Self‐reports, peer ratings and construct validity

Self‐reports of behaviour have been criticized as fraught with problems that seriously undermine the construct validity of conventional personality measures. The problems are related to a putative absence of, or distortion of, a person's knowledge about himself or herself. A proposed solution is to use peer reports of personality instead, because such observer ratings are presumably more impartial and free of the distorting influences affecting self‐reports. We review some past research on moderators of the agreement between self‐reports and peer ratings of personality, arguing that those findings support the validity of self‐ratings. We conclude that peer ratings, although highly useful as adjuncts to other methods of assessment, are no substitute for self‐reports as a source of personality information. Copyright © 2010 John Wiley & Sons, Ltd.