Explicit-reflective nature of science (NOS) instruction has demonstrated a positive impact on student learning. Although explicit-reflective NOS instruction often consists of questions that draw students' attention to NOS ideas, there are few recommendations in the science education literature about how the form of these questions might inform NOS educational methods and pedagogy. While some questions in the literature simply point students' thinking in a general direction, other questions require students to justify particular NOS ideas, or prompt them to choose between two positions. Given that NOS questions often seem to differ in the degree to which they direct student thinking, this study sought to examine the nature of student responses to different question types. Data was collected through writing. Four different versions of a reading were created with questions related to scientific method. Each version included a different question type: three drawn from Gallagher and Aschner's (1963) category system (evaluative, convergent, and divergent), and a general question that did not reference NOS. Readings were randomly distributed to 285 sixth grade students. Responses were analyzed using the provisional codes: normative, descriptive, misconception, off-topic. Using a Chi-squared test of independence and corresponding percentages, clear and statistically significant differences were observed in student responses to different question types. Drawing on the data from this study, convergent questioning seems more suited to the purpose of guiding students to an accurate conception of NOS, while divergent and evaluative NOS questions may make better assessment questions.