When you have several observations, such as 30 or 40 ,t test is the best, but if you have to compare two means,chi square is your choice. Good luck.
What's your research question? How many research groups? How many variables? What type of data? There's a lot more which needs decided before you pick a test.
Check the assumptions of the test you wish to run and use the test for the highest level of measurement for which the assumptions are reasonably satisfied. If you have a large number of observations, the law of large numbers will almost always allow you to use the independent t-test which is far more powerful than a nonparametric test. Also keep in mind that the t-test compares means whereas the the Chi-square is comparing counts.
Hi Hardi Abdul Rahaman; it depends on the nature of your data. If your data is interval and ratio data, then you can use a T-test with other assumptions of normality in place. However, for categorical data, the chi-square is more suitable. You will have to check the assumptions for both these tests, but the nature of your data should direct the test you will use. I hope this makes sense and good luck.
This has more to do with not only the number of observations or groups, but more specificallly the level of measurement of your dependent variables. A Chi square is for comparing nominal data while a T-test compares interval or ratio data.
- Badges
- Science topic
- Similar topics
- Engineering
- Modeling