RESUMEN
Doing what is promised or accurately reporting what has been done increases with correspondence training. Early research showed that training produced positive correlations between saying and doing, but did not always produce useful generalized correspondence to new, untrained say-do sequences (Baer, 1990). Recent research (reviewed here in detail) questioned early procedures and introduced new issues: functional criteria, functional verbal responses, baseline measures, and response chains in correspondence training. Correspondence research was almost abandoned after 1992. This review suggests new procedures and directions to revive this important research area. Specific suggestions are to combine correspondence procedures with similar features from ongoing research in compliance and self-instruction; consider important guidelines for future research derived from recent studies; consider interactions between applied and basic research with correspondence issues; compare correspondence between different participant cohorts; examine the use of descriptive and functional terms; examine punishment procedures; and provide a behavioral analysis of the relations between verbal and nonverbal behaviors related to correspondence.