The Employment Recommendation Questionnaire: III. Validity of Different Types of References¹

Summary A series of investigations were made on the comparative value of different kinds of reference persons who submit employment recommendation questionnaires (ERQ) on applicants for employment in the federal service. ERQ scores on applicants for seven professional and one trade position revealed a general tendency for acquaintances to give the most favorable evaluations, with previous subordinates, co‐workers, and employers next in order. Agreement between the appraisals of supervisors and acquaintances is the case of printers did not yield a correlation significantly greater than zero. Agreement between acquaintances and references familiar with the applicant's work performance (supervisors and co‐workers), while significantly greater than zero, was extremely low. A comparison of the validity of different types of references in predicting later job success of applicants in five trades revealed that supervisors and acquaintances had a very small but statistically significant validity, while personnel officers, co‐workers, and relatives manifested no significant relationship. In the same five trades, supervisors and co‐workers who had known the applicant for a year or longer showed a numerically slightly greater correlation with job success than did reports from similar references of shorter acquaintance, but the coefficient was not statistically significant.