Aminu Idris Harbau, Ibrahim Shu'aibu & Abubakar Abdullahi Yamusa
ANALYSES OF THE STANDARD ERROR OF MEASUREMENT FOR SECONDARY SCHOOL CERTIFICATE QUALIFIYING EXAMINATION AND SENIOR SCHOOL CERTIFICATE EXAMINATIONS ECONOMICS TEST ITEMS IN KANO,
NIGERIA
Abstract
The study is aimed at analysing the Standard Error of Measurement (SEM) of
economics test items of 2015 SSCQE and 2016 SSCE economics test items
employing the Item Response Theory (IRT). An expo facto research design was
employed in the analysis. The design was adopted becausethe studyused
secondary data collected from the Kano State Education Resource
Department (KERD) and National Examination Council (NECO). It aimed at
establishing how standard the items were.The population of the study consists
of the fifteen thousand four hundred and eighty one (15,481) students who sat
for 2015 economics SSCQE and 2016 SSCE in Kano State owned senior
secondary schools. It consisted of ten thousand nine hundred and fifty (10,950)
male students and four thousand and thirty one (4,531) female students. A
sample size of 1,000 students was selected using multistage sampling
technique. The data collected was coded and run using IRTPRO package to
determine the SEM indices of Kano State 2015 SSCQE. Findings from the
study revealed that the SEM of 2015 SSCQE and 2016 SSCE economics test
items based on two-parameter logistic (2PL) model was high. This indicates
consistency in measuring student ability in economics. The study found that
the massive failure in Kano State Certificate Qualifying Examinations over
the years was not caused by the quality of the SSCQE items but by other
variables which are not related with psychometric properties of the test items.
It was therefore concluded that there was no superiority in terms of item
quality between the items generated by KERD. As such performance in
SSCQE may determine performance in NECO. Given the various advantages
of IRT over the other popular measurement frameworks, the government
should encourage KERD to adopt this measurement framework. IRT can
enable KERD to place examinees on the correct ability level to describe the
test items and the abilities of the examinees. This will enable KERD to meet the
best practices of global examination standard and overcome the lingering
superiority complex of one examining body over the others.
Keywords: Standard error of measurement, IRT, KERD, Qualifying, NECO, Economics
Aminu Idris Harbau, Ibrahim Shu'aibu & Abubakar Abdullahi Yamusa
ANALYSES OF THE STANDARD ERROR OF MEASUREMENT FOR SECONDARY SCHOOL CERTIFICATE QUALIFIYING EXAMINATION AND SENIOR SCHOOL CERTIFICATE EXAMINATIONS ECONOMICS TEST ITEMS IN KANO, NIGERIA
Abstract The study is aimed at analysing the Standard Error of Measurement (SEM) of economics test items of 2015 SSCQE and 2016 SSCE economics test items employing the Item Response Theory (IRT). An expo facto research design was employed in the analysis. The design was adopted becausethe studyused secondary data collected from the Kano State Education Resource Department (KERD) and National Examination Council (NECO). It aimed at establishing how standard the items were.The population of the study consists of the fifteen thousand four hundred and eighty one (15,481) students who sat for 2015 economics SSCQE and 2016 SSCE in Kano State owned senior secondary schools. It consisted of ten thousand nine hundred and fifty (10,950) male students and four thousand and thirty one (4,531) female students. A sample size of 1,000 students was selected using multistage sampling technique. The data collected was coded and run using IRTPRO package to determine the SEM indices of Kano State 2015 SSCQE. Findings from the study revealed that the SEM of 2015 SSCQE and 2016 SSCE economics test items based on two-parameter logistic (2PL) model was high. This indicates consistency in measuring student ability in economics. The study found that the massive failure in Kano State Certificate Qualifying Examinations over the years was not caused by the quality of the SSCQE items but by other variables which are not related with psychometric properties of the test items. It was therefore concluded that there was no superiority in terms of item quality between the items generated by KERD. As such performance in SSCQE may determine performance in NECO. Given the various advantages of IRT over the other popular measurement frameworks, the government should encourage KERD to adopt this measurement framework. IRT can enable KERD to place examinees on the correct ability level to describe the test items and the abilities of the examinees. This will enable KERD to meet the best practices of global examination standard and overcome the lingering superiority complex of one examining body over the others. Keywords: Standard error of measurement, IRT, KERD, Qualifying, NECO, Economics