A controlled reversal-learning study found that methamphetamine altered modeled learning dynamics, with effects depending on baseline performance. That is a more specific result than a general claim that participants made better decisions or earned more points.
Kirschner and colleagues used a randomized, double-blind, within-subject design. Participants completed a drug-free baseline task and later performed under oral methamphetamine, 20 mg, and placebo. The task required adapting decisions as the relationship between symbols and reward changed.
The study analyzed 94 participants and used computational measures to characterize how strongly new information influenced subsequent choices. Baseline testing allowed drug effects to be examined in relation to performance before either experimental condition.
The authors reported changes in dynamic adjustment of learning from prediction errors, particularly among moderately low baseline performers. The low-baseline subgroup’s advantage in total points was only a trend, reported at p = 0.07. It should not be presented as a statistically significant improvement in that outcome.
Performance improved from the baseline session to later sessions, demonstrating practice effects. Therefore, improvement relative to baseline cannot automatically be credited to methamphetamine. The relevant drug comparison is with placebo, with attention to session order and the analysis used.
A single average drug effect may obscure different responses among initially stronger and weaker performers. At the same time, subgroup analyses and model-dependent explanations introduce uncertainty. A model parameter describes behavior under assumptions; it is not a direct recording of neurotransmitter activity.
The experiment concerns healthy-volunteer performance after one oral dose on a particular task. It does not establish that people with dependence, sleep deprivation or additional drug exposures will respond similarly.
The paper supports a qualified account of learning flexibility. It does not show that methamphetamine improves all reasoning, judgment or real-world decisions, and it supplies no concentration-based impairment rule.
Related reading: Selected performance improvements in human methamphetamine experiments.