Abstract
The purpose of the present study was to examine the applicabillity of the Edwards Personal Preference Schedule (EPPS) to Japanese subjects with special reference to its social desirability control. The following results were obtained: (1) When social desirability value was measured of 135 statements of the EPPS, a fairly high r of. 96 was found between the two sex groups, a slightly lower r of. 85 between normal males and delinquent males, and a considerably lower r of. 48 between Japanese and Americans.(2) Similar relationships were obtained with respect to percentage of endorsement of the statements in selfratings. No comparison was made between the two ethnic groups since the American data were not available.(3) Inter-statement r was not high in endorsement for most of the need categories.(4) Correlation between desirability and endorsement measured on independent groups was high (.81) both for males and females, and the values roughly corresponded with that obtained by Edwards.(5) When the statements were paired exactly as in the American EPPS, an intraclass r of. 20 was found between social desirability values of the paired statements, which was quite low in comparison with. 85 in American data. However, it increased up to about.7 in the Berrien revision in which statements were revised in order to equalize desirability values. Similar results were obtained by different analyses.(6) When the EPPS in the usual forced-choice form was adminstered under two different instructions (social desirability and self-ratings) on independent groups, subjects were found to respond differently under two types of instuctions. In addition, the social desirability conception of the EPPS statements based on the forced-choice form was substantially related to that based on the form of inventory.(7) So far, social desirability control of the EPPS was studied based on independent group data. This way of analysis is valid if the scale is applied only for group studies, such as cross-cultural investigations. However, in personality assessment, social desirability must be controlled using individual data, or at least the data based on the same groups. Although the correlation based on independent groups are not directly comparable with that based on the same group, the present study indicated that correlation between social desirability and endorsement of the EPPS statements was low in this form of the inventory and especially so in the forced-choice form when it was computed on the same subjects. The findings were discussed in relation to previous studies.
