A Comparative Analysis of Reliability Methods

摘要

The study is concerned with the inappropriate usage of the reliability methods. It also identified the most commonly used of the reliability methods and the most appropriate methods for agricultural science objective test items. Reliability of a test is the extent that repeated measurements give consistent results for individual. A test is said to be reliable if it measures what it is supposed to measure consistently. There are many methods of measuring reliability which are: Test-retest, parallel form, Split-half, other methods of measuring internal consistency include Kuder Richardson KR20 and Kuder Richardson KR21.The study made a comparative analysis of reliability in four selected secondary schools in Ado-Ekiti Local Government Area of Ekiti state. Forty (40) items in agricultural science objectives test were constructed by the researcher. The objectives test was administered to the students in the four selected schools. The scores were compared across the sample schools. The descriptive analysis such as mean, standard deviation and comparative analysis of reliability methods were used. The hypothesis was held significant at 0.05 probability level. The result of the analysis shows that there was no significant differences between test-retest, parallel form, split half and Kuder-Richardson methods in agricultural science objectives test. The findings revealed that significant different exist in reliability methods employed in the sampled schools as a result of the findings, the instrument is recommended to be used in analyzing reliability method.