The use of data analytics in predicting student success in first-year courses has been gaining momentum in South Africa. With that, various assessment data points are used to map out incoming students' academic profiles and help design curriculum support mechanisms. Scores on the highstakes assessment subdomains could be used to explain the variance in course performance and suggest that subdomains with the greatest relative importance have the strongest association with the variance in performance in a course. Calculating subdomain scores (subscores) and establishing the meaningful weighting of each subdomain in a test is important. This study compares how different subscore calculation methods, such as IRT and number-correct, affect reporting and the ranking of subdomains in a relative importance analysis of a first-year mathematics cohort.