|Title||An examination of the comparative reliability, validity, and accuracy of performance ratings made using computerized adaptive rating scales|
|Publication Type||Journal Article|
|Year of Publication||2001|
|Authors||Borman, WC, Buck, DE, Hanson, MA, Motowidlo, SJ, Stark, S, Drasgow, F|
|Journal||Journal of Applied Psychology|
|Keywords||*Computer Simulation, *Employee Performance Appraisal, *Personnel Selection, Adult, Automatic Data Processing, Female, Human, Male, Reproducibility of Results, Sensitivity and Specificity, Support, U.S. Gov't, Non-P.H.S., Task Performance and Analysis, Video Recording|
This laboratory research compared the reliability, validity, and accuracy of a computerized adaptive rating scale (CARS) format and 2 relatively common and representative rating formats. The CARS is a paired-comparison rating task that uses adaptive testing principles to present pairs of scaled behavioral statements to the rater to iteratively estimate a ratee's effectiveness on 3 dimensions of contextual performance. Videotaped vignettes of 6 office workers were prepared, depicting prescripted levels of contextual performance, and 112 subjects rated these vignettes using the CARS format and one or the other competing format. Results showed 23%-37% lower standard errors of measurement for the CARS format. In addition, validity was significantly higher for the CARS format (d = .18), and Cronbach's accuracy coefficients showed significantly higher accuracy, with a median effect size of .08. The discussion focuses on possible reasons for the results.