Question

1. Why would a t-test be conducted prior to an intervention being implemented in an experimental research? What purpose do t-tests serve for after the intervention is implemented?

Answer 1

- FA t- test assusmes that the population arrives from normal distribution but unlike the z test , it does not have the knowledge of sample variance only .

- In an experimental research generally the population under study are divided in to two groups .

* Group 1 : The experimental group on which the intervention ( experiment ) is carried out to test the effect of experiment that is BEFORE AND AFTER effects .

* Group 2 : A group which receives a control experiment or placebo .

- For group 1 , we apply paired t - test to study effect of experiment that is wether conducting the experiment is worthwhile or not . To make this more clear let us take one example .

Example ., A group of 100 sedentary overweight individuals are surveyed to test the effect of exercise on their body . Their weights are recorded at the start of experiment before data (3 months) and post 3 months exercise regime after data ( after 3 months). A paired t - test is conducted with the null hypothesis Ho= mew d = 0 against alternative hypothesis . H1 = mew d > 0 which would test wether exercise has resulted in reduction of weight . Here mew1 being average weight of the 100 survey individulas after the experiments . mew 1 - mew 2 = mew d .

- A test statistic is computed and the p - value decides wether the after was leading the reduction in weight . If yes then the experiment could be carried forward .