Deese hypothesis corrected

In 1959 Deese published work on two hypotheses. The first one, that the probability of free recall intrusions is proportional to the mean association strengths of these intrusions to the presented lists, was found to be true and published in Journal of Experimental Psychology. The second one, attempted to extend this relationship to correct recalls but he believed it to have failed and published this failure in the little read Psychological Reports.Here I hypothesize that the Deese proportionality relationship between the probability of free recall intrusions and their item-item association strengths may actually be correct if we use the post-study association strengths. I test this “Deese-Tarnow” relation indirectly by inferring the increase in studied item-item association strengths in Deese’s experiment and if I use the same constant of proportionality as for the intrusions, I find a reasonable learning curve: increases are largest for word lists with the smallest pre-study and vice versa.If this corrected hypothesis turns out to be true, it would imply that there is no difference in short term memory between correct and incorrect free recall other than the inter-item association strengths: short term memory “truth” would seem to not exist.Experimental predictions are given.