ITSSIM: Interrupted Time-Series Simulation

How Does ITSSIM Work?

ITSSIM uses Monte Carlo simulation to estimate an effect size for an AB single-case data set. Essential to ITSSIM is the assumption that one observed single-case data set can be explained by many plausible intervention effects. Therefore, unlike other single-case statistical methods, ITSSIM effect sizes are not calculated from observed data; instead, ITSSIM effect sizes are calculated from many simulated time-series that are based on a range of conditions which could plausibly account for the observed data. An ITSSIM analysis proceeds in three stages:

  1. Parameter Estimation. Level, trend, variability, and autocorrelation parameters (and their standard errors) are estimated from the observed data.
  2. Time-Series Simulation. 100,000 artificial time-series are simulated based on the A phase parameter estimates, and another 100,000 artificial time-series are simulated based on the B phase parameter estimates. These time-series represent different null effects and intervention effects that could plausibly explain the observed A phase data and B phase data, respectively.
  3. Effect Size Calculation. Effect size statistics (including d and r) are estimated from the two distributions of simulated time-series means.


How to Cite: Tarlow, K. R. (2018). ITSSIM: Interrupted Time-Series Simulation, Version 1.0. College Station, TX: Author. Retrieved from

See Also: Tarlow, K. R., & Brossart, D. F. (in press). A comprehensive method of single-case data analysis: Interrupted Time-Series Simulation (ITSSIM). School Psychology Quarterly. (preprint)

© 2021 Kevin Tarlow