A card-sorting task in which a test figure is matched to one of two target figures by shape or by colour. After practice and a block on one dimension, the rule switches; a mixed block then cues the dimension on every trial. Measures cognitive flexibility.
Configure parameters and run an interactive preview exactly as participants will experience it. No data is recorded.
Adjust parameters below, then start the preview on the right.
Task parameters
Practice: 4 trials on the first dimension, then 4 on the second, with feedback; 3 of 4 correct pass, with up to 3 attempts per dimension. Then 5 pre-switch trials on the second dimension and 5 post-switch trials on the first, each needing 4 of 5 correct. A participant who does not meet one of these criteria is shown a neutral end screen, and the result records where the task stopped (discontinued). Then the mixed block, in a pseudorandom order with 2–5 dominant trials before each non-dominant trial: 68 trials, about 5 min. The dominant dimension is the post-switch dimension, i.e. the first one practised.
Sources (Zelazo et al., 2014, the adult NIH Toolbox validation): practice of 4 trials per dimension with 3 correct to pass and as many as 3 chances, the task stopping if a dimension is not passed; 5 pre-switch and 5 post-switch trials with 4 of 5 correct to go on; 50 mixed trials (40 dominant + 10 non-dominant) with 2–5 dominant trials before each non-dominant one. Zelazo et al. (2014) scored only the first 30 mixed trials, and the released NIH Toolbox gives ages 8 and up 30 mixed trials, hence the shorter option. The first dimension is random per participant, the per-device form of the counterbalancing in Zelazo et al. (2014).
Trial timing
Each trial, as in Fig. 1 of Zelazo et al. (2014): a blank of 800 ms, a fixation cross for 1000–1500 ms, the cue word (SHAPE or COLOUR) alone for 1000 ms, then the test figure and the two targets, without the cue word, for up to 10,000 ms; a trial without a response by then is recorded as an omission. The figure gives the fixation as a range; here each trial draws its duration uniformly from it (our reading). The response window can be shortened (1000–10000 ms); there is no unlimited option, because no published version leaves the test figure on screen without a limit.
Practice feedback duration: no published value. Zelazo et al. (2014) describe a feedback screen after each practice response but not how long it stays; 1000 ms is our choice.
Two targets stay at the bottom: a blue circle (left) and an orange square (right). The test figure, an orange circle or a blue square, matches one target by colour and the other by shape, so the two rules always point to different sides. Blue and orange are distinguishable with red–green colour deficiency. The cue word is shown only; the NIH Toolbox also speaks it (a recording on each trial in the child version, Zelazo et al., 2013; instructions read aloud in the adult study). The test figure onset is taken after the screen has been drawn and responses at touch-down, but this timing has not yet been validated on real devices.
Outputs: accuracy in the pre-switch, post-switch and mixed blocks; perseverative errors (post-switch responses by the pre-switch rule); mixed-block accuracy and median RT of correct responses for dominant and non-dominant trials; switch costs in the mixed block (median RT and accuracy, switch − repeat trials); where the task stopped, if it did. As in Zelazo et al. (2014), RTs under 100 ms or more than 3 SD from the participant’s mean RT in the test blocks are left out of these RT summaries; the number left out is reported. The generic accuracy and median RT exclude practice; the generic median RT is not trimmed.
This is a researcher preview. No participant data is recorded.
Simulated participant view
A card-sorting task in which a test figure is matched to one of two target figures by shape or by colour. After practice and a block on one dimension, the rule switches; a mixed block then cues the dimension on every trial. Measures cognitive flexibility.
No data is recorded
Two target figures stay at the bottom of the screen as large touch targets: a blue circle and an orange square. On each trial a fixation cross appears, then the cue word SHAPE or COLOUR alone, then the test figure without the cue word (an orange circle or a blue square) that matches one target by colour and the other by shape, as in the published trial sequence. The participant taps the matching target. A screen before each block gives its rule.
Useful when a study needs the established DCCS procedure of the NIH Toolbox: rule switching between two sorting dimensions, perseveration after a rule switch, and the cost of switching in a mixed block where one dimension is dominant. Suited to baseline and occasional sessions in adults.
Not a diagnostic instrument and not a source of NIH Toolbox norms: the task is built from the published procedure with generic figures, and the NIH Toolbox computed score is not reported. Not a quick EMA probe. For a task with a free switch ratio and no blocks, use Task Switching.
Keep the standard mixed block (50 trials: 40 dominant + 10 non-dominant) and the random first dimension, which counterbalances which dimension becomes dominant, as in Zelazo et al. (2014). When time is tight, run only the first 30 mixed trials: Zelazo et al. (2014) scored these 30, and the released NIH Toolbox gives ages 8 and up 30 mixed trials. Compare post-switch accuracy with pre-switch accuracy for perseveration, and the mixed-block measures for switching.
As in Zelazo et al. (2014), practice has 4 trials on the first dimension and 4 on the second, with feedback; 3 of 4 correct pass, with up to 3 attempts per dimension. Then 5 pre-switch trials on the second dimension, 5 post-switch trials on the first, and the mixed block, in which the post-switch dimension is dominant. A participant who does not pass practice in 3 attempts, or gets fewer than 4 of 5 right in the pre- or post-switch block, sees a neutral end screen, and the task stops there, as in the published procedure; the result records where. RTs under 100 ms or more than 3 SD from the participant's mean are left out of the RT summaries, as in Zelazo et al. (2014). Low post-switch accuracy with perseverative errors means the old rule persisted. In the mixed block, slower and less accurate responses on non-dominant and switch trials reflect the cost of shifting away from the dominant rule; the first mixed trial follows a break and counts as neither switch nor repeat. Every test figure conflicts (it matches one target by shape and the other by colour), so a wrong response always follows the other rule. Adults are often near ceiling in accuracy, so the RT measures carry most of the variance. Differences from the published procedure: generic figures instead of the NIH Toolbox pictures; the cue word is shown only, whereas the NIH Toolbox child version also plays it from a recording on each trial (Zelazo et al., 2013) and in the adult study the instructions were read aloud; the duration of the practice feedback is not published. Responses are timed from the drawn test figure to the touch-down, but this timing has not yet been validated on real devices.
Add Dimensional Change Card Sort (DCCS) to your ESM study on SMAAT — free to get started.