The effect of feedback type and task complexity on the acquisition of procedural knowledge in an expert system setting using an audit task

This study examines the effect three different feedback types have on the acquisition of procedural knowledge controlling for task complexity. It addresses main research questions: (1) does expert system feedback type affect the acquisition of procedural knowledge; (2) does task complexity affect the acquisition of procedural knowledge; and (3) do expert system feedback type and task complexity interact to affect the acquisition of procedural knowledge.;There is evidence in the accounting and psychology literature that task complexity affects judgment outcome (Bonner 1994; Abdolmohammadi and Wright 1987). However, it is not known to what extent task complexity affects the acquisition of procedural knowledge and to what extent task complexity and expert system feedback type interact.;A total of 129 student subjects participated in a five-stage laboratory experiment. The experiment consisted of a 3 x 2 between-subjects design, where expert system feedback type varied three ways and task complexity two ways. Subjects were randomly assigned to the treatment cells.;The results indicate that subjects in the expert system groups acquired significantly greater procedural knowledge than subjects in the non-expert system groups. However, the type of feedback did not significantly affect acquisition of procedural knowledge, when task complexity was not controlled. Finally, this study notes that task complexity does interact with feedback type.;This study contributes to the research in five ways. First, it controlled for task complexity ex ante using Bonner's model of task complexity. Second, this study finds that task complexity affects acquisition of procedural knowledge only in the presence of an expert system. Third, this study was able to partially replicate Eining and Dorr's (1991) finding that expert system use does assist subjects in acquiring a significantly greater amount of procedural knowledge. Fourth, this study notes that in the absence of task complexity, feedback does not significantly affect the acquisition of procedural knowledge. Finally, it finds there is an interaction between task complexity and expert system feedback type. When the task is complex, the feedback type does not matter, but when the task is simple, subjects who received the detailed-text-with-example feedback acquired a significantly greater amount of procedural knowledge.