Game-based training of flexibility and attention improves task-switch performance: near and far transfer of cognitive training in an EEG study

Abstract

There is a demand for ways to enhance cognitive flexibility, as it can be a limiting factor for performance in daily life. Video game training has been linked to advantages in cognitive functioning, raising the question if training with video games can promote cognitive flexibility. In the current study, we investigated if game-based computerized cognitive training (GCCT) could enhance cognitive flexibility in a healthy young adult sample (N = 72), as measured by task-switch performance. Three GCCT schedules were contrasted, which targeted: (1) cognitive flexibility and task switching, (2) attention and working memory, or (3) an active control involving basic math games, in twenty 45-min sessions across 4-6 weeks. Performance on an alternating-runs task-switch paradigm during pretest and posttest sessions indicated greater overall reaction time improvements after both flexibility and attention training as compared to control, although not related to local switch cost. Flexibility training enhanced performance in the presence of distractor-related interference. In contrast, attention training was beneficial when low task difficulty undermined sustained selective attention. Furthermore, flexibility training improved response selection as indicated by a larger N2 amplitude after training as compared to control, and more efficient conflict monitoring as indicated by reduced Nc/CRN and larger Pe amplitude after training. These results provide tentative support for the efficacy of GCCT and suggest that an ideal training might include both task switching and attention components, with maximal task diversity both within and between training games.

Fig. 1

Schematic overview of the study design

Fig. 2

Example trials of the TS for crosstalk blocks. The RSI conditions varied between but not within blocks. The no-crosstalk blocks (not shown here) only featured neutral distractors

Fig. 3

Changes in RT (pretest-posttest) per training group (a) separate for no-crosstalk and crosstalk, collapsed over all task conditions (b) for crosstalk, separate for each distractor condition and collapsed over all other task conditions. All RTs improved significantly at p < 0.005. Error bars denote SE. * p < 0.05

Fig. 4

Stimulus-locked ERPs. a Grand average CSD maps for N2 and P3b intervals. b ERPs at FCz, per Group and Block, for pretest (solid lines) and posttest (dashes lines), vertical lines indicate stimulus onset. c Average N2 peak amplitudes at FCz, per group and session. d N2 peak amplitude change (posttest–pretest) per group for crosstalk only. Error bars denote SE. * p ≤ 0.05

Fig. 5

Correct response-locked ERPs. a Grand average CSD maps for CRN interval. b ERPs at Fz, per group, for pretest (solid lines) and posttest (dashes lines), vertical lines indicate response onset. c Average CRN peak amplitudes, per group, session, block and distractor. d CRN peak amplitude change (posttest–pretest) per group and distractor, for crosstalk only. Error bars denote SE. * p ≤ 0.05

Fig. 6

Error response-locked ERPs. a Grand average CSD maps for Ne/ERN and Pe intervals. b ERPs at FCz, per group, for pretest (solid lines) and posttest (dashes lines), vertical lines indicate response onset. c Average Ne/ERN and Pe peak amplitudes, per group and session. d Pe peak amplitude change (posttest–pretest) per group for crosstalk only. Error bars denote SE. * p ≤ 0.05