By Anigbo, L .c
This study is concerned with the development and standardization of a set
of test batteries captioned Mathematics Achievement Test Batteries
(MATBAT). The batteries are based on the primary school curriculum
modules for primary 4. Adopting Instrumentation Design, an initial pool of
two hundred and seventy (270) items was constructed and trial tested in
nine (9) sets employing the partial multiple matrices sampling design.
After the trial testing, one hundred and thirty five (135) items survived
the item analysis and were composed into three sets that were used for the
final test administration employing the genuine multiple matrices
sampling design. Both the trial testing and the final administration were
carried out with primary 4 pupils randomly drawn from primary schools
in Enugu State. Two hundred and seventy (270) pupils were used for the
trial testing while nine hundred (900) pupils were used for the final
administration. Five research questions and three hypotheses guided the
study. The research questions were answered by descriptive statistics
while the null hypotheses were analyzed using dhe analysis of variance
(ANOVA) statistic. The findings of the study inchide: MATBAT, made up of 135
n~ultiple-choice items con~posed in three sets with the following psychometric
characteristics: difficulty indices ranging from 0.25 to 0.75, discrimination indices
ranging from 0.1 3 lo 1.0, distractcr indices of options ranging fro111 0.1 to 0.6. The
Kuder-Richardson (K-R 20) reliability indices of the subsets range from 0.68 to 0.93. The contents of the modules in set 1 and set 2 werc all positively correlated to one another and the correlations were all significant at the .05 levels. Some n~odulesin set 3 were negatively correlated to one another; and there were systematic decreases in the mean scores in the modules towards the end of each set. These may suggest that Set 3 is factored while the length of the tests may account for decreasing mean achievement scores towards the end of the sets. Statistically significant difference (p.05) in the mean scores of the pupils from the different schools ownership types were detected. The difference was in favour oC privately owned schools. The means achievement scores of the pupils in Set 1 was significantly highcr 0?=.05) than that in Set 3 while the difference between the mean scores of Set 1 and Set 2 was not significant. NJoreover the pupils' mean score in Set 2 was significantly higher than that of Set 3. The interaction of ownership and sets in the mean achievement scores were not significant ((p=.05).