Comparing the epistemological underpinnings of students' and scientists' reasoning about conclusions