Between-individual comparisons in performance evaluation: A prospect theory approach