Skip to main content
Published Online:https://doi.org/10.1027//1015-5759.17.1.17

Summary: Supervisor ratings are useful criteria for the validation of selection instruments but may be limited because of the presence of rating errors, such as halo. This study set out to show that supervisor ratings which are high in halo remain successful criteria in selection. Following a thorough job analysis, a customer service questionnaire was designed to assess the potential of retail sales staff on three “orthogonal” subscales labelled Dealing with people, Emotions and energy, and Solitary style. These subscales were uncorrelated with supervisor ratings made about 8 weeks later. However, the supervisor ratings were correlated with an overall scale derived from the three scales of the customer service questionnaire. These results support the view that supervisor ratings generally consist of global impressions and suggest that these global impressions are useful measures of overall performance. This field study confirms laboratory results that halo does not necessarily reduce rating accuracy.

References

References