Conducting t_tests
tags: #statistics/inferential/ttest
What is a t-test?
Always use Welch's t-test - accommodates for unequal variances while producing identical results to the regular t-test if you do not want to check all assumptions.
A t-test is a bivariate inferential statistic technique between a quantitative DV and a qualitative categorical IV used in hypothesis testing to compare the difference between means, as mediated by their variances.
- i.e., t-tests tells us how do these means differ while controlling for their variances
There are three main types of t-tests:
-
Two-Sample (Independent T-tests) - which compares the mean of two groups or population; two sub-types:
- Equal Variances
- Unequal Variances: Welch's T-test[1]
-
One-sample t-tests - examines whether the mean of a population is statistically different from a known or hypothesized value.
-
Paired t-test - pairwise comparison of the means on the same subject; commonly used for within-subject designs
T-tests can be used as part of exploratory data analysis (EDA) to compare the means of two groups of continuous data and to assess whether the difference between them is statistically significant.
Navigation: Conducting T-tests
Welch's t-test does not require assumptions of equal variances - as a result, it is a far more robust/rigorous t-test ↩︎