A comparison of data collection techniques used with discrete trial teaching

Abstract

This study was designed to examine the comparative value of three discrete trial teaching data collection techniques: Continuous recording, time sample, and estimation. The data collection was conducted by behavior interventionists while teaching children diagnosed with autism spectrum disorder skills using discrete trial teaching. Utilizing a counter-balanced design, data collection techniques were examined in regards to their accuracy, that is, their correspondence to the independent measurement of a primary observer collecting contemporaneous trial-by-trial data. Also assessed were the relative impacts of the various techniques on efficiency of therapy and rate of children's acquisition. Finally, interventionists rated their preference of and satisfaction with each of the three techniques. Continuous recording was generally the most accurate, although the other methods were accurate to a degree when used by trained recorders to justify their usage in applied settings. Estimation was the most efficient and time sample was the most preferred.