Figure 1
Trials Timelines for all Experiments
Notes. The pair of trials at the top left illustrates Exp. 1 procedure, where the stimuli were horizontally aligned, the context was either a red or a blue box, the SOA increased in 66 ms steps, and no feedback was provided. In this example, a task repetition (a letter task, in this example) is performed in the first trial. Hence, in the next trial, the SOA increases from, for example, the minimum SOA (66 ms) to the next SOA (132 ms). In this second trial, the example illustrates a wrong response to the switch target (the digit, in this example). The top right trial pair illustrates Exp. 2 procedure, where the stimuli and context were as in Exp. 1, but the SOA increased in 50 ms steps, and feedback was provided for erroneous responses. In this example, a task switch is performed in the first trial, having a hypothetical SOA of 50 ms. Hence, in the next trial, the SOA is reset to the minimum SOA (50 ms). In this second trial, the example illustrates a wrong response to the repetition target. The bottom left trial pair illustrates Exp. 3 procedure, where stimuli were vertically aligned, the context was one of 8 images (only 2 are visible here), and feedback was provided as in Exp. 2. In this example, a task switch is performed in the first trial, having a hypothetical SOA of 100 ms. Hence, in the next trial, the SOA is reset to the minimum SOA (50 ms). In this second trial, the example illustrates a wrong response to the repetition target. The bottom right illustration shows one of the possible key-response mappings, where the vowel/consonant task is performed with the left hand and the parity task with the right hand. Note that, in Exp. 1 and 2, this mapping could only be used when the letter was on the left of the digit (as in this example). When the digit was on the left of the letter, the parity task was performed with the left hand, and the letter task with the right hand.
Figure 2
Mean Task-repetition rate based on Context Relation in all Experiments
Notes. Rep. = Repetition. The upper panel depicts the mean task-repetition rate in context repetitions versus switches in the three experiments. Vertical bars represent 95% confidence intervals of the difference of the means. The lower panel depicts violin plots as well as individual participants’ means of the effect of interest, namely task-repetition rates in context repetitions minus task-repetition rates in context switches.
Figure 3
Mean RTs and Error Rates based on Context Relation in all Experiments.
Notes. The vertical bars represent 95% confidence intervals of each raw mean.
Table 1
Mixed Models on Reaction Times and Errors of Experiment 1.
| FIXED EFFECT | REACTION TIMES | ERROR | |||||
|---|---|---|---|---|---|---|---|
| ESTIMATE | t value | p | ESTIMATE | z value | p | ODD RATIO | |
| (Intercept) | 632.019 | 46.889 | <.001 | –2.798 | –26.865 | <.001 | 0.061 |
| Task Rep. | 22.161 | 9.625 | <.001 | 0.265 | 6.3 | <.001 | 1.304 |
| Context Rep. | 0.979 | 0.422 | 0.673 | 0.019 | 0.521 | 0.602 | 1.019 |
| SOA | 0.034 | 3.71 | <.001 | 0.074 | 2.111 | <.05 | 1.077 |
| Response Rep. | –2.053 | –0.811 | 0.417 | –0.303 | –6.572 | <.001 | 0.738 |
| Task Rep.:Context Rep. | 1.171 | 0.545 | 0.586 | 0.05 | 1.225 | 0.22 | 1.051 |
| Task Rep.:SOA | 0.083 | 9.726 | <.001 | 0.109 | 2.682 | <.01 | 1.115 |
| Context Rep.:SOA | 0.014 | 1.635 | 0.102 | 0.004 | 0.116 | 0.908 | 1.004 |
| Context Rep.:Response Rep. | –6.072 | –2.403 | 0.016 | –0.027 | –0.593 | 0.553 | 0.973 |
| SOA:Response Rep. | –0.001 | –0.114 | 0.909 | –0.075 | –2.03 | <.05 | 0.928 |
| Task Rep.:Context Rep.:SOA | 0.003 | 0.396 | 0.692 | –0.047 | –1.182 | 0.237 | 0.954 |
| Context Rep.:SOA:Response Rep. | 0.006 | 0.797 | 0.425 | 0.023 | 0.62 | 0.536 | 1.023 |
[i] Note. Rep. = Repetition. The variables are sum-coded such that Switch = 1 and Repetition = –1. SOA values were standardized. The model formula was Task_rep*Context_rep*SOA + Response_rep*Context_rep*SOA + (1|pp).
