The main aim of this study was to observe the sensitivity for discriminating old and new words for three word types (critical, related. unrelated) in Deese-Roediger-McDermott (DRM) lists. With this aim, for all three kinds of DRM paradigm word types we paired one presented word on study phase against one word that was not presented in each trial in a two-alternative forced choice test. We tried to answer three questions related to false positive responses in the DRM paradigm: First, do false positives stem from a response bias or do the participants have lower sensitivity to distinguish nonstudied from studied words? We used a forced-choice recognition task in order to isolate the effect of sensitivity. Second, is a potential reduction in sensitivity related to recollection or familiarity? We asked participants to classify their responses as "remember", "know", or "guess" in order to explore this issue. Third, is there a difference in sensitivity for prior study in auditory and visual list learning tasks and their distribution into the three kinds of recognition responses? For the first question of the study as a result of the research we observed lower sensitivity for the critical words of the DRM lists than words in unrelated lists. When the findings they classified in terms of recognition memory were examined. it was observed that remember responses clearly differentiated for the three types of words. It was an answer for the second question of the study that sensitivity reduction was related with recollection more than familiarity. Modality did not make a difference in any measure.