Comparison of Predictable and Unpredictable Trial Sequences during Small-Group Instruction

Abstract

The use of predictable (round robin) and unpredictable (random) trial sequences during small-group instruction was evaluated in three experiments in teaching word and abbreviation identification to four students with learning disabilities; a fifth student participated in part of Experiment I before moving to another school. In Experiment I, a progressive time-delay procedure was used to teach word reading in a small group, and the effects of a single-trial, predictable sequence was compared to a single-trial, unpredictable sequence. In Experiment II, a progressive time-delay procedure also was used to teach abbreviation identification in a small group, and the effects of a multiple-trial, predictable sequence was compared to a multiple-trial, unpredictable sequence. In Experiment III, a model-test procedure was used in a small group, and the effects of the multiple-trial, predictable sequence was compared to the single-trial, unpredictable sequence. An adapted alternating-treatments design was used in all experiments. Results indicate that the progressive time-delay procedure was reliably implemented and was effective in establishing criterion-level responding by all group members. Students also learned words taught to other students through observation. In Experiment I, the two trial sequences did not differ substantially, and in Experiment II mixed effects were found. With the model-test procedure in Experiment III, two students initially produced higher levels of correct responding in the multiple-trial, predictable sequence; however, no substantial differences were found in observational learning. Across all investigations, no consistent effects of the trial presentation methods were noted.