Irrelevant Thoughts And Attentional Inertia

In the Dimensional Change Card sorting (DCCS) task, 3-year-olds can usually sort cards successfully by a first rule - whether by shape, color, size, etc. When asked to switch then to another rule, most 3-year-olds will perseverate by continuing to sort cards according to the first and now-irrelevant rule. This occurs even when the current rule is repeated every single time they're asked to sort a card! Children will even correctly repeat the name of the rule they should be using, and then proceed to actually sort the card by the old rule.

By age 4, however, most kids are able to successfully switch to a second rule (though many will still have trouble when asked to switch again). What then changes between 3 and 4 to allow this shift away from a remarkably strange behavior?

A new study by Wolfgang Mack begins to answer this question. In a first experiment, 76 3 year old children completed the DCCS, but 35 received a modified version of the task: for those children in the experimental group, the "target cards" (essentially, labels to differentiate the sorting locations) were removed for 3 seconds before being reaffixed in between when children were asked to sort by the first and the second rules. Children in the experimental performed significantly better than those in the control group - as though reaffixing the target cards helped punctuate the shift from an old rule to a new rule.

A second experiment with 107 3-year-olds involved a slightly different manipulation: instead of reaffixing the target cards in between "pre-switch" and "post-switch", kids in the experimental group were simply asked what they had for breakfast that morning. In another experimental condition, children were explicitly trained on sorting cards by both color and shape before beginning the DCCS. Again, children in both experimental groups did significantly better than those performing the standard DCCS task (although the significance of the effect was noticeably lower than in the first experiment, despite the larger sample size).

Mack concludes that the verbal information conveyed by the experimenters at the end of the first sorting rule and at the beginning of the second is merely insufficient for many 3 year olds to understand what they should be doing. Accordingly, providing pre-training may help through this route.

Clearly, language is important - but why then should some four year olds still have difficulty switching back to the first rule, or difficulty in the border version of DCCS (where the sorting rule is determined by whether the cards have a thick or thin border) if they are able to successfully switch earlier in the game? Such findings complicate any theory which puts verbal misunderstanding at center stage in the phenomenon of perseveration.

Mack showed that switching performance can be improved simply by asking an irrelevant question, but provides no explanation for this fascinating finding. One idea, based on Kirkham et al's attentional inertia account of perseveration, is that children are merely fixated on the first rule and need some sort of strong external prompt to think of something else. Yet, from this perspective, it seems strange that removing the target cards should have an even stronger effect (in which children might be expected to remain focused on the preswitch dimension, at least more so than when considering their breakfast).

Instead, one might endorse a "contextual change" account in which the experience of sorting cards according to the first rule is exerting strong proactive interference on children's ability to maintain the first rule. Any change in context, particularly those which were most strongly paired with the previous sorting experience (for example, the target cards), may release subjects from this proactive interference - even if those contextual changes are only momentary. According to this perspective, pretraining may work because it dilutes the association between the environmental context and the preswitch rule, or it may improve performance based on some other mechanism.