An investigation of how multilingual learners process questions during evaluations

Abstract

In a country like South Africa where multilingualism is the norm, multilingual language development is still not well understood. This then raises the question of how educational practices can best scaffold learning for multilingual children. As it stands a great number of school-going children in South Africa continue to be taught through multilingual language practices in class and then during formal assessments and examinations, they are evaluated in the second or third languages. Implying that learners have to master complex concepts in a language that is still being acquired. This can result in difficulties in learners’ abilities to use language for learning, including their comprehension and answering of questions. Consequently, this study focuses on how a sample population of multilingual 10–11-year-olds (primary school, grade 5) and 14–15-year-old (high school, Grade 9) learners from Mamelodi and Nellmapius townships in Tshwane, Gauteng, process questions during evaluations. A mixed methods design was used to investigate their processing of questions (which includes both the answering and comprehension of questions). The study applied the question comprehension theory informed by several question processing models (such as Robertson’s (1994) theory of simultaneous understanding, answering and memory interaction model, and Pollitt and Ahmed’s (1999) Question-Answering process model), and the bottom-up framework (Kempen, 1999) in the analysis of learners’ question processing. As a framework for language processing, the bottom-up approach presented a staged process of how an individual can construct meaning and arrive at the answer. The results of the study highlight the ways that learners of different ages manoeuvre in the process of answering questions that demand different cognitive skills in their second and third languages. Results of the study showed that while all learners were able to provide a response to the questions. However, not all responses were accurate. This indicates that there is a breakdown in the processing and comprehension of the question. Moreover, there was no difference between the 10–11-year-olds and 14–15-year-old learners in processing / answering of knowledge and comprehension questions