Scaling and regression models of statistical moderation

 

T. O. Abe

 

Research Article | Published January 2014

Journal of Educational Research and Reviews, Vol. 2(1), pp. 7-11

 

 

Postgraduate School, College of Education, Ikere-Ekiti, Ekiti State, Nigeria.

Email: dr_abe07@yahoo.com 

 

………………………..............……………………………………………………………………........…………

 

View Full Text - PDF

 

This study investigated scaling and regression models of statistical moderation of internal assessment scores. This is an ex-post-facto research design in which there was no treatment and manipulation of subject, instead it involves the collection of data from records, cluster sampling technique was adopted to select five hundred (500) students that were made up of fifty students each from ten schools. The two questions and two hypotheses were analyzed using Pearson-product-moment correlation statistical technique. The results revealed that both scaling and regression model do not utter the strength of relationships among the internal assessment scores in Mathematics, English language and Integrated Science and at the same time there were significant relationships among the internal assessment scores in Mathematics, English Language and Integrated Science before and after the applications of statistical moderation by scaling and regression models. Based on these findings, this study does not recommend scaling and regression models as means of moderating internal assessment scores because they do not utter the strength of relationship as expected in order to ascertain the quality control of teachers’ assessment scores in Nigeria Secondary Schools.