Resumen de A study of prompting versus confirmation in machine and text presented programmed learning under two conditions of responding
Sixty‐four secondary school children were given a program on probability either by machine or in text form. For some groups the response blanks were already filled in‐‐others constructed their responses. A further condition was making responses either overtly or covertly. Children of above average intelligence constructing their answers were significantly better in post‐test performance if they made covert rather than overt responses. Machine users were poorer than text users because of poor scores made by the response‐prompted machine group (not paralleled in the corresponding text group). Though covert constructed responses were again superior to overt on a test given nine weeks later, the difference was not significant. Details of interactions between the variables and times taken by each group are given.
