Interrupted time series

Interrupted time series analysis (ITS), sometimes known as quasi-experimental time series analysis, is a method of statistical analysis involving tracking a long-term period before and after a point of intervention to assess the intervention's effects. The time series refers to the data over the period, while the interruption is the intervention, which is a controlled external influence or set of influences. ^{[1] [2]} Effects of the intervention are evaluated by changes in the level and slope of the time series and statistical significance of the intervention parameters. ^[3] Interrupted time series design is the design of experiments based on the interrupted time series approach.

The method is used in various areas of research, such as:

• political science: impact of changes in laws on the behavior of people; ^[2] (e.g., Effectiveness of sex offender registration policies in the United States)
• economics: impact of changes in credit controls on borrowing behavior; ^[2]
• sociology: impact of experiments in income maintenance on the behavior of participants in welfare programs; ^[2]
• history: impact of major historical events on the behavior of those affected by the events; ^[2]
• psychology: impact of expressing emotional experiences on online content; ^[4]
• medicine: in medical research, medical treatment is an intervention whose effect are to be studied;
• marketing research: to analyze the effect of "designed market interventions" (e.g., advertising) on sales. ^[5]
• environmental sciences: impacts of human activities on environmental quality and ecosystem dynamics (e.g., forest logging on local climate). ^{[6] [7]}

See also

• Quasi-experimental design

References

1. Ferron, John; Rendina‐Gobioff, Gianna (2005), "Interrupted Time Series Design" , Encyclopedia of Statistics in Behavioral Science, American Cancer Society, doi:10.1002/0470013192.bsa312, ISBN   978-0-470-01319-9 , retrieved 2020-03-09
2. McDowall, David; McCleary, Richard; Meidinger, Errol; Hay, Richard A. Jr. (August 1980). Interrupted Time Series Analysis. SAGE. pp. 5–6. ISBN   978-0-8039-1493-3.
3. Handbook of Psychology, Research Methods in Psychology, p. 582
4. Bollen; et al. (2019). "The minute-scale dynamics of online emotions reveal the effects of affect labeling" . Nature Human Behaviour. 3 (1): 92–100. doi:10.1038/s41562-018-0490-5. PMID   30932057. S2CID   56399577.
5. Brodersen; et al. (2015). "Inferring causal impact using Bayesian structural time-series models". Annals of Applied Statistics. 9: 247–274. arXiv: 1506.00356 . doi:10.1214/14-AOAS788. S2CID   2879370 . Retrieved 21 March 2019.
6. Li, Yang; Liu, Yanlan; Bohrer, Gil; Cai, Yongyang; Wilson, Aaron; Hu, Tongxi; Wang, Zhihao; Zhao, Kaiguang (2022). "Impacts of forest loss on local climate across the conterminous United States: Evidence from satellite time-series observation" (PDF). Science of the Total Environment. 802 149651. Bibcode:2022ScTEn.802n9651L. doi:10.1016/j.scitotenv.2021.149651. PMID   34525747.
7. Li, Yang; Zhao, Kaiguang; Hu, Tongxi; Zhang, Xuesong. "BEAST: A Bayesian Ensemble Algorithm for Change-Point Detection and Time Series Decomposition". GitHub .