Misconceptions held and errors made by South African learners in answering science questions in the trends in mathematics and science study (TIMSS).
This thesis presents results of a secondary analysis of South African Grade 8 learners' responses to the TIMSS 1999 Population 2 Science questions, and the results of a test and group interviews with learners from three high schools in Pietermaritzburg. The study is motivated by the need to understand the reasons for the poor performance of SA Grade 8 natural science learners in the TIMSS 1999 study. The aim of the study was to identify possible misconceptions held and errors made by SA science learners and to pinpoint pos sible causes and sources of these misconceptions and errors. An analysis of the SA learners' responses contained in the TIMSS 1999 Population 2 data and the test results from the Pietermaritzburg schools was used to identify possible misconceptions held and errors made by SA learners on certain science topics. The questions used in the Pietermaritzburg study were selected from the TIMSS 1999 population 2 science quest ions. The questions were selected based on perceived minimal levels of guessing by learners when answering these questions in the TIMSS 1999 study. The three schools selected for inclusion in this study in Pietermaritzburg were chosen based on their former classification as Indian, Coloured and Black schools. The majority of pupils at each of these three schools still reflect this racial composition that was in place at the time the TIMSS study was conducted. The Pietermaritzburg data was collected by administering a written test consisting of selected TIMSS 1999 science questions to a randomly selected group of ten learners at each school follow ed by interviews with the learners and the head of the Science Department at each school. Learner ' notebooks and the year planners used by teachers at the three schools were analysed primarily to check if these were in line with the SA intended curriculum and to check for any possible sources of misconceptions. The analysis of the data was guided by the constructivist theory and the conceptual framework used in the TIMSS data. The TIMSS conceptual framework focused on the intended, the implemented and the achieved curriculum. The study found that learners hold misconceptions on different concepts of the natural science curriculum. Some causes of misconceptions and errors among learners were identified. Some of the misconceptions highlighted indicate for instance that learners believe that boiling water is a form of a chemical reaction; a seed develops from a root; proteins are the same as vitamins, proteins and vitamins provide energy for the body. Some possible causes of misconceptions and errors among learners were identified. These possible causes of misconceptions and errors are varied with some for example, linked to poor content coverage by teachers, learners' everyday experience, the lack of learning resources and materials at some schools.