Comparison of six response-elimination techniques following VR reinforcement training in humans

Abstract

The present study employed humans in an examination of the relative response-eliminating effects of the following six procedures: Extinction training, omission training, response cost, omission training plus response cost, extinction training plus response cost, and omission training with a gradual introduction. Results indicated that all procedures eliminated responding significantly more effectively than did extinction training. In addition, omission training plus response cost, extinction training plus response cost, and omission training with a gradual introduction reduced responding significantly more effectively than did either omission training or response cost, which did not differ significantly from each other. Finally, omission training plus response cost and extinction training plus response cost decreased responding significantly more efficiently than did omission training with a gradual introduction, although they were not significantly different from each other. Much research attention has concentrated on assessing the relative response-eliminating effects of omission training, in which S is reinforced for omitting a previously reinforced response, and extinction training. In general, these studies (cf. Miller & LeBlanc , 1972; Topping & Larmi, 1973; Topping, Pickering, & Jackson, 1971, 1972; Uhl & Garcia, 1969; Uhl & Sherman, 1971; Zeiler , 1971) have employed infrahuman Ss and have found that omission training is the more efficient technique, especially with regard to the durability of the response elimination. Because of the paucity of data on the elimination of responding in humans, Johnson, McGlynn, and Topping (I973) recently employed college students in a comparison of four response elimination procedures. Their results indicated that response cost (punishment), omission training plus response cost, and omission training reduced responding significantly faster than did extinction training. Although the effects of response cost and omission training plus response cost did not differ significantly from each other, both procedures eliminated responding significantly more rapidly than did omission training. The present experiment was designed to (a) replicate and extend portions of the Johnson et al (I 97 3) study, as well as provide additional response elimination data on humans; (b) provide a more sensitive comparison of response cost and omission training plus response cost in