Academic language features in mathematical modelling tasks raise difficulty in reading comprehension for secondary students

Abstract School-related reading requirements can be challenging and are entwined with the general ability to perform academically in subjects such as mathematics. However, it is empirically still unclear in how far linguistic requirements affect regular secondary school students ability to interact with subject content. The project reported here investigated the effects of academic German language characteristics on text comprehension difficulty of regular students in a German comprehensive school (“Gesamtschule”, 10 % heritage language users) in 25 text-heavy modelling tasks on mathematical functions. For each text, three versions with different amounts of academic lexical and syntactic features were constructed, while the content was not altered. N = 407 regular secondary school students (grades 7–10) solved text comprehension items specifically relevant to the mathematical task solution. General measures of language and mathematical proficiency and cultural capital were elicited. The main findings were that a), considerable numbers of students did in fact experience text comprehension difficulties of mathematically relevant information in mathematical modeling tasks; b), a systematic increase of academic language features in the task texts increased comprehension difficulty even when students characteristics were controlled for; c) higher grade level, mathematical and general language competence, and male gender, predicted strongly the ability to solve the comprehension items correctly; and d) number of books at home and language use at home did not moderate text comprehension after all other variables were controlled.