Card Sorting Test (WCST-style) Online — Free Cognitive Flexibility Task

In the card sorting test you match each presented card to one of four key cards laid out at the top of the screen. You are never told the matching rule; you only get correct or wrong feedback after every choice. The rule is one of color, shape, or number, and after ten correct in a row it silently changes to the next one. Early on, the task reveals your ability to infer a rule; later, it reveals cognitive flexibility (set shifting): letting go of a rule that was working the moment it starts failing and switching to a new one.

The paradigm originates in the card sorting task published by Berg and Grant in 1948, and became an established executive-function measure after Milner linked frontal lobe damage to perseverative errors. Soundary's version is our own web implementation of that principle; it does not use the commercial WCST kit or its scoring norms and standard scores. The key cards are also deliberately different from the original red triangle, green star, yellow cross, and blue circle: here they are one purple square, two sky-blue hexagons, three orange diamonds, and four pink half-moons.

The rule cycles color, then shape, then number, then back to color. The test ends when you complete six categories or reach the 64th trial, and usually takes five to ten minutes. The 24 response cards are constructed so that no card shares two attributes with any key card at once, so every choice records unambiguously which rule you applied.

The headline result is the number of categories completed (0 to 6). Each time you pass a rule with ten correct in a row, you gain one, and the total trial count and error count are shown beneath it. Five or more indicates smooth rule discovery and switching, three or four is intermediate, and two or fewer suggests finding or changing rules took a while.

The key index in the detailed results is perseverative errors: errors made after a rule change that would have been correct under the immediately preceding rule, shown as a count and as a percentage of all trials. It captures how strongly you stay stuck on what worked a moment ago. All other errors are counted separately as non-perseverative. Conceptual-level responses are the share of responses that fall within runs of three or more consecutive correct answers, reflecting how much you applied the rule with understanding rather than by chance. Failure to maintain set counts the times a run of five or more correct was broken by an error before reaching ten, moments where you knew the rule but let it slip.

The interpretive text reads five or more categories with a perseverative error rate of 15% or less as flexible shifting, three or more categories as intermediate, and anything else as having had difficulty shifting; conceptual-level responses are read as high at 70% or more, intermediate at 40% or more, and low below that. The composite score kept for trend tracking is a 0 to 100 value that weights the proportion of categories completed at 70 points and the conceptual-level response rate at 30 points. All of these boundaries are reference values chosen by Soundary and bear no relation to the standard scores or percentiles of the commercial WCST.

Because the heart of this task is solving it without knowing the rules, one attempt teaches you the three criteria and how they switch, so results improve markedly from the second time on. The practice effect is far larger than in most cognitive tasks, so leave at least several weeks between retakes and treat the first attempt as your reference point rather than comparing session to session. Fatigue, sleep loss, and distraction increase failures to maintain set, where you know the rule but slip, while hasty tapping increases non-perseverative errors.

Read the number of categories together with perseverative errors. Few categories with a low perseverative rate means the early rule inference was slow; a high perseverative rate means you could not let go of a method you already knew. Comparing this with everyday life, whether changing plans feels hard or you tend to stick to one approach, supports self-understanding. Alongside the Stroop and Go/No-Go tasks for inhibitory control and a working-memory task, it sketches which parts of executive function are relatively strong or weak.

Card sorting is a flagship executive-function measure in clinical neuropsychological assessment, but this web version is not that test and makes no diagnosis. Regardless of your result, if planning, reordering tasks, or noticing your own mistakes has become noticeably harder over the past few months, we recommend a formal neuropsychological assessment through a psychiatrist or neurologist.