (Frum, 2000). A bestselling self-help book during this time claimed that others’ opinions limited the potential of the individual; its seminal chapter was titled “You Don’t Need Their Approval” (Dyer, 1976; for a review of the waning of conformity, see Chapter 1 of Twenge, 2006). As Fukuyama (1999, p. 13) put it, the trend since the 1960s has been toward “the liberation of the individual from unnecessary and stifling social constraints.” Roberts and Helson (1997) found that responses to certain California Psychological Inventory items shifted over this time period, with more recent generations more likely to endorse individualistic and narcissistic items and less likely to endorse items promoting conformity and social norms. Those who scored high on this “secular trends index” were described by others as rebellious, unconventional, uninhibited, and pleasure seeking, whereas those low on the index were described as formal, conscientious, and conventional.

In addition, the late 1960s and 1970s were characterized by societal disintegration, manifested in a wobbly economy, a heightened crime rate, and more transient social relations (for example, the increase in the divorce rate). In general, social connections decreased while threats in the environment increased during the period between the 1950s and the late 1970s (Twenge, 2000). Fig. 1 shows the pattern of change in a z-scored composite of seven social statistics including the divorce rate, birth rate, crime, youth suicide, unemployment, inflation, and the Dow Jones Industrial Average. (This index has been used in previous birth cohort studies; e.g., Twenge & Campbell, 2001. The statistics were obtained from the Statistical Abstract of the United States). This index of social statistics shows a steep trend downward between the 1950s and the late 1970s.

Thus several types of evidence suggest that the need for social approval will be high in the 1950s but will steadily decrease throughout the 1960s and 1970s. Numerous sources document an increase in individualism and a decrease in conformity over this time, and this is also reflected by quantifiable social statistics such as the crime rate and the divorce rate.

However, these social trends improved somewhat during the 1980s and 1990s as American society regained some social integration. For example, the divorce rate decreased somewhat, unemployment and inflation stabilized, and crime decreased during the 1990s (see Fig. 1). Other sources also suggest an increase in social integration after the 1980s, when outright rebellion went out of style (Strauss & Howe, 1991). For example, college students in the 1980s and 1990s were more likely to support the death penalty and less likely to support the legalization of marijuana than their counterparts in the 1970s (Dey, Astin, & Korn, 1992). Drug and alcohol use among teenagers declined during the 1980s and 1990s (Statistical Abstract, 2005). However, conformity did not return to 1950s and early 1960s levels, as individual freedom was valued throughout this period (Twenge & Campbell, 2001). For example, high school students’ trust in others and church attendance continued to decline during the 1980s, while their tolerance for school cheating and out-of-wedlock births increased (Bachman et al., 1998, Bronfenbrenner et al., 1996).

Thus the social integration of American society leveled off during the 1980s and 1990s—it did not continue to decrease sharply as it did during the 1970s, but it also did not return to 1950s levels. We hypothesize that the need for social approval will mirror these shifts in society, decreasing sharply between the 1950s and the late 1970s, and leveling off after 1980.

An exploration of changes in the need for social approval will also address an unanswered question: are people just more willing to admit to certain characteristics during some time periods, or do people from different birth cohorts really have different personalities? Are previously discovered birth cohort trends due to self-report bias or to “real” change? If MCSD scores are correlated with traits when matched by year, this will suggest that the need for social approval might explain some of the change. If they are not correlated, this would suggest that the need for social approval cannot explain the birth cohort differences.

Consider birth cohort differences in anxiety and neuroticism. Twenge (2000) found that both college student and child samples self-reported significantly more anxiety symptoms from the 1950s to the 1990s. The shift in scores was almost a standard deviation. However, it is possible that people simply became more willing to admit to anxiety. Swindle, Heller, Pescosolido, and Kikuzawa (2000) favored the need for social approval explanation for their survey data on nervous breakdowns, which found that 7% more Americans in 1996 (compared to 1957) said they felt they were going to have a nervous breakdown. The authors argued that the results were due to a decrease in the stigma associated with mental illness and describe their results in terms of how “willing” respondents were “to admit” to a nervous breakdown, although they acknowledge that an actual change in diagnosable disorders could also be responsible for the shift. Examining changes in the need for social approval will help determine if response bias or actual changes in psychopathology caused the shifts in anxiety.

In this article, we examined change over time in MCSD and CSDQ scores. The results will show how the need for social approval has changed in the United States between 1958 and 2001, and whether these changes in individuals’ need for social approval covary with the state of society.

Study 1 involves 203 samples of American college students responding to the MCSD, collected from the literature between 1958 and 2001. In Study 2, we examined 38 samples of children (ages 8–12) who completed the Children’s Social Desirability Questionnaire (CSDQ), designed as a conceptual equivalent to the MCSD (Crandall et al., 1965). The samples of children in Study 2 were collected to verify the pattern of change in the need for social approval in a population of a different age that is also less prone to sampling bias.

