Poster B121, Sunday, March 26, 8:00 – 10:00 am, Pacific Concourse
Dissecting stimulus-dependent and stimulus-independent factors in an implicit learning task reveals a mixture of performance enhancing and performance eroding processes on different time scales
Balázs Török1,3, Karolina Janacsek2,3, Dávid G. Nagy2,3, Gergő Orbán3, Dezso Nemeth2,3; 1Budapest University of Technology and Economics, 2Eötvös Loránd University, 3Hungarian Academy of Sciences
Performance deterioration during a continuous period of training (termed reactive inhibition) can confound measures of learning in experiments. This may lead to incorrect conclusions for sleep-related consolidation (Rickard et al., 2008). In our study we present a parametric model producing subject-by-subject and trial-by-trial predictions for performance, aimed to dissociate learning from non stimulus-dependent reaction time changes. One hundred and eighty subjects participated in our experiment. The Alternating Serial Reaction Time (ASRT) task was used to measure perceptual-motor learning. We administered one minute long continuous training blocks separated by 15-20 seconds of breaks. Three sessions of 15 blocks were recorded with a 3-5 minute break between sessions. Performance improvement over the 5 minute break is often associated with an early boost (Brawn et al. 2010), however our model can explain such illusory improvements by a quadratic formulation of reactive inhibition. This form suggests a linear increase in reaction times during a block of continuous training with an increase of this slope between blocks. Our results show that including reactive inhibition in our model significantly improves predictive power for 90% of participants. Moreover, reactive inhibition can explain a larger share of variance seen in individuals’ reaction time data. We also exhibit evidence for independence of statistical learning measure used in the ASRT task from reactive inhibition. We discuss methodological implications for learning experiments.
Topic Area: PERCEPTION & ACTION: Motor control