Abstract
Literature describes word problems as an integral concept in mathematics education, as they are a strong tool for developing mathematical proficiency. Additionally, it acknowledges word problems as a complex and challenging mathematical concept for children. In response to the difficulties that children face with word problems, particularly additive word problems, interventionists globally have described the effects of different problem-solving instructions that have resulted in promising gains in learners' conceptual gains. Interventions such as the Schema-Based Instruction and Conceptual Model-Based Problem Solving have shown promising effects on problem-solving instructions that require structural reasoning. Regarding word problem-solving instruction in South Africa, the keyword approach has been a persistent method that has proven to be ineffective, especially when engaging with complex word problems. Employing a Mixed-methods Exploratory Sequential Design, confirms that firstly the ineffectiveness of this approach through a qualitative analysis of Grade 3 teachers’ practices (n = 5). Secondly, it shows the effects of the Equilibrated Development Approach on Grade 3 learners' (n = 140) conceptual development in compare problems. The results reveal what two other researchers have found in South Africa: an intervention that uses both structural and numerical reasoning to improve word problem solving improves the conceptual knowledge of the average learner in the class. Recommendations include future investigation that target lower performing learners in quintile 1-4 schools to understand why that is.