Interrupted Time Series (ITS)

Interrupted time series (ITS) analysis is a quasi-experimental method that estimates the effect of an intervention by testing for changes in the level and slope of an outcome series after a known start date.

단절적 시계열(ITS) 분석은 알려진 시작 시점 이후 결과 시계열의 수준과 기울기 변화를 검정하여 개입의 효과를 추정하는 준실험 방법이다.

How do I evaluate whether a policy or intervention changed a trend over time?

정책이나 개입이 시간에 따른 추세를 바꿨는지 어떻게 평가하나요?

Interrupted time series analysis, usually fitted as segmented regression, compares the pre-intervention trend with the level and slope after the intervention. It fits outcomes measured repeatedly at regular intervals before and after a clearly defined start date when no control group is available.

When to use Interrupted Time Series (ITS)

• A policy, program or event with a known start date
• Monthly, weekly or yearly outcome measurements both before and after the change
• Population-level data without a randomized comparison group

When to use something else

• A comparable untreated group is available: use difference-in-differences or a comparative ITS
• Assignment is determined by a cutoff on a score rather than time: use regression discontinuity design

Assumptions and requirements

• Enough time points before and after the intervention
• Pre-intervention trend would have continued without the intervention
• No other co-occurring events at the same time
• Autocorrelation and seasonality handled in the model

What the analysis reports

• Pre-intervention slope
• Immediate level change
• Change in slope after the intervention
• Plot of observed and counterfactual trends
• Autocorrelation diagnostics (e.g., Durbin–Watson)

How to report it

Following the intervention, [outcome] showed an immediate level change of [value] (95% CI [[lower], [upper]], p = [value]) and a slope change of [value] per [time unit] (p = [value]).

Related methods

• Difference in Differences (DiD)
• Regression Discontinuity Design (RDD)
• ARIMA Model

Run Interrupted Time Series (ITS) on Quorant