t-test

#In Data Analyst series

What is t-test?

The t-test is a parametric technique used to test the statistical significance of the difference between two means.

When applying the t-test, it is important to note that only two means can be compared.

Types of t-tests

Mainly 3 types of t-tests as following:

• Single sample t-test
• Independent Groups t-test
• Dependent Groups t-test

Let’s dive into each one of them:

Single sample t-test

It tests whether the sample mean of a variable differs significantly from the mean of that value in the population.

Assumption: The measurements of the dependent variable are at least at the interval scale level. The measurements of the dependent variable show a normal distribution in the universe.

For example:

Does the average score of the employees in a company towards the institution differ significantly from the population mean score of 70?

Two-way Hypothesis

H0: The average score of the staff’s attitude towards the institution does not differ significantly from the population average of 70. ( µ = 70)

H1: The average score of the personnel’s attitude towards the institution differs significantly from the average score of 70, which is the population average. (µ ≠ 70)

By running one-sample t-test, if p-value <.05, we would reject the null hypothesis.

Independent Groups t-test

The independent samples t-test tests the statistical significance of the difference between the means obtained from two unrelated samples.

Assumption: The measurements of the dependent variable are at least at the interval scale level. The measurements of the dependent variable show normal distribution in both groups.

The samples whose means are to be compared are independent.

Homogeneity of variance (common variance is used if homogeneous, separate variance is used if heterogeneous. As the sample size increases, both values get closer to each other). (We want the variances are almost the same so that we can run the analysis)

Dependent Groups t-test

The dependent sample t-test tests the statistical significance of the difference between the mean of two related samples.

This test is used for either repeated measurements of the same experiment or measurements from paired samples.