Dependent Sample t-Test: Interval Estimation - Quick Tab

Select the Quick tab of the Dependent Sample t-Test: Interval Estimation dialog box to access options to implement the procedure (Steiger & Fouladi, 1997) for computing a confidence interval for E_s, the standardized effect size.  In this case, E_s is a quantity that is difficult to interpret unless certain simplifying conditions (e.g., equal variances on both measurements, known correlation between measurements hold). We include the interval primarily for mathematical completeness.

Observed t

Enter the observed value of the t-statistic in the Observed t box.

Sample Size (N)

In the Sample Size (N) box, enter the sample size used in the study.

Conf. Level.

Enter the confidence level for the confidence interval calculation in the Conf. Level box. Note that confidence limits are presented for two quantities:

Delta

The upper and lower limits are for a confidence interval on δ, the noncentrality parameter of the noncentral t-distribution.

Standardized Effect

These upper and lower limits are for E_s. Keep in mind that in this case, E_s is standardized by a quantity that is a function of both within-group standard deviations and the correlation between measurements. When both standard deviations are equal and the correlation between measurements is equal to .50, this quantity is equal to the difference between the two means divided by the within group standard deviation, as in the independent sample t-test.