A Comprehensive set of Postestimation Measures to Enrich Interrupted Time-series Analysis. (March 2017)

• Record Type: Journal Article
• Authors: Linden, Ariel
• Is Part Of: Stata journal. Volume 17:Number 1(2017)
• Page Start: 73
• Page End: 88
• Publication Date: 2017-03
• Subjects: st0389_3 -- itsa -- interrupted time-series analysis -- quasiexperimental designs -- causal inference -- counterfactual; Statistics -- Periodicals; Statistics -- Computer programs -- Periodicals; 001.422
• DOI: 10.1177/1536867X1701700105
• Languages: English
• ISSNs: 1536-867X

Abstract

While the primary goal of interrupted time-series analysis (ITSA) is to evaluate whether there is a change in the level or trend of an outcome following an interruption (for example, policy change, intervention initiation), a series of additional measures may be relevant to the analysis. In this article, I seek to fill a gap in the ITSA literature by describing a comprehensive set of measures that can be computed following ITSA models, including those that fulfill the primary goal and those that provide supplementary information about trends. These measures can be calculated using theitsa command; this article therefore serves as a complement to "Conducting interrupted time-series analysis for single and multiple group comparisons" (Linden, 2015, Stata Journal 15: 480–500), which introduced theitsa command. Specific ITSA postestimation measures described in this article include individual trend lines, comparisons between multiple interventions, and comparisons with a counterfactual.