Interrupted Time-Series Designs for Evaluating Health Communication Campaigns

Interrupted time series (ITS) constitute a class of powerful quasi-experimental designs for evaluating health communication campaigns. This article: 1) describes the basic elements of time-series designs in general and interrupted time-series designs in particular; 2) discusses briefly some of the more widely recognized ITS designs in order of increasing methodological rigor; 3) introduces the two major types of time-series statistical analysis, and; 4) provides a detailed example of a sophisticated ITS design used to evaluate the effectiveness of anti-marijuana campaigns in two communities. While ITS designs have been used infrequently in campaign evaluation, enough such studies have been conducted to demonstrate their utility in this domain. It is hoped that future campaign evaluations will make much greater use of this sophisticated technique which appears tailor-made to investigate many of the questions of greatest interest to campaign researchers and practitioners.