Problem-Solving Ability as a Predictor of Academic Achievement in Mathematics among Higher Secondary Students: A Structural Equation Modelling Approach

Abstract

Background: Problem-solving skills have long been viewed as the core of learning mathematics, but empirical modelling of the multidimensional nature of problem-solving and its predictive role in mathematics achievement has been little evident in higher secondary education.
Purpose: This study hypothesises problem-solving skill as a second-order latent construct of numerical, logical, figural, and analytical dimensions and the predictive effect of the concept on the performance of mathematics in higher secondary students.
Methods: A cross-sectional design was used in this quantitative study. A standardised Problem-Solving Ability Test and a test on mathematics achievement were used to collect data on 300 higher secondary students in Tamil Nadu. The analysis of data was performed with the help of correlation analysis, regression analysis, Confirmatory Factor Analysis (CFA) and Structural Equation Modelling (SEM).
Results: A significant positive correlation was found between problem-solving ability and performance in mathematics (r = +.742, p <.001). The SEM model showed an excellent fit (kh2/df =2.28, CFI = .95, TLI =.94, RMSEA =.047). Adaptability to problems was an excellent predictor of mathematics success (b =.74, p <.001) and contributed 55% of the variance (R2=.55).
Conclusion: The results support the hypothesis that problem-solving skills are a hierarchical cognitive structure and a good predictor of mathematics performance among high-level school learners.
Future Scope: Future studies should consider the structural model by adding motivational factors such as self-efficacy and mathematics anxiety, and longitudinal designs to investigate causal associations.