Accuracy of physical fitness assessment of female students of higher educational institutions

Abstract

The objective of the research was to increase the accuracy of physical fitness assessment of female students. The
methods of the research were as following: theoretical analysis and generalization; fitness testing;
linear interpolation method; least squares method; method of graph analysis; correlation and
regression analysis. Material. The scales quality was determined by the coefficients of
determination and the average approximation errors in percents. These indicators were obtainedfor the scale parameters ‘result-assessment’ and ‘assessment-assessment’ as well as for the physical fitness assessment of 1st-3rd years students (n=116). Results: it was found that the scales of
the four test exercises, in which the scores are calculated by linear regression equations, have
sufficient accuracy and on average 0.41±0.30 points or 15.6% higher assess the physical fitness of
female students than the scales that are made by the levels of preparedness. Conclusions. The
scales in which the scores were set by the linear regression equations were better selected (D =
99.70÷100.00%) and showed high accuracy (a = 0.00~1.01%). In contrast, the scales in which the
scores are calculated according to the level of training, at D = 92.92÷94.70 %, In contrast, scales
that score by fitness level, with D=92.92-94.70% and Ā=12.29-13.62%, do not accurately
determine the physical fitness of female students (Ā=7.74%). It was set that the scales’ accuracy
could be measured by multiple regression equations. Physical fitness assessments of female
students, which were formed by the equations of multiple regressions, coincided with the
assessments of level linear regressions, on which these formulas were obtained, to the third or
second decimal place. The multiple regression equation, which is compiled from estimates of fitness
levels, gives an average score of 0.60±0.38 less than the linear regression equations.