Nelson D L, Gergenti E, Hollander A C
J Autism Dev Disord. 1980 Sep;10(3):311-21. doi: 10.1007/BF02408290.
A color-coded "extra prompt" procedure was compared to a "no extra prompt" procedure in teaching autistic children and adolescents how to lace shoes. One randomly assigned group of 10 autistic subjects first learned to lace shoes whose laces and eyelets were color-coded red and white, and then encountered the no extra prompt condition in which color codes could no longer be depended upon to solve the position discriminations required to lace properly. In a counterbalanced fashion, the other group of 10 autistic subjects reached criterion on the non-color-coded, naturalistic shoe before experiencing the extra prompt condition. Analysis of variance and post-hoc analyses suggest that subjects who first learned under the color-coded, extra prompt condition encountered significant difficulty in transferring their newly acquired skill to the naturalistic, non-color-coded condition, whereas subjects who learned initially without the extra prompts had little difficulty with the subsequent color-coded condition. A follow-up procedure requiring all subjects to choose between the color prompt and the position cue revealed that 11 of 20 subjects consistently chose the color cue, even though it resulted in improper lacing. It is recommended that clinicians avoid the use of highly salient, non-criterion-related prompts in teaching certain types of adaptive skills to autistic children.