Table 2
Mixed Models on Reaction Times and Errors of Experiment 2.
| FIXED EFFECT | REACTION TIMES | ERROR | |||||
|---|---|---|---|---|---|---|---|
| ESTIMATE | t value | p | ESTIMATE | z value | p | ODD RATIO | |
| (Intercept) | 625.625 | 39.042 | <.001 | –3.126 | –37.272 | <.001 | 0.044 |
| Task Rep. | 21.538 | 10.259 | <.001 | 0.081 | 1.775 | 0.076 | 1.084 |
| Context Rep. | 0.569 | 0.279 | 0.78 | –0.042 | –0.987 | 0.324 | 0.959 |
| SOA | 0.051 | 5.676 | <.001 | –0.006 | –0.105 | 0.917 | 0.994 |
| Response Rep. | 6.396 | 3.061 | 0.002 | –0.216 | –4.915 | <.001 | 0.806 |
| Task Rep.:Context Rep. | 0.929 | 0.489 | 0.625 | 0.026 | 0.589 | 0.556 | 1.026 |
| Task Rep.:SOA | 0.104 | 12.206 | <.001 | –0.058 | –1.071 | 0.284 | 0.944 |
| Context Rep.:SOA | 0.009 | 1.052 | 0.293 | –0.062 | –1.194 | 0.232 | 0.94 |
| Context Rep.:Response Rep. | –3.397 | –1.628 | 0.103 | 0.003 | 0.066 | 0.947 | 1.003 |
| SOA:Response Rep. | –0.011 | –1.695 | 0.09 | 0.021 | 0.561 | 0.575 | 1.021 |
| Task Rep.:Context Rep.:SOA | 0.01 | 1.174 | 0.24 | –0.035 | –0.67 | 0.503 | 0.966 |
| Context Rep.:SOA:Response Rep. | 0.005 | 0.81 | 0.418 | –0.006 | –0.161 | 0.872 | 0.994 |
[i] Note. Rep. = Repetition. The variables are sum-coded such that Switch = 1 and Repetition = –1. SOA values were standardized. The model formula was Task_rep*Context_rep*SOA + Response_rep*Context_rep*SOA + (1|pp), and it was fitted with the bobyqa optimizer as opposed to the default, which did not converge (nloptwrap).
Table 3
Mixed Models on Reaction Times and Errors of Experiment 3.
| FIXED EFFECT | REACTION TIMES | ERROR | |||||
|---|---|---|---|---|---|---|---|
| ESTIMATE | t value | p | ESTIMATE | z value | p | ODD RATIO | |
| (Intercept) | 699.921 | 30.37 | <.001 | –2.955 | –33.724 | <.001 | 0.052 |
| Task Rep. | 37.655 | 14.279 | <.001 | 0.202 | 5.007 | <.001 | 1.224 |
| Context Rep. | 2.693 | 1.08 | 0.28 | –0.044 | –1.155 | 0.248 | 0.957 |
| SOA | –0.014 | –1.37 | 0.171 | 0.009 | 0.184 | 0.854 | 1.009 |
| Response Rep. | 11.459 | 4.934 | <.001 | –0.058 | –1.464 | 0.143 | 0.944 |
| Task Rep.:Context Rep. | –1.363 | –0.543 | 0.587 | 0.009 | 0.24 | 0.811 | 1.009 |
| Task Rep.:SOA | 0.108 | 10.679 | <.001 | 0.005 | 0.108 | 0.914 | 1.005 |
| Context Rep.:SOA | 0.012 | 1.193 | 0.233 | –0.08 | –1.79 | 0.073 | 0.923 |
| Context Rep.:Response Rep. | –0.54 | –0.233 | 0.816 | 0.027 | 0.689 | 0.491 | 1.027 |
| SOA:Response Rep. | –0.007 | –1.076 | 0.282 | –0.062 | –1.815 | 0.07 | 0.94 |
| Task Rep.:Context Rep.:SOA | 0.019 | 1.961 | 0.05 | –0.05 | –1.054 | 0.292 | 0.951 |
| Context Rep.:SOA:Response Rep. | –0.01 | –1.453 | 0.146 | 0.045 | 1.309 | 0.191 | 1.046 |
[i] Note. Rep. = Repetition. The variables are sum-coded such that Switch = 1 and Repetition = –1. SOA values were standardized. The model formula was Task_rep*Context_rep*SOA + Response_rep*Context_rep*SOA + (1|pp).