Abstract The present study examined performance on Wason's abstract selection task. Specifically, it explored further our hypothesis that the difficulty of the task can be traced to two sources: selective encoding of the problem information, and inadequate analytic processing of the encoded information. The results of Experiment 1 replicated our earlier finding of good performance when an explicated rule with categorical elements is used along with violation selection instructions and a decision justification format. In addition, Experiment 1 demonstrated that the type of elements used in the rule (categories vs specific instances of categories) did not influence the rate of correct performance. Experiment 2 examined the effects of explicated rules that allow matching bias to operate. The rate of correct selections was lowered, but only significantly so for the categorical rules. Performance in both cases, however, was still well above baseline. The results are discussed in terms of our selective encoding...