Post Your Answer

Navika, errors can be independent or dependent depending on the value and the group mean. For example, if there are 6 values in a group, then the error is not independent. That is, some or the other factor might result in the increase or decrease of the value.

Yes. Unpaired T-test assumes that the populations have the same variance. This in result leads to same standard deviation.

Unpaired T-test has same variance for the two populations. You can use prism test to determine the variance. The P value from the test answers the questions such as variance of the populations and random selection of samples. Small value of P suggests that the variances are not similar. However, if you have plenty of data you can interpret the results and determine if the variances are equal or not.