Reliability and content validity of commercial tests and their correlation to pupils’ perfomance in mathematics Kajiado county , kenya

Abstract

This study was carried out to establish the reliability and content validity of commercial tests and their correlation to pupils‟ performance in mathematics: the case of public primary schools, Isinya District, Kajiado County. It involved nine teachers who were the key i nformants selected from different schools and were actively involved in mathematics teaching in class 8 or held re sponsibilities like mathematics panel heads, in their schools in Isinya District. The teachers analyzed the n ine commercial papers used for research in this project, and also indicated the Mathematics mean scores attained by the class 8 pupils in the years 2013 and 2012. A mixed descriptive survey research design was app lied and data was collected thro ugh questionnaires with key informants. Quantitative data analysis was applied to survey data collected via questionnaires. The frequency distribution was de scribed while data from questionnaires was qualitatively analyzed using tables and Fleiss Kappa method. The findings of the study revealed that the poor reliability and content validity of commercial papers compared to K.C.P.E, had an e ffec t on the subject mean scores leading to wrong conclusions‟ by the mathematics teachers or the school administration. Other factors like training in test construction have a varying influence on the validity and reliability of Commercial tests. Experienced teachers who had prior training in test development (33.3%) applied a number of this knowledge when analyzing the commercial papers . T his was evident from the findings of the study that commercial pap ers 006, 008 and 009 score rated low with k= 0.115 , 0.157 and 0.162 respectively. The commercial exams mean scores did not have any correlation with the standardized exam K.C.P.E which had an a verage mean difference of 6.2%, i t was concluded that commercial tests are gene rally invalid and unreliable. Due to lack of test construction knowledge by teachers as revealed by the study (66%). This is a contributing factor for the teachers choosing to use commercial tests. The study recommended research could be done on reliability and content validity of commercial papers and their correlation to pupils‟ performance in other subjects . A further research may be done to unveil the reasons that hinder public schools from purchasing standard set exams.