The Effect of Candidates' Perceptions of the Evaluation Method on Reliability of Checklist and Global Rating Scores in an Objective Structured Clinical Examination