Comparison of simulation training on self and using a doll for teaching generalized menstrual care to women with severe mental retardation.

This study examined the effects of two simulation methods of general case instruction in teaching generalized menstrual care; long-term maintenance was assessed up to 18 months. Within a split multiple-baseline design across subjects, three middle-school students and one adult woman with severe or profound mental retardation initially were assigned either to on-self instruction or to instruction using a doll. The two learners receiving doll training also participated in instruction on self. A modified forward chaining procedure with identical task analysis steps was used with each method. The range of positive and negative training examples used also was identical for the two techniques except that doll-size underwear and sanitary pads were used for doll training. Participants were taught (by demonstrating on themselves or by manipulating the doll) to change stained underwear and stained pads and not to change underwear or to obtain a pad in the absence of a stain. The dependent variables consisted of simulation and in vivo (natural menses) assessment probes and reflected the percentage of trials in which all task analysis steps were performed correctly. Correct performance on critical menstrual-care steps alone also was assessed. Results indicated limited generalization after training using a doll. While all four learners demonstrated high levels of generalized responding following on-self instruction, the effect decreased over time for two participants. Implications for defining the dependent variable, selecting teaching examples, and enhancing maintenance are discussed.