Comparison between measures of the Response Class “Properly separate Litter”

Abstract

The most used measures to study the response class “proper separation of litter” on researches in behavior analysis are weigh­ing and counting the recyclable items deposited in bins. Even though different measures are used, there seems to be no discussion in the literature regarding the relative appropriateness of different measures to the dependent variable. The present study set to make a comparison between the aforementioned measures and direct observation of the participants’ behavior. To perform such a comparison the study was designed as a systematic replication of the study on recycling behavior by Austin et al. (1993). Participants were mem­bers of two buildings of a university and eventual passers-by in the area. A multiple baseline AB design was employed. Specific were introduced as the independent variable. The correlation coefficient between direct observation and items counting was higher than the coefficient between direct observation and weighing. The results attained by the original research were not replicated. The results reported here seem to suggest that future research should assign greater confidence to the counting of items.