Behaviorally anchored ranking scale versus letters of recommendation for evaluating residency candidates

A behavioral scale and traditional letters of recommendation were compared as tools for evaluating applicants for a hospital pharmacy residency program. In designing the behaviorally anchored ranking scale (BARS), a list of desirable characteristics of hospital pharmacy residents was compiled and descriptions were written of effective, average, and ineffective performance for these characteristics. Twelve characteristics in three categories (professionalism, special skills, and character attributes) were used. Each person who had written a letter of recommendation for the 1983-84 hospital pharmacy residency program at the University of Texas Medical Branch in Galveston was also asked to complete a BARS form. Pharmacy administrators involved in the resident-selection process used the BARS criteria to evaluate all letters of recommendation; responses on the BARS forms were then analyzed. For 18 applicants, 46 letters of recommendation and 39 BARS forms were completed. Many of the BARS characteristics were not addressed in letters of recommendation. None of the letters commented on leadership quality. Ability to work with others was the BARS characteristic most often mentioned in the letters (58.7%). Of characteristics that were mentioned in both the letters and the BARS, 82.4% were ranked equally in both. Of all the letters and scales, 59% were not in agreement with each other. The BARS provided a clearer picture of the characteristics and qualities of a hospital pharmacy residency candidate than did the letters of recommendation.