Abstract
This study was carried out to investigate the effect of test items practice and test item compromise on mathematics achievement and retention among secondary school students in Ardo-Kola local government area of Taraba State. Non-equivalent posttest control group of quasi experimental design was adopted for the study. Forty students were selected from two secondary schools to participate in the study. One standardized instrument was used for data collection. The instrument is post-test mathematics achievement (POMA). Two research questions were stated for the study and two hypotheses tested at 0.05 level of significant. The reliability co-efficient of 0.72 was established for the instrument using split half method. Data were analysed using mean, standard deviation and analysis of co-variance. Result from the study revealed that the items practice group performed higher (mean= 78.80, SD= 22.66) than the items compromised group (mean= 37.10, SD=13.91). Also, the retention score of the test items practice group is higher (mean= 81.15, SD=19.26) than the compromised group (mean=44.15, SD=10.68). There was also a significant difference in the academic achievement and retention scores of test items practice group and compromised group. Based on these findings, mathematics teachers and school administrators are to pay more attention to frequent practice of mathematics items at school and home to enhance students' retention and achievement than their involvement in compromising test items during examinations in our schools.