SOURCE:

Faculty: Education
Department: Educational Foundation

CONTRIBUTORS:

Agbonkpolo, U.M;
Esomuno, N.P.M;
Okoye, R;

ABSTRACT:

The main purpose of the study was to find out experimentally whether the segmentation marking method (SMM) of scoring essay tests would be more effective than the whole script marking method (WSMM) with regard to the minimization of systematic errors, random errors and marking time. Thus the standard error of measurement (SEM), the average mark change (AMC) and the average marking time (AMT) were used to measure the effectiveness of the two methods. The post test – only control group design, in which two equivalent groups are subjected to different treatments, was used in the study. The two marking methods (treatments) were explained and randomly given to two equivalent groups, each consisting of 24 NECO markers (one group to one method), to mark thirty two examination scripts in a conference marking. After the marking the SEM, AMC and AMT were computed. Seven hypotheses were formulated and tested at 0.05 alpha level. The statistics used were the t – test, Pearson correlation, and regression statistics. The findings showed that both groups had similar levels of SEM (random errors) in the scores they awarded. Marking experience, tend to reduce SEM in SMM though the effect was not significant. However as marking experience increases, SEM reduces in SMM and increases in WSMM, making the former have less SEM than the latter among markers with more than eight years of marking experience. Finding shows that SMM had slightly higher AMC (random and systematic errors) than WSMM unlike the parity in their SEM. This shows that the systematic errors were slightly more in the SMM than in the WSMM, probably due to the fact that if a script was over - rated or under - rated with the SMM, this error was exported to subsequent scripts, because of the direct comparative approach. Poor examinee’s hand writing tend to increase AMC in SMM more than in WSMM, which implies that examinees’ handwriting was more influential in the direct comparative scoring of their responses, than in whole script scoring, although the difference was not significant. And as examinees’ handwriting clarity increases, AMC becomes smaller in SMM than in WSMM. Finding also shows that SMM had greater AMT than WSMM though the difference was not significant. SMM markers complained about their method; this confirmed that they were not used to SMM. Based on the above findings and the previous findings that training matches experience in marking reliability the researcher suggested that intensive training in the use of SMM be given to less experienced markers, to put them in the same pedestal with the more experienced markers, among whom the SMM was shown to be more reliable than the WSMM; for the realization of the hope of a better marking reliability with the use of SMM. This could be boosted by warning examinees against illegible handwriting to minimize systematic error.