ON MAKING THE SHORTLIST FOR THE SELECTION OF CANDIDATES

Let x(1) less-than-or-equal-to ... less-than-or-equal-to x(n) and y(1) less-than-or-equal-to ... less-than-or-equal-to y(m) be samples of or der statistics from two populations F1 and F2, respectively. When the scores, say, are pooled and ordered, we address the question 'what is the probability that the r largest x order statistics appear on a shor tlist of k individuals?' This question underlies the selection of stud ents, or of candidates for a job, and has implications in the selectio n process. We study this model for general F1 and F2. Tables are given to calculate probabilities when F1 and F2 are normal distributions th at differ only by location, and also when F1 and F2 are normal distrib utions with different means and variances, but there is only one x-val ue.