The 'less-is-better' effect in capuchin monkeys (Sapajus sp.): Some data and a contrast-like hypothesis

Given a choice between a simple option offering a preferred-food item (e.g., a grape, G) and a combo option offering the same preferred-food item plus a less-preferred food item (e.g., a grape + a slice of cucumber, GC), animals often behave suboptimally by either being indifferent between the two options or by preferring the simple option-the "less-is-better" effect. To explain indifference, the selective-value hypothesis assumes that, in the choice context, the subjective value (V) of the less-preferred food is zero (i.e., V_GC = V_G + 0 = V_G). To explain the less-is-better effect, the average-quality hypothesis assumes that the value of the combo equals the average of its components' values [i.e., V_GC = Average(V_G,V_C) < V_G]. No unified account explains both sets of experimental findings. To test these hypotheses further, we presented six capuchin monkeys (Sapajus sp.) with a variety of choice tests, some simple (GC vs. G) and some complex (2G1C vs. 3G1C; or 2C1G vs. 3C1G), with a final sample of four individuals per test. The results confirm that capuchin monkeys also behave suboptimally, revealing either indifference or the less-is-better effect. Crucially, our findings suggest that the value of the less-preferred food, C, may become negative through a contrast-like effect. By expanding the selective-value hypothesis to accommodate situations where the less-preferred item's value may be reduced to zero or become negative, we suggest a unified, process-based account of the two sets of research findings.