Introduction: C-Tests are employed globally to evaluate general language proficiency, and modifications of the typical C-Test have been proposed. Given the strong connection between language and mathematics, this study, besides investigating the difficulty of C-Test modifications and the impact of the test-taker’s gender, examines the relationship between modified C-Tests and mathematical performance across different content areas.
Methods: 190 seventh-graders were surveyed on their mathematical skills and given C-Tests with variations in gap placement (beginning vs. end) and topic (mathematical vs. general). Two evaluation methods were examined (correct/incorrect vs. word-recognition).
Results: GLMM analyses show that (a) the initial gap is the key factor contributing to test difficulty, (b) girls outperform boys, (c) boys benefit more from mathematical text topics, girls from gaps at the end of words, (d) arithmetic shows the strongest, geometry the weakest relationship to C-Test performance, (e) students with higher mathematical skills benefit significantly from mathematical text topics, and (f) students with lower mathematical skills benefit significantly from gaps at the end of words.
Discussion: Findings show that formal characteristics and gender impact C-Test performance and must be kept in mind when determining language proficiency through it. Furthermore, this study extends current research by demonstrating how specific language skills measured through modified C-Tests relate to different mathematical skills and therefore once more advocates the integration of language in mathematics education by presenting the C-Test as a possibility for this.