~ 237 ~ International Journal of Physical Education, Sports and Health 2018; 5(1): 237-241 P-ISSN: 2394-1685 E-ISSN: 2394-1693 Impact Factor (ISRA): 5.38 IJPESH 2018; 5(1): 237-241 © 2018 IJPESH www.kheljournal.com Received: 12-11-2017 Accepted: 15-12-2017 Dr. Rajeev Choudhary Professor and Head, School of Studies in Physical Education, Pt. Ravishankar Shukla University, Raipur, Chhattisgarh, India Correspondence Dr. Rajeev Choudhary Professor and Head, School of Studies in Physical Education, Pt. Ravishankar Shukla University, Raipur, Chhattisgarh, India Application of “independent t-test” by using SPSS for conducting physical education researches Dr. Rajeev Choudhary Abstract Different types of researches are conducted in the field of physical education, such as case studies, philosophical researches, historical researches, survey studies and experimental studies. As per the trend and interest of today’s researchers, survey studies are the most preferred studies. There are different types of surveys; most commonly used surveys are relationship studies, descriptive studies, prediction studies and comparative studies. Out of all types of surveys, comparative studies are preferred. The easiest and smallest comparison can be between two groups. T-test is used to compare two groups, but both groups should have independence. Along with independence, normality and homogeneity of variance is required. If normality and homogeneity of variance is not fulfilled, there are alternate options. Keywords: Independent t-test, normality, homogeneity of variance Introduction Prologue T-test is a parametric statistics used to conduct comparative studies. It may be a comparative survey or an experiment, since different verities of t-tests are available to analyze data. Different assumptions are applicable to different t-tests. In precise form, it can be said that “t- test is a statistical techniques used it find out the significant difference among the means of two samples or two observations”. A Common assumption should be fulfilled before applying t-test i.e. normality of data. There are different ways to test the normality of data i.e. descriptive statistics, QQ Plot, and formal tests. So, first steps should be “testing of normality”. If the assumption of normality is not fulfilled, parallel non parametric technique should be applied. Types of t-tests S. No. Name Parallel non parametric statistics 1. Independent t-test Mann Whitney U test 2. Dependent t-test Wilcoxon Signed rank test 3. One sample t-test Sign test T-test: type- I (Independent t-test) Independent t-test in used to find out the significance difference between the means of two independent samples. There is one more assumption of independent t-test i.e. homogeneity of variance and this required assumption is tested by “levene statistics”. Example A researcher wants to find out the significant difference between the blood glucose level of two samples (Males and females). Researcher used “Static Group Comparison Design”.