Math word problem solving by English learners and English primary students in an intelligent tutoring system

AbstractMath word problem solving in an online tutoring system was compared for high school students who were native speakers of English English primary and their peers who were learning English English learners. Word problems were written in English, the language of instruction. Data records for word problems that had been solved by students in both language groups were located and compared. Results indicated that the English learners were less likely to answer correctly, had more incorrect answer attempts, and took longer per problem on average than English primary students. When word problems were matched for math operation, students in both language groups performed worse on problems with more challenging text. There were no differences for the two language groups with regard to self-reported math motivation, plans to attend college, or off-task 'gaming' behaviour, suggesting that the lower performance of the English learners could not be attributed to lower effort.