Analyzes uses and effects of coding schemes in research on counseling process and outcome. Concludes that interpretation of counseling events in terms of the categories of a coding scheme attenuates an adequate description of those events whenever the researcher uses the coder's interpretation of events as if it were veridical with the participant's interpretation. (Author)