Additional comments on multiple-baseline designs in instructional research.

Abstract

A number of measurement and design issues that are critical to the use of multiple-baseline procedures in evaluating instructional interventions were highlighted. First, issues related to the interaction between length of baseline assessment and the following outcomes were presented: (a) deceleration in behavior across baseline, (b) prediction of behavior change, (c) error analyses performance on instructional stimuli, and (d) reactivity of observation. Finally, an attempt was made to match the variety of multiple-baseline designs to specific questions often asked by instructional researchers.