Abstract
Our study was designed to examine the relationship between testosterone (T) levels and the retention of a movement-timing task for participants who self-controlled their feedback schedule during the acquisition period. Ten males completed an acquisition and retention phase in a double-blind and placebo controlled repeated measures protocol. Each week, participants received either 11mg of placebo or testosterone (NatestoTM). The tasks required participants to accurately click the left mouse button over five targets in exactly 3000ms. Each target had an index of difficulty of 5bits. Participants completed 5 blocks of 12 trials during acquisition and self-controlled their feedback after each trial. An hour after acquisition, participants completed 12 trials without feedback. A week later, participants performed the task in a mirror-reversed order and in the opposite drug condition. The amount of feedback requested decreased from the first to the third block of acquisition, F(4,36)=3.75,p<0.012. However, there were no main effects or significant interactions involving testosterone for the number of feedback requests or the absolute error on trials with and without feedback. Constant error, F(4,32)=4.48,p<0.006, and absolute error, F(4,32)=4.28,p<0.007, decreased significantly from block one to block four and five during acquisition. Absolute error in retention was significantly shorter on placebo (880ms) than on T (1479ms), F(1,9)=5.34,p<0.046. However, the correlation between absolute error and number of errors was non-significant under placebo, r=-0.416,p=0.23, but significant while on T, r=-0.755,p<0.012, indicating that participants lengthened their time in order to enhance accuracy while on T. Results are discussed in the context of instructional settings for typical and special populations.Acknowledgments: Natural Sciences and Engineering Research Council of Canada (J. Carré) and Northern Ontario Heritage Fund Corporation (T. Ortiz).