Question

MVE problem 2(statistics)
a) What is the difference between an dependent and an independent t-test?
b) Explain what a p-value of 0.02 means? Give an example of an independent t-test.
c) Explain the benefit of using 2 levels for each factor? When do you usually use such patterns.
d) Why is MLR not used in multivariable calibration for spectroscopy (eg NIR) data for Y variable

Answer 1

The main difference between dependent and independent t-test is is that the sample observations are related in the former where as sample observations are independent in the later .

eg: If we are supposed to compare the weight of  two groups of people say grp1 and grp2 , these groups are taken from the same populations but consists of diffrent observations , in this case we will use independent t test since the groups are independent .

similarly if we were to compare the weight of people in grp1 before and after diet then we will have to use dependent t-test or paired t -test because in this case same participants are present in both the groups and hence our samples become dependent .

Some of the other differences are as follows :

1. sample size is equal for both the groups in dependent t-test whereas the groups can have different sample size in independent t-test .

2. In dependent t-test we need the sample mean and variance of only the difference value between the two groups , whereas we require sample mean and variance of both the groups in independent t-test in order to conduct the test.