The Baseline Corrected Tau single-case effect size was developed to address the problem of baseline trend in interrupted time series (AB) data. This statistic uses nonparametric methods which are efficient in small samples and robust to outliers and serial dependency.

Baseline Corrected Tau uses a two-step process to estimate an effect size for an AB single-case design. First, monotonic baseline trend is estimated and (if necessary) corrected using Kendall's Tau rank correlation coefficient. If a statistically significant baseline trend is present, baseline trend may be corrected across both A and B phases using the nonparametric Theil-Sen estimator, which is based on Tau. Second, an effect size is calculated as a Tau correlation between a dummy code variable (A phase = 0, B phase = 1) and either the original or corrected data.

To use the calculator, begin by entering the A phase and B phase data below. After testing for baseline trend, a recommendation will be made regarding whether or not to correct for baseline trend. An effect size may then be calculated.

PHASE A (BASELINE) |
PHASE B (TREATMENT) |

**How to Cite:** Tarlow, K. R. (2016). Baseline Corrected Tau Calculator. http://www.ktarlow.com/stats/tau

**See Also:** Tarlow, K. R. (2017). An improved rank correlation effect size statistic for single-case designs: Baseline Corrected Tau. *Behavior Modification, 41*(4), 427-467. http://dx.doi.org/10.1177/0145445516676750

**R Code:** Tarlow, K. R. (2017, June). *Baseline Corrected Tau for single-case research* (R code). http://ktarlow.com/stats

**Updates**

2017-11-20: An error was corrected that led to inaccurate p-values in cases where all baseline data points were equal, i.e., where the baseline was flat.

2017-10-02: An error was corrected that led to inaccurate p-values in cases where the baseline trend coefficient (Tau) was negative.

© 2020 Kevin Tarlow