Abstract
Research that investigated validity and effectiveness of corrections using social desirability scales was reviewed. Social desirability scales seemed to detect response distortion, but the linear relation between response distortions and social desirability scales presupposed in the corrections using social desirability scale scores was seen only in a small subset of personality dimensions. The methods to correct individual scores using social desirability scores were effective in the aggregate level to some degree, but ineffective in the individual level, and failed to produce corrected scores that could approximate individuals' honest scores. Social desirability corrections using part or partial correlations did not improve criterion-related validity of personality constructs. It would be necessary to reexamine traditional correction methods that used social desirability scales, starting with the issue of what social desirability scales really measured.
