“…More recently, and more directly aimed at multi-rater appraisal, London (1995) identi® ed the following list of items as being important: c content of appraisal (relevant, familiar, observable behaviours) c format of appraisal form c involvement in design (more likely to yield commitment) c clarity of procedure and purpose c training to make aware of common errors and biases c rater anonymity c use for development or rewards c inclusion of self-assessment c frequency of feedback c format of report; level of personal interaction in delivery of results Conceptually, the factors from these and other previous studies divide into three classes of variables which could potentially affect the success (as de® ned above) of peer evaluations ± and, to some extent, any kind of evaluation system. These are the organizational context (Maurer and Tarulli, 1996), the design of the system itself (Kane and Lawler, 1978), and the implementation of the system by individuals and groups within the organization (Zedeck and Cascio, 1982). Drawing on the studies and on preliminary ® eld-work, under each of these broad headings I chose a set of items to be investigated as potential conditions for the success of peer evaluation.…”