Fig. 3. The factor of interval contrasted the difference between working and long-term memory function. All drugs increased reaction times and decreased correct recognitions in comparison to placebo, but had similar effects among themselves for the most part (except for a greater effect of midazolam, indicated with *). All drugs increased the error rate in categorizations (false alarms, panel  B ), but there was no difference among drugs. Behavioral responses revealed normal function of working memory in the presence of drug (panel  A ; 22 ms faster recognitions of pictures at 6 s; F = 5.83). The functioning of working and long-term memory to support recognitions during the continuous recognition task (CRT) was equivalent (panel  C ; factor of interval not significant [NS]). In other words, the CRT was successful in encoding long-term memories in the presence of sedation. Midazolam produced more impairment than other drugs (significantly increased reaction times [  P = 0.013  vs. placebo; NS  vs. other drugs] and decreased memory function [  P = 0.003  vs. , placebo, lower dose;  P < 0.001  vs. placebo higher dose;  P < 0.04  vs. other drugs, higher dose]). Notably, in terms of memory function, dexmedetomidine was no different from placebo even at the higher dose (panel  C ). All drugs increased forgetting of pictures encoded during the CRT compared with placebo (panel  D ). The degree of forgetting was the same among active drugs, as the calculation (recognition CRT – recognition end of day) normalized the starting point for forgetting. There was less forgetting of pictures recognized from long-term memory than those recognized from working memory (as shown by the factor of interval F = 8.72). Preferentially better long-term memory may have been exaggerated in the case of thiopental (shown by the differences in the third and fourth bars in each cluster), indicating a lack of a conscious amnesic effect (a drug by interval interaction suggested by  P = 0.10, with thiopental being different from placebo [  P = 0.02]).