**Methods and Materials**
*Content*
- *Kool replication UiT - Script English.pdf:* A script of the instructions given to the participants, translated to English. Please refer to the procedure video component/video as well.
- *Kool replication UiT - Handout instructions Norwegian.pdf:* A small hand-out of the task instructions given to the participants (please see the Procedure video wiki).
- *Kool replication UiT - Handout instructions English.pdf:* A small hand-out of the task instructions given to the participants, translated to English.
- *Kool replication UiT - Debrief questions Norwegian.pdf:* Debrief questions given to the participants after the experiment.
- *Kool replication UiT - Debrief questions English.pdf:* Debrief questions given to the participants after the experiment, translated to English.
- *MathLab task files:* A folder containing the relevant information to run the computer experiment. It contains files used in MathLab (with [Psychtoolbox][3]) to run the experiment. These are the same as in the original experiment, with two small changes in the code. So please use the *deck_randposHome.m* file directly under the MathLab task files folder, and not the one in the dst_multidecks.zip file/folder. For further instructions see *Kool replication UiT - MathLab researcher instructions.pdf*.
*Changes in the MathLab code*
The code given by original researchers is outdated in 2 areas:
- Change line 57 to: RandStream.set**Global**Stream(RandStream('mt19937ar','seed',randSeed)); (Underline and bold added, underline not possible to show here). The only word changed here is “Global”.
- Make sure priority on line 104 is 0 instead of 7. Linux uses a 7 priority, Windows cannot. Keep it / change it back to 7 if you are using OSX (Mac).
*Recorded data*
For each of the participants, the following will be recorded:
- Number of trials
- Overall ratio of Low Demand versus High Demand
- Ratio of Low Demand versus High Demand in trial block 1
- Ratio of Low Demand versus High Demand in trial block 2
- Ratio of Low Demand versus High Demand in trial block 3
- Ratio of Low Demand versus High Demand in trial block 4
- Ratio of Low Demand versus High Demand in trial block 5
- Ratio of Low Demand versus High Demand in trial block 6
- Ratio of Low Demand versus High Demand in trial block 7
- Ratio of Low Demand versus High Demand in trial block 8
- Number of Low Demand choice resulting in repeated task
- Number of Low Demand choice resulting in switched task
- Number of High Demand choice resulting in repeated task
- Number of High Demand choice resulting in switched task
- Total accuracy (correct responses in all trials)
- Accuracy when the Low Demand choice resulted in a repeated task
- Accuracy when the Low Demand choice resulted in a switched task
- Accuracy when the High Demand choice resulted in a repeated task
- Accuracy when the High Demand choice resulted in a switched task
- Reaction time overall
- Reaction time when the Low Demand choice resulted in a repeated task
- Reaction time when the Low Demand choice resulted in a switched task
- Reaction time when the High Demand choice resulted in a repeated task
- Reaction time when the High Demand choice resulted in a switched task
*Predictions / Analysis*
Individuals minimizing cognitive demand would be expected to show a consistent bias toward the low-demand alternative across runs.
To validate the task-switching manipulation, meaning to check that a switch is more effortful compered to a repeated task, we compare reaction times via a two-way repeated measures analysis of variance (ANOVA) with factors for trial type (repetition vs. switch) and circle choice (high demand or low demand). Error rates for the high- and low-demand circles will be compared in a Wilcoxon signed-rank test. To test for circle preference, we test the low-demand selection rates for individual subjects against the chance rate of 0.50 in a Wilcoxon signed-rank test.
We will also compute each subject’s proportion of low-demand selections across all trials. Internal consistency will be assessed by calculating Cronbach’s , treating the eight runs of the DST as subtests.
[3]: http://psychtoolbox.org
