This paper outlines how one measure of response strength to very simple, one-dimensional stimuli may be applied to the study of the attentional behavior of infants. An analysis using Luce's "choice theory" is then applied. The problems of interpretation which may arise when multidimensional stimuli are available for choice are then considered. The method for determining a consistent scale value of the one-dimensional stimuli is described. Formulas are derived for predicting observation time for arbitrary combinations of the strength of the response to three one-dimensional stimuli, and it is shown that even when just a few stimuli and attention values are known, "a large number of predictions flow immediately from a choice analysis." The importance of the stage of visual discrimination ability of the infant, and of the interatrial interval, in interpreting results is noted. A numerical example, when response strengths to two stimuli and to all other (non-defined) stimuli are known, is given.