Christian Subcultural Differences in Item Perceptions of the MMPI-2 Lie Scale

Article excerpt

Interpretation of the MMPI-2 Lie scale assumes all groups understand the items similarly. This study hypothesized that a Christian subculture would view certain items differently due to religious salience. In Study 1, it was confirmed that those immersed in a Christian subculture differed from the MMPI-2 norm sample's lie scale as predicted on five religiously salient items. Factor analysis supported two factors, with these five items plus one other loading on a religious factor and the remaining nine loading on a nonreligious factor. Study 2 replicated this two-factor model in a more general sample and found that Christians viewed the religiously salient items as more objectionable than the other nine, and only these six items correlated with five measures of religiousness. Findings suggest that a Christian subculture may interpret some MMPI-2 lie scale items differently than others, making interpretation of their lie scale scores questionable.

The MMPI-2 Lie scale consists of behaviors that, while being generally viewed as inappropriate, are assumed to occur in normal populations. However, different subcultures have different standards for behavior, potentially leading some subcultures to have contrasting perceptions regarding a subset of the items. For example, Wen (1970) found lie scale item validity differences for specific items, based on race and sex, using the Eysenck Personality Inventory (1903). Similarly, Maiocco (1990) found that Mexican Americans scored higher on the MMPI2 lie scale than African Americans and Caucasians, while Callahan (1998) found Mexican American women scored higher than Caucasians. If differing perceptions occur, scores for people from such a subculture might reflect assumptions other than those underlying the scale.

A clear subcultural difference for Christians between total scores has been established for the MMPI lie scale (note that the lie scale was unchanged from the MMPI to the MMPI-2). Early research by Brown and Lowe (as cited in Dahlstrom, & Welsh, 1900) found that college Bible students scored higher on the MMPI lie scale than non-religious students. Similarly, a study of a Christian religious sect using the MMPI (Delay, Pichot, Sadoun & Perse, 1955) found a significant elevation in lie scale responses for group members. Based on his own comparable findings, Hood (1975) also proposed that the MMPI lie scale might have questionable validity for religiously committed persons. More recently, Reichwald (1999) reported that the MMPI-2 lie scale significantly correlated with a religious measure called the Shepherd Scale (Basset et al., 1981), designed to differentiate Christians from non-Christians. These findings suggest the need to further evaluate whether the MMPI-2 lie scale is affected by Christian subcultural differences.

With some exceptions (e.g., Wen, 1970), the finding of significantly higher lie scores in religious respondents has generally been interpreted as a manifestation of a general approach to the entire item set. A significant amount of research has focused on how Eysenckian lie scales perform for general and religious samples with varying conclusions. Some have interpreted high lie scores as reflecting a naive tendency to lie or "fake good" (Dahlstrom & Welsh, 1900; O'Donovan, 1909; Rigby, 1987). In this case the interpretation would be that religious respondents tend to rely more on naive, idealized self-misrepresentation regarding presentation to others. Similarly high lie scores have been linked to: (a) lack of insight (Crookes & Buckley, 1970; Eysenck, Nias, & Eysenck, 1971; Kirton, 1977; Rump & Court, 1971) which is connected to immaturity (Francis, Pearson & Kay, 1983; Francis, Pearson, & Kay, 1988; Francis, Pearson, & Stubbs, 1985) and (b) social acquiescence or conformity (Finlayson, 1972; Massey, 1980; O' Hagan, 1981; Powell, 1977). On these two accounts religious respondents emerge as either lacking in self-insight and maturity or as more socially conforming. …


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.