+265(0)111 624 222
research@unima.ac.mw
Chirunga-Zomba, Malawi
+265(0)111 624 222research@unima.ac.mwChirunga-Zomba, MalawiAuthor: Nkhoma, Mike Supervisor(s): Jimmy Namangale
Abstract
The study investigated the potential of raising validity evidence of Physical Science tests through teachers’ peer instruction. Of particular relevance in the study was the effect of teachers peer instruction on content and construct related validity evidence of Physical Science classroom tests. The study employed a mixed methods approach applying quantitative and qualitative methods. The quantitative methods used a pre-test and post-test one group experimental design, with peer instruction and validity evidence as independent and dependent variables respectively. Quantitative data was collected mainly from 62 tests with a total of 3016 items administered to 1543 learners of 2007 MSCE Physical Science class. The tests were constructed and administered by 17 Physical Science teachers to the class they were teaching. The teachers were purposefully sampled from Secondary Schools in the Southern Region of Malawi. Qualitative methods were applied to descriptive data generated through in-depth interviews with the teachers. Item relevance, item representativeness and item cognitive representativeness in tests teachers constructed before they attended peer instruction were compared with item relevance, item representativeness and item cognitive representativeness in tests they constructed after peer instruction. Similarly proportion of shared variance due to common factors was the attribute of interest for comparison for construct related validity evidence in the same tests. Content validity evidence of teachers’ tests constructed after attending peer instruction increased significantly in terms of item representativeness. The proportion of shared variance in tests constructed after attending peer instruction also increased significantly, an indicator that construct related validity evidence had increased. The implication of the findings was that there is potential of raising the validity evidence of Physical Science tests through peer instruction.
