Models for “how People Predict Three Balls Fall into Three Cells”

Abstract

2 models are presented for Lieberman's (1963) study of how Ss predict 3 distinguishable balls will fall into 3 cells. Both models assume that there are differential preferences for the cells. These are represented by cell preference probabilities which govern the placement of balls into cells. In the pattern model this assumption is supplemented by a set of preferences (represented by probabilities) for patterns in which the balls may be arranged. In the repetition model the cell preferences are supplemented by a duplication probability which represents the notion that S considers, before placing the 2nd and 3rd ball, whether or not they will fall into an already occupied cell. Comparisons are made with a simple equiprobable model.