A Frequency-Domain Analysis of Skill Learning in Unimpaired Participants: Replication and Extension of Findings

Event related spectral oscillations from scalp-derived Electroencephalograph (EEG) were investigated for practice-based changes across seven days of practicing a novel skill called alphabet addition. This task uses arithmetic problems consisting of both letters and numbers, and is based on the alphabetic sequence instead of the number line. The problems varied in difficulty and were presented with a candidate answer. For each problem, participants used a button-press to indicate whether the candidate answer was true or false (e.g., A+1=B, True; C+5=E, False). On one third of the problems in all sessions, participants were probed to indicate the cognitive strategy they used to produce their answer (memory, counting up in the alphabet, or other). Practice sessions occurred on the first six days, and EEG recordings were collected prior to the practice session on the first, fourth, and seventh day. One set of problems was presented repeatedly in every session, and different sets of new, unpracticed problems were presented during the second and third recording sessions. In replication of previous studies that recorded EEG on only the first and final days of training, it is expected that participants will shift cognitive strategies from counting up in the alphabet to memory retrieval with practice, resulting in faster reaction times and higher accuracy regardless of problem difficulty (Khazen, 2019; Romero et al., 2008). Replication of theta band (i.e., 4-8 Hz) synchronizations is also expected for practiced problems, as theta oscillations are associated with semantic memory and thus increased use of the memory retrieval strategy with practice. Beta-band (i,e., 16-25 Hz) desynchronizations, a marker of increased cognitive load, are expected for difficult problems early in practice and for unpracticed problems presented in later sessions. These desynchronizations should diminish for all practiced problems as memory replaces counting up, and the task becomes less cognitively demanding. The intermediate recording session in this study, along with the strategy probes and the inclusion of additional, unpracticed problems in the second and third recording sessions, allows for an assessment of the linearity of the oscillatory EEG changes and their theorized associations with specific cognitive processing.