A framework is presented for understanding how errors arise in institutional research, and principles are suggested to avert errors in work processes of institutional research offices. Error is defined as a product which fails to meet professional standards, has unacceptable process limitations, or is unacceptable to a client. The following four principles are suggested to make work processes more error-sensitive and error-free: (1) clarifying environmental constraints and redesigning how work is done; (2) empowering agents to overcome organizational, perceptual, or knowledge-base limitations in defining, managing, or interpreting the results of work processes; (3) designing heuristics which emphasize openness of information and feedback loops; and (4) using technologies which generate intraprocess feedback. Environmental constraints include the organizational climate for work as well as political, legal, professional, and other external constraints. Four examples of project redesign are discussed: a survey on student charges, a comparative analysis of faculty salaries, an enrollment planning process, and redefinition of a comparator group. (Contains 12 references, 4 tables, and 2 charts). (SW)