“…Acceptable reliability levels were only rarely reported for supervisor ratings (Kingdon, 1975), peer ratings (Martin & Carkhuff, 1968), and self-ratings (Bailey et al, 1977;Carter & Pappas, 1975;Perlman, 1972). Supervisor, peer, and self-ratings have been found to covary significantly, or to yield parallel pre-and postchange patterns in many instances (Berenson et al, 1966;Bishop, 1971;Borman & Ramirez, 1975;Dilley, 1964;Friesen & Dunning, 1973;Martin & Gazda, 1970), especially after extensive training programs (Burck, Jacobs, Saubra, Stone, & Thomson, 1973).…”