I am interested in education research, sometimes i found that using of t-test is common in educational research. In my finding almost undergraduate student will chose t-test to analyze their hypothesis. Since i do not have much knowledge about statistics, i'd like to asked is there any specific requirements of using t-test in education research? and in which condition this test cannot be used?
Thank you
Best,
Hello Nabilah,
t-tests compare means of two groups chosen by the investigator. For example students in a control group versus students in the experimental group. The assumptions for the general linear model (including regression, ANOVA (which includes t-tests), and ANCOVA) are that the errors (residuals) are normal and homogeneous (Eisenhart 1947, Seber 1966, Neter et al 1983 pp 31& 49, Quinn and Keough 2002 pp 110 & 280), and independent. Evaluation of assumptions in texts, where it occurs, often entails a residual versus fit plot (for homogeneity) and a normal score plot or Quantile-Quantile plot (for normality of the residuals).
It is a common misconception that one must meet these assumptions before doing the analysis. In fact one can only check the assumptions from the residuals.
Before applying a t-test, it is worth considering the response variable (what you are measuring) in a qualitative way. Is it binary (yes/no) response? If so the assumptions won't be met. Is it on an ordinal scale ? Such as a Likert scale, 1-5 (in ResearchGate search: Likert scale t-test). If so, the t-test is not the best way to compare the 2 groups and the assumptions may not be met. Is it on a cardinal scale (interval or ratio scale variables)? if so, the t-test is best place to start, for comparing means of two groups.
For current practice in your field have a look at published research similar to yours in highly ranked educational research journals.
https://www.scimagojr.com/journalrank.php?category=3304
Good luck with your research,
David S
Refer this article:
https://study.com/academy/lesson/what-is-a-t-test-procedure-interpretation-examples.html
Hi David and Santosh,
Thank you very much for all your answers and reference article. It help me a lot in my working to monitor the students.
Best,
Nabilah
Hi Nabilah
1. Testing the significance of random variables/parameters is very important issue in all sciences.
2. T-test is one of these measures.
3. T-test is not one test, it is a group: for one sample, two samples, related samples, unrelated samples, parametric test and non-parametric test.
4.You need to study:
1. Landau, S. and Everitt, B. S.(2004). A Handbook of Statistical analyses using SPSS.
2.Howitt, D. and Cramer, D.(2008). Introduction to SPSS.
Regards,
Zuhair
Hi Nabilah
One of the above book and an additional manual are attached.
- Badges
- Science topic