Psychological Science Double-blind Tests of Subliminal Self-help Audiotapes

Three replications of a double-blind experiment tested subliminal audiotape products that were claimed to improve memory or to increase selfesteem. Conditions of use adhered to manufacturers' recommendations, and subjects ( N = 237) were limited to persons who desired the effects offered by the tapes. Actual content and labeled content of tapes were independently varied, so that some subjects who believed they were using memory tapes were actually using self-esteem tapes, and vice versa. After a month of use, neither the memory nor the self-esteem tapes produced their claimed effects. Nevertheless, a general improvement for all subjects in both memory and self-esteem (a nonspecific placebo effect) was observed, and more than a third of the subjects had the illusion of improvement specific to the domain named on the tape's label. Subliminal self-help (SSH) audiotapes are widely advertised as being able to produce many desirable effects, including weight loss, smoking cessation, anxiety reduction, and improvement of sexual function. In ordinary use, purchasers of SSH tapes know and desire the target effect, which is clearly marked on the tape label and in accompanying literature. However, in listening (perhaps daily) to the tape, the user may hear nothing relevant to the target effect. The audible content of the tape typically consists of relaxing material, often music or recorded nature sounds. There a re numerous repor ts of claimed therapeutic effects of subliminal audiotapes, but no such report has apAddress correspondence to A.G. Greenwald, Dept. of Psychology, NI-25, University of Washington, Seattle, WA 98195. Telep h o n e : ( 2 0 6 ) 543-7227, e m a i l : peared in a competitively refereed psychology journal. Further, many of these reports have been produced by researchers associated with manufacturers of the tapes (Becker & Charbonnet, 1980; Borgeat & Chaloult, 1985; DocheBudzynski & Budzynski, 1989; Taylor, 1988; VandenBoogert, 1984; see review by Eich, in press). The present research was conducted to provide a rigorous assessment of therapeutic effectiveness of selected SSH audiotapes. Four aspects of procedure were essential to the experimental protocol. First, a double-blind method was used in order to separate effects of subliminal content from passible placebo effects. Second, conditions closely resembled the conditions of ordinary use of SSH tapes, in order not to omit contextual aspects of the ordinary procedure that might be necessary to claimed effects. Third, the research used subjects who were motivated to achieve the goals claimed by the tapes, because such motivation is plausibly a precondition for occurrence of claimed effects and is a condition that applies generally in the marketplace. And, fourth, the research used both pretest and posttest measures of status relative to the target goals, in order to obtain the increased sensitivity afforded by analyses of covariance with pretests used as covariates. Among the target goals for which SSH tapes were available from several manufacturers, improving memory and increasing self-esteem were selected both because it was easy to locate subjects who desired these goals, and because several well-established measures that could be used in a pretest-posttest design were available. Some manufacturers generously provided multiple copies of their tapes for the research; with manufacturers' permission, we also duplicated mail-ordered tapes for use in the research. Audible identifications of contents (at the beginnings of some of the tapes) were carefully and selectively erased by an audio laboratory prior to giving any tapes to subjects.'