The purpose of this experiment was to determine whether a constant information processing rate would occur when subjects in verbal discrimination (VD) learning were presented a mixture of items of different lengths. Forty-two Naval Postgraduate School students served in a VD experiment with a random mixture of two-, three-, and four-word items at presentation rates of 1/2 or 1/4 bits of information per second. Half the items had similar and half, dissimilar words. The VD list had six each, of two-, three-, four-word items resulting in six bits of information for the two-word items, 9.51 bits for the three-word items, and 12 bits for the four-word items for an overall presentation load of 27.51 bits of information. The information content of two-, and three-word items was normalized to agree with the 12 bit base of four-word items for analysis. The similarity variable was disregarded, since it was not significant as a main effect. The results showed that, at each presentation rate, two-, three-, and four-word items were learned at the same information processing rate. Moreover, the interaction between presentation rate and blocks of trials showed an expected multiplicative function such that the performance of one group could be predicted with 93.5 percent accuracy. (Author)