Thinking when you think you’re not thinking

Dijksterhuis, Bos, Nordgren and van Baaren, R. B. (2006) investigated the effect of deliberation strategy on how successfully people chose the best second-hand car from a selection of 4 in a lab-based exercise. Participants saw a list of attributes of the cars, some positive, some negative, one at a time in random order. One of the cars had 75% positive attributes, two had 50%, and one had 25%. Some examples of attributes:

The Hatsdun has good mileage
With the Hatsdun it is difficult to shift gears
The Kaiwa has good mileage
The Kaiwa has poor handling
The Dasuka has poor mileage
The Dasuka is available in very few colors

The participants were randomly assigned to groups. Following the display of the features, members of one group were asked to “think about the cars” and members of the other group had to solve anagrams. After four minutes they had to choose the car they thought was best. The experiment was run for both “simple” decisions, where there were 4 attributes per car, and “complex” decisions, where there were 12 attributes per car.

For the simple condition, they found no difference between the two groups. For the complex condition they found that the group who solved anagrams were more likely to pick the best car (around 60% correct versus 25%).

The experiment was repeated, but this time participants were asked to rate the quality of each of the cars on an analogue scale, which was sampled and scaled to give a score from -25 to +25. The score they gave the worst car was subtracted from the score they gave the best car, for instance a subject who absolutely adored the best car and absolutely detested the worst car would get a score of 25-(-25) = 50. Someone who got confused and gave the opposite would get -50. Again they found an advantage of the unconscious strategy in the complex condition, with means of 5 versus under 1. Note that people didn’t seem to do terribly well, though the researchers did get a statistically significant difference between groups. It would be nice to get a look at the data in more detail.

This is intriguing for a number of reasons. I wonder if the distraction task could viewed as a dual-task, analogous to the sort of task used by memory researchers to fractionate various types of working memory. I’d like to know the effect of the task chosen on the accuracy of the decisions. Also intriguing is the effect of individual differences on the decisions, e.g. memory capacity, and also preferences for making decisions. I wonder if people could be trained to improve their conscious problem solving strategy.

The clinical relevance of this kind of thing has been investigated (independently—the literature doesn’t seem to intersect) by Donaldson and Lam (2004). They used a battery of questionnaires to assess mood and whether people ruminated when they felt down. They gave people a social problem solving task with questions like: what would you do if you discovered a friend was avoiding you? Their most interesting result is that both the trait of ruminating and the induction of rumination had a detrimental effect on social problem solving ability.

All interesting stuff!


Dijksterhuis, A., Bos, M. W., Nordgren, L. F. & van Baaren, R. B. (2006). On Making the Right Choice: The Deliberation-Without-Attention Effect. Science 311(5763), 1005-1007.

Donaldson, C. & Lam, D. (2004). Rumination, mood and social problem-solving in major depression. Psychological Medicine 34, 1309-1318.