This study investigated the effects of augmenting and deleting elaborations in an existing self-instructional text for a micro-computer database application, "Microsoft Works User's Manual." A total of 60 undergraduate students were randomly assigned to the original, elaborated, or unelaborated text versions. The elaborated version included: (1) reorganizing content by separating conceptual and procedural information; (2) adding examples, restatements, explanations, an introduction and summary; (3) adding text-to-graphic references to support both conceptual and procedural information; (4) adding several examples of using different combinations of connectors to link selection rules; and (5) removing side-margin headings. Changes in the unelaborated version included: (1) identifying statements of procures and rules; (2) deleting all elaboration such as examples, restatements, and descriptions of what the software does as a result of procedures; (3) deleting unnecessary words; (4) deleting introductory text; and (5) deleting side-margin subheadings. Students completed six tasks using one of the three instructional texts, and then completed a six-item attitude survey. The absence of a significant difference in the task performance under the three text versions suggests that not all people require elaborations to develop a good understanding of procedural learning. Background information on elaboration includes a discussion of learner-generated versus author-provided elaboration, and the minimalist versus maximalist approach to training users in the use of computer programs. Six tables summarize findings. (Contains 67 references.) (MAS)