The aim of this research is to determine whether students' test performance and approaches to test questions change based on the type of mathematics questions (visual or verbal) administered to them. This research is based on a mixed-design model. The quantitative data are gathered from 297 seventh grade students, attending seven different middle schools in Cankaya, Ankara. Of all the students who participated in the research, qualitative data were gathered from 10 of them. In this research, seventh grade mathematics achievement test was developed by the researchers in visual and verbal forms. 10 of the students were selected and interviewed by utilizing semi-structured interview form. From the findings of the study, there was significant difference between the test scores and response time of the two forms in favor of visual form. The difference between item statistics is changeable in terms of the function of the visual. The results of the interviews showed that students have positive views toward visual questions. The students' perceptions of the use of visual in mathematics questions are examined in three main categories: preferability, comprehensiveness and responsibility of the questions.