The effect of blinded versus nonblinded interviews in the resident selection process.

OBJECTIVE: To evaluate whether resident applicants' academic performance biases the assessment of nonacademic qualities. METHODS: In this prospective, descriptive study, 2 blinded (personal statement only) and 1 nonblinded (application) 30-minute interviews were compared for candidates ranking into Top 10, Upper Third, Middle Thirds, Lower Third, and Do Not Rank classes. RESULTS: A total of 234 candidates were interviewed from 2005 to 2007. The association between blinded interviewers for the categories was 87%, 63%, 68%, 73%, and 90% (P  =  .0000), respectively. Comparing blinded to nonblinded interviewers showed an association of 75% (63%), 71% (86%), 68% (58%), 66% (79%), and 72.7% (82%) (P  =  .0000), respectively. A strong degree of agreement (Cohen κ, 0.75) for the 2 ranking scores resulted in 90% agreement for Top 10 and Upper Third and 85% for Middle Third and Lower Third categories. No correlation was found between United States Medical Licensing Examination scores and final ranking; moderate agreement was found between ranking and deans' letters (Cohen κ, 0.59, P  =  .0000). CONCLUSION: Candidate rankings on nonacademic attributes were not affected by interview type.