In my probability tutorial I used an example of comparing the average weight loss in a standard diet group versus the average weight loss in the new diet group. In that example the Independent variable is the “Group”, which has two categories (standard diet versus new diet) and the dependent variable is “Weight Loss”, measured in pounds.
Knowing that the amount of weight loss within a particular diet group will vary from one person to the next, we can’t just compare each individual’s weight loss in one group with each individual in the other group. So, we need to summarize the data somehow.
The t-test summarizes the measurements (e.g. amount of weight loss) by computing the average separately for each group. Then, statistical analysis software (e.g. SPSS) applies the appropriate mathematical formulas to carry out the t-test, which ultimately results in a p-value (probability value). By convention, if p < 0.05, the sample results provide strong evidence that the averages are statistically significantly different for the two groups.