Both of these analyses gathered data from the literature, using a modified meta-analysis technique called cross-temporal meta-analysis or within-scale meta-analysis (e.g., Twenge, 2000, Twenge and Campbell, 2001). Cross-temporal meta-analysis examines the correlation between mean scores on psychological measures and the year those data were collected (either reported or estimated). In this case, we gathered data from journal articles and dissertations that reported mean scores for their samples on the MCSD or CSDQ. If there is a significant correlation (or curvilinear relationship) between the mean scores and year, this will demonstrate that the need for social approval has changed with historical time.

Section snippets

Study 1

Study 1 gathered data from samples of American college students who completed the MCSD between 1958 and 2001. Using samples from the literature eliminates the problem of many birth cohort studies, which have often relied on retrospective accounts. Many studies of birth cohort effects in depression, for example, ask respondents to recall past episodes of depression (e.g., Klerman and Weissman, 1989, Lewinsohn et al., 1993). In contrast, this method examines questionnaire responses collected at

Study 2

In Study 2, we sought to replicate the birth cohort change in the need for social approval in a sample of children, examining scores on the Children’s Social Desirability Questionnaire (CSDQ; Crandall et al., 1965). Although it is unlikely that changes in college samples can explain the results of Study 1 (given that these populations have remained relatively homogeneous), a replication with children would help rule out that possibility. Enrollment in elementary and middle schools remained

General discussion

Both college student’s and children’s need for social approval decreased during the 1960s and 1970s and leveled off after 1980. Thus younger generations are less concerned with being polite, conventional, and acceptable to others. The replication among child samples suggests that the Study 1 results were not the product of college student samples or any changes in those populations. In fact, the changes in child samples were even stronger than those for college students. The decline appeared in

Conclusions

The changes in the need for social approval demonstrate the cultural demands that come into play when people complete personality and attitude assessments. In a society of close social connections, seeking the approval of others is very important, as the public face of the individual must be appealing to others and not offend anyone. In a looser culture, people place greater importance on being “true to my real self” (Yankelovich, 1981; p. 69), and that becomes the culturally accepted value. In

References (68)

  • A. Caspi

    Personality in the life course

    Journal of Personality and Social Psychology

    (1987)
  • I. Choi et al.

    Causal attribution across cultures: variation and universality

    Psychological Bulletin

    (1999)
  • V.C. Crandall et al.

    A children’s social desirability questionnaire

    Journal of Consulting Psychology

    (1965)
  • D.P. Crowne et al.

    A new scale of social desirability independent of psychopathology

    Journal of Consulting Psychology

    (1960)
  • D.P. Crowne et al.

    The approval motive

    (1964)
  • E.L Dey et al.

    The American freshman: Twenty-five year trends, 1966–1990

    (1992)
  • W.W. Dyer

    Your Erroneous Zones

    (1976)
  • J.E. Ellingson et al.

    Social desirability corrections in personality measurement: issues of applicant comparison and construct validity

    Journal of Applied Psychology

    (1999)
  • D. Frum

    How we got here: The 70s, the decade that brought you modern life (for better or worse)

    (2000)
  • F. Fukuyama

    The great disruption: Human nature and the reconstitution of social order

    (1999)
  • Gough, H.G. (1991). Scales and combinations of scales: what do they tell us, what do they mean? Paper presented at the...
  • S.J. Heine et al.

    Culture, dissonance, and self-affirmation

    Personality and Social Psychology Bulletin

    (1997)
  • S.J. Heine et al.

    Interjudge agreement, self-enhancement, and liking: Cross-cultural divergences

    Personality and Social Psychology Bulletin

    (2002)
  • S.J. Heine et al.

    Beyond self-presentation: evidence for self-criticism among Japanese

    Personality and Social Psychology Bulletin

    (2000)
  • R. Helson et al.

    Personality change over 40 years of adulthood: hierarchical linear modeling analyses of two longitudinal samples

    Journal of Personality and Social Psychology

    (2002)
  • D.L. Horton et al.

    The effect of instructional set and need for social approval on commonality of work association responses

    Journal of Abnormal and Social Psychology

    (1963)
  • L.D. Jamner et al.

    Discourse between verbal and non-verbal pain indices: the influence of repressive coping strategies

    Psychophysiology

    (1986)
  • S. Kitayama et al.

    Individual and collective processes in the construction of the self: self-enhancement in the United States and self-criticism in Japan

    Journal of Personality and Social Psychology

    (1997)
  • G.L. Klerman et al.

    Increasing rates of depression

    Journal of the American Medical Association

    (1989)
  • K.C. Kling et al.

    Gender differences in self-esteem: a meta-analysis

    Psychological Bulletin

    (1999)
  • H. Langinvaionio et al.

    Finnish twins reared apart: 3. Personality factors

    Acta Geneticae Medicae of Gemellogiae

    (1984)
  • P. Lewinsohn et al.

    Age-cohort changes in the lifetime occurrence of depression and other mental disorders

    Journal of Abnormal Psychology

    (1993)
  • M.R. Lipsey et al.

    Practical meta-analysis

    (2001)
  • J.C. Loehlin

    Genes and environment in personality development

    (1992)
  • Cited by (0)

    We thank Jeff Bryson and Niels Christensen for their comments on an earlier version of the manuscript, and Mark Reid for designing the graphs.

    View full text