Regression analysis to assess the throwing ability with selected anthropometric and functional variable for team handball players

Abstract

The purpose of the study was to develop regression model to assess the throwing ability in team handball on the basis of selected physical and anthropometric variables for inter university players. For the present study 26 male handball players of Lakashmibai National Institute of Physical Education, Gwalior of age 21±5, on the basis of purposive sampling were selected as subjects. Multiple regression analysis was used to classify best indicators among selected variables for long throwing in handball. Five variables were used as discriminating factor to classify best indicators for long throwing ability i.e. upper arm length, lower arm length, hand length, humerus bicondylar diameter and hand grip strength. To measure upper arm length, lower arm length and hand length, humerus bicondylar diameter and hand grip strength, steel measuring tape, sliding caliper and Hand grip dynamometer is used respectively. After applying multiple regression analysis long shooting ability is significantly correlated with Hand Grip strength and Upper Arm Length at 1% level and Humerus Bicondylar Diameter at 5% level. Comparing with two models, the second model, the value of R2 is 0.587, which is maximum and therefore second model shall be used to develop the regression equation. Since F- value for the second model is highly significant, it may be concluded that the model selected is highly efficient also. The variable having higher magnitude of absolute function value is more powerful in discriminating two group means that leg explosive strength is most powerful in discriminating the two groups. In second model t values for the entire two regression coefficient are significant as their p value are less than 0.05 (p