Table 1.
Explanation of the design principles from the Van Rijt et al. [56] study, adopted in the current study.
Fig 1.
General overview of switching replications design.
Group 2 thus serves as a control to group 1. When the control group later receives the treatment, the original treatment group then serves as a continued-treatment control [67].
Table 2.
Example of a question (see [29]) with the different multiple-choice options (translated from Dutch by the authors).
Table 3.
Model comparison for the three core variables: Understanding score, blind score and no concept score.
Table 4.
Mean and 95% Highest Density Interval (HDI) for the posterior distributions of the parameter estimates based on the best fitting models for understanding score (Model 5), blind score (Model 1) and no concept score (Model 6).
Fig 2.
Median, 66% and 95% CI for the posterior distribution of means for both groups (Group 1 = intervention between Occ1 & Occ2; Group 2 = intervention between Occ2 & Occ3) on Understanding score (A) and No Concept score (B). Blue intervals indicate means after implementation of the intervention; red intervals indicate means before the implementation of the intervention.
Fig 3.
Posterior distribution for the absolute effect sizes (Cohen’s d) of the intervention for Understanding scores (A) and No Concept scores (B) with the median and 66% and 95% CI (Effect size labels according to Calin-Jageman and Cumming [85]). Group 1 = intervention between Occ1 & Occ2; Group 2 = intervention between Occ2 & Occ3.
Table 5.
Axial codes and subcodes related to key assigment 1 (N = 196 students; N = 146 performed and readable tasks).
Table 6.
Axial codes and subcodes related to key assignment 2 (N = 196, N = 142 performed and readable tasks).
Table 7.
Axial codes and subcodes related to key assignment 3 (N = 196, T1: N = 123; T2: N = 121 performed and readable tasks).