Data analysis

For all pairwise comparisons (i.e., comparisons between two within- or between-subjects conditions), we used dependent and independent t-tests. For within-subjects comparisons that involved more than two conditions and/or more than one factor, we used linear mixed models (LMMs). We chose LMMs instead of within-subjects ANOVAs because LMMs allow for a more precise modeling of the data (via the inclusion of random intercepts and slopes for each participant) and because, with LMMs, it is possible to include non-categorical factors (e.g., time course). For between-subjects comparisons that involved more than two conditions and/or more than one factor, we used a multiple linear regression (MLR). We chose a MLR instead of a between-subjects ANOVA because we aimed to perform a step-wise regression to assess the variance explained by our questionnaire scales on top of the variance explained by the experimental factors. For comparing frequencies, we used a standard Chi-squared test.

We will first report the results for Experiments 1 and 2, focusing on the knowledge manipulation (computer capacity known/unknown), see Fig. 3. Next, we will report an analysis that combines data from Experiments 1 and 2 with the two experiments from our previous study (Wahn et al., 2023). This analysis gives us a 2 × 2 between-study design with the between-study factor “Bonus task” (with, without), with a bonus task present in the current but absent in our previous study, and the within-study factor “Computer capacity” (known, unknown); see Fig. 6. Finally, we will report the results for Experiment 3, focusing on the incentive manipulation (bonus task with/without incentive). To do so, we compare Experiment 3 (bonus task without incentive) to Experiment 2 (bonus task with incentive) and to Experiment 2 from our previous study (no bonus task); see Fig. 7.

Offloading percentage for Experiment 1 (left) and Experiment 2 (right). Errors bars are Standard Error of the Mean

Offloading percentage, displayed as a function of Bonus task (with, without) and Computer capacity (known, unknown). Note that the data in the left panel (“with bonus task”) represent Experiment 1 (Unknown) and 2 (Known) from the present study (compare Fig. 3), whereas data in the right panel (“without bonus task”) are taken from Experiments 1 and 2 from our own previous study (Wahn et al., 2023). Errors bars are Standard Error of the Mean

Offloading percentage as a function of monetary incentive and bonus task availability. Errors bars are Standard Error of the Mean