Hypothesis tests
CogStat supports a series of hypothesis tests.
- For the Bayesian tests, CogStat uses the default priors as in JASP, so that the two software gives the same result (unless you modify the priors in JASP).
- See more details about the special version of some hypothesis tests on the Differences in calculations between CogStat and other programs page.
Available tests
Currently CogStat supports the following hypothesis tests (tests in parenthesis are included in several analysis):
- Shapiro-Wilk test
- One-sample t-test
- with Bayesian one-sample t-test
- Wilcoxon signed-rank test
Explore relation of a variable pair:
- Pearson correlation hypothesis test
- with Bayes Factor for Pearson correlation
- Spearman correlation hypothesis test
- Chi-squared test
Compare repeated measures variables:
- (Shapiro-Wilk test)
- paired t-test
- with Bayesian paired t-test
- paired Wilcoxon test
- repeated measures ANOVA
- Mauchly’s sphericity test (assumption)
- Holm-Bonferroni corrected post-hoc tests
- Friedman test
- Durbin-Conover test
- McNemar’s test
- Cochran Q-test
- (Shapiro-Wilk test)
- modified t-test
- Mann-Whitney test
- Levene’s test (assumption)
- two-samples t-test
- with Bayesian two-samples t-test
- Welch’s t-test
- one-way ANOVA
- post hoc Tukey’s HSD tests
- Kruskal-Wallis test
- post hoc Dunn’s test
- Two-way ANOVA
- (Chi-squared test)