Computational protocol: The early effects of external and internal strategies on working memory updating training

Similar protocols

Protocol publication

[…] The background characteristics of the three groups at pretest were compared with one-way ANOVAs for continuous variables and χ2 tests for categorical variables. Independent samples t-tests were employed to evaluate potential differences in motivation and alertness during the intervention period between the two training groups.As regards the pre-post analyses, ANCOVAs with posttest performance as the dependent variable, pretest performance as the covariate, and group as the between-subjects factor were conducted for each pre/post task. Significant main effects of group were followed up with planned contrasts. The first contrast compared the strategy training group to the active control group (−1, 1, 0), and the other contrast compared the strategy training group to the passive control group (−1, 0, 1). (Note that unlike in the pre-registration where we stated that post hoc tests will be used, we instead chose to employ planned contrasts. We considered a priori planned contrasts to be more adequate here, after all: we had a clear hypothesis on the strategy training effects. In practice, post hoc contrasts would yield the same results). To correct for multiple comparisons, Benjamini-Hochberg adjusted p-values were applied for group comparisons on each pre-post outcome measure.To examine whether the present sample had enough power to detect significant group (between-subjects) × time (within-subjects) interactions if present in the pre- and posttest measures, we conducted a post-hoc power analysis using G*Power version 3. The power to detect a large (f = 0.40) or medium effect (f = 0.25) was > 0.99 based on the sample size and the use of the within-subjects correlation of r = 0.68 (which was the largest observed correlation across all pre-and posttest variables). We also ran a power analysis with the smallest correlation among the pre- and posttest variable (r = 0.32). In this case, the power to detect a large or medium effect was 0.98. For ANCOVA analyses, the power was 0.97 for detecting a large effect size (f = 0.40) and 0.66 for a medium effect size (f = 0.25). Earlier studies with WM strategy intervention have not explicitly reported results from power analyses–. Nevertheless, the two previous studies reporting effect sizes have shown large effects (ηp2 range 0.18–0.27) with groups ranging from 27 to 79 participants,. Thus, the sample size of the present study should be sufficient to achieve adequate statistical power for our strategy intervention.The ANCOVA analyses were conducted using IBM SPSS Statistics 23 software. Planned contrasts, figures, and inter-rater reliability analyses (when evaluating self-reported strategies) were performed in the R environment version 3.3.3, using the packages ggplot2, and rel. Prior to the data collection, the study protocol was preregistered at AsPredicted.org (https://aspredicted.org/b8rp8.pdf). […]

Pipeline specifications

Software tools G*Power, SPSS, Ggplot2
Application Miscellaneous
Organisms Homo sapiens