Improving the selection interview

Whether interviewers are selecting students for courses, or applicants for jobs, they are likely to find themselves in the paradoxical situation of being fairly confident in a judgement which most research suggests is likely to have little predictive validity. The interview stands firm as the most frequently used selection technique, in spite of the avalanche of research evidence which threatens its existence. Ever since a Mr. Scott working on salesman selection in 1915 found that none of his interviewees agreed with each other about the relative merits of the applicants, the interview has been shown wanting in most respects as a means of assessing people. However if we can't beat it perhaps we had better join it; but let's join it armed with as much research evidence as we have available, so as to ensure that it's used as well as possible.