SummaryThe purpose of this part of the research project on clinical inference in nursing was to attempt to identify what the message units of various cognitive tasks in nursing are. The effort failed.No single cue was found to convey more than a trivial amount of information to the nurse-subjects. Nor were groups of cues, arranged in various ways, found to be related to the inferences made by the nurse-subjects about SOP. And, finally, it was learned that these nurse-subjects did not consciously discriminate between the usefulness of various cues, nor did their confidence in their decisions vary over cases.The study contained various flaws: 1) the cases were not representative, in a strict sense, of cognitive tasks encountered by nurses; 2) the data were crudely presented in the cases, and were not in a form usually seen by the nurse; 3) the statistical analysis was inadequate; the number of cases (100) was not sufficiently large for each analysis made; and 4) the various cue-groupings employed were, to a large degree, arbitrary. It may well be that, if none of these deficiencies existed, the basic message units of cognitive tasks would have been discovered. It is doubtful, however, that this would have been the case, for the following reasons: 1) for the most part, the defects listed above would not have prevented the appearance of regularities between cues (and/or cue-groupings) and inferences if they were actually there to be discovered; 2) there were extremely wide individual differences in the six nurse-subjects in their inference patterns drawn from the same 100 cases—if basic message units do exist, the nurse-subjects employed here do not know of them, or do not agree on them; and 3) the nurse-subjects did not discriminate among the usefulness of cues. If basic message units are common knowledge, the nurse-subjects did not employ them in a consistent fashion.Thus, the results of the present study are not conclusive. They do indicate, however, that questions concerning the message units involved in cognitive tasks encountered by the nurse can be investigated. The question of what information units are used by the nurse should be answered.