A comparative analysis of general case simulation instruction and naturalistic instruction.

Effects of naturalistic instruction and simulation instruction for the acquisition, generalization, and maintenance of housekeeping and janitorial work skills in young adults with moderate and severe mental retardation were assessed. The performance of the four participants (under the two methods of instruction for two job task sequences) was assessed using a multiple probe design across behaviors. Both the naturalistic and simulation instructional methods involved noninstructional probes, instruction sessions, generalization probes, and withdrawal of instruction. Results of noninstructional probes parallel to simulation instruction sessions indicated only a minimal improvement over initial (base-line) noninstructional probes. By contrast, noninstructional probes parallel to naturalistic instruction showed a marked improvement for all four participants with correct responding reaching or exceeding the 90% level. Naturalistic instruction was an effective means of producing skill acquisition, while generalization of the learned skill was marginal.