Курсовая работа: Психологические факторы социальной желательности

Several articles about the validity of shorter versions of the 33-item M-C Social Desirability Scale were reviewed. Strahan and Gerbasi (1972)1 constructed three short forms of the M-C 33. They found the M-C 20 to be almost as internally consistent as the original measure. They would recommend the M-C 1(10 items) and the M-C 2(10 items) in situations where interview time is very limited and “attendant drop in reliability is tolerable.” Reynolds (1982)2 analyzed 6 short forms, three of which were their own and three of which were recommended by Strahan and Gerbasi. They found that the two strongest forms, psychometrically, were the 13-item form which they created (r=.76) and Strahan and Gerbasi’s 20-item form (r=.79). As between the two, they recommend the 13-item because of its brevity. Silverstein (1983)3 evaluated three of Reynold’s short forms and three of Strahan and Gerbasi’s. He found all three of Reynolds forms were “better than random” while this could be attributed to only one of Strahan and Gerbasi’s forms. However, he makes no firm conclusion but rather ascertains that readers themselves must decide if Reynolds Form C (with a corrected validity of .80) is an adequate substitute for the original M-C instrument. Zook and Sipps (1985)4 analyzed short forms of the M-C(33) that were developed by Strahan and Gerbasi (1972), as well as by Reynolds (1982). They paid special attention to differences in response based upon gender, and found that the short forms could be substituted in for the long form without significant loss of reliability. They specifically recommend the Reynolds 13-item short form because it was the form most thoroughly studied at the time of this article’s publication and because it appeared to be better than all other short forms. Fraboni and Cooper5 (1989) concluded that both age and sex account for small but significant amounts of variability in the M-C scores. The M-C 1(10) developed by Strahan and Gerbasi, which is least affected by age and socioeconomic status, is recommended in situations where internal consistency reliability is less important than practical problems (such as respondent burden) that are created by a longer survey. Both the M-C 1(10) and M-C 2(10) were piloted in order to see if there was anything offensive or confusing about either scale. In an elderly population and African American adults recruited in black churches, both scales were found to be acceptable. Among young to middle-aged women surveyed at a community health center, the M-C 2(10) as more well received. Since a preference for the M-C 2(10) was shown in one of the three cases, the M-C 2(10) is being recommended for all BCC sites that wish to measuresocial desirability. Use of the same scale will facilitate uniformity across sites. It was also decided that the instrument would be administered as close to baseline as possible. Please see below for the scale as well as its scoring algorithm. If you should have any questions regarding the Marlowe-Crowne Social Desirability Scale, please contact James R. Hebert (Department of Epidemiology and Biostatistics, Arnold School of Public Health, University of South Carolina)

1 Strahan, R. and Gerbasi, K.C. Short, homogeneous versions of the Marlowe-Crowne Social Desirability Scale. Journal of Clinical Psychology , 1972, 28, 191-193.

2 Reynolds, W.M. (1982). Development of reliable and valid short forms of the Marlowe-Crowne