I've reviewed the papers by Grant, D. A., & Berg, E. (1948) and Fey, E. T. (1951).

These papers from what I was able to find, lay out the basis of the Wisconsin Card Sorting Test (WCST). There are three possible rules for sorting the cards; color, shape, or count. I am curious why the WCST places 4 stimulus cards in front of the patient when 3 carefully chosen cards would be sufficient. With 4 cards it seems that unless the examiner makes special preparations, there might be occasions when there is more than 1 correct stimulus card to choose, or during a transition from certain rule to the next the patient may make an incorrect guess and won't be able to tell which rule to eliminate.

Can anyone explain why there has to be 4 stimulus cards? Could I run an equally effective WCST with 3 stimulus cards?

Grant, D. A., & Berg, E. (1948). A behavioral analysis of degree of reinforcement and ease of shifting to new responses in a Weigl-type card-sorting problem. Journal of experimental psychology, 38(4), 404.
Fey, E. T. (1951). The performance of young schizophrenics and young normals on the Wisconsin Card Sorting Test. Journal of consulting psychology, 15(4), 311.


2 Answers 2


I cannot tell you why they chose four cards, but I do know that it is perfectly fine to use three, (e.g. Niv and colleagues, 2015).

To give you more info on why the WCST is used in the study by Niv. The study wanted to see how you could best model learning (and relearning) multidimensional information, to be able to better understand how people make sense of the multidimensional information provided by the real world. Therefore, they had participants perform the WCST with three cards. Multiple reinforcement learning algorithms were then fitted to the behavioral data, and they found that a reinforcement learning model that includes decay (as to model memory decay) was best able to do so.

I believe it would have been perfectly fine to have used four cards in that study as well. It would make it more difficult to learn on which dimension to focus, which would probably also be modeled.

There are also variations of the WCST (e.g. Catherwood and colleagues, 2014). In that study they instead of took apart four dimensions and showed a 2x2 matrix with each unit depicting a different dimension.

In short

Depending on the goal of your experiment, you can decide how you would like to use the WCST specifically.

  • $\begingroup$ +1 The original task most likely had 4 cards because there are - as you mention - 4 (2x2) scenarios. $\endgroup$
    – Arnon Weinberg
    Jul 25, 2016 at 22:43
  • $\begingroup$ Thanks, I thought it would be ok to use 3. Just to be certain, I have requested the original 1948 paper on the WCST and once I am able to review it I'll report back. $\endgroup$ Jul 26, 2016 at 13:17

I wanted to report further findings in my search for the answer to this question. I will not accept this as the answer but I think anyone else who looks into this issue may find this information useful.

I obtained the original paper by Berg (1) published in 1948 in the journal of general psychology. The paper makes no explicit mention of why 4 stimulus cards are shown. Perhaps the choice was to follow previous methods in similar tests such as the Goldstein-Sheerer cube test (2).

The results of the Berg paper, a post-hoc style analysis, grouped the participants into 3 groups based on quantitative performance (A,B, and C). Group A quickly determined the rules of the test and performed well on the remainder of the test. Group B took longer and verbally reported trying many hypotheses of the rules before discovering there were 3 rules to choose from. Group C had no idea what was going on, some subjects perseverated 100 trials with their own incorrect rule.

I therefore speculate that if there were only 3 stimulus cards that group B would be somewhat eliminated. Subjects who could determine the rules would get there faster by elimination since there could only be 1 correct stimulus card at a time. One virtue then, of having 4 cards, is to parse out subjects who have a more keen ability to "think/shift" which could be more useful/powerful if the goal of the test is to quantify shifting ability.

Furthermore, I took the online test located here and found 18 out of 60 trials where there was more than 1 matching possibility yet I was able to succeed at the test because I was able to track which rule I had been using.

I agree with Robin Kramer that the number of stimulus cards can be altered depending on the purpose of the test.

  1. "A simple objective technique for measuring flexibility in thinking." Berg, A, 1948, Journal of General Psychology, Vol 39, 15-22.

  2. "Abstract and concrete behavior: An experimental study with special tests." Goldstein, K., Scheerer, M., 1941, Psychology Monologues, Vol 53, 1-151

  • $\begingroup$ Thank you for sharing this! There is no shame in accepting your own answer. you were the one finding the original paper ;) $\endgroup$ Jul 27, 2016 at 20:48
  • $\begingroup$ I'll wait a little longer to see if anyone else has input then I'll accept answer $\endgroup$ Jul 28, 2016 at 14:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.