• If students circled a response, then you will need to tally the number of students who chose each of the responses represented. Then tally the responses to related questions and create a summary of responses.

  For example, the following four statements can be clustered into two categories.

  1. Science, as it is practiced in the real world, is objective and unbiased.
  2. Chemists work to uncover universal laws that already exist in nature.
  3. Chemists construct theories that explain what they observe in nature.
  4. It is important to be skeptical about the results of scientific experiments.
  The first two statements reflect an idealized view of science as truth. Whereas the last two statements are represent a more realistic view of the nature of science. Students' responses to question (1) and (2) versus (3) and (4) would provide a measure of the level of sophistication in students' views of the nature of science.

• If you are using a form that can be scanned, the resulting data can be presented in an electronic file. This data can then be entered into a data analysis program that will allow you to analyze the results of individual statements and aggregate the results across similar statements as above. Simple frequency distributions and percentage of responses in each category are a good place to begin to view your students perceptions in a variety of categories. For example, if most of your students agreed with the statement: "Doing labs in this class was like following a recipe in a cookbook." you might want to change your laboratory activities.

If you use the surveys as pre/post measurements, you may find a statistical program useful in measuring any statistically significant changes in your students' views or attitudes. This type of pre/post survey is particularly useful if you have incorporated curricular changes into your course. A simple comparison of your student means from each statement will allow you to see trends in your students' attitudes toward your course and its components. For example, assume that the mean for statement 11 in Table 2, " It was clear how the lab experiments fit into this course, " was low (less than or equal to 2 where 1=strongly disagree and 2=disagree). This would indicate that your students do not understand how the lab fits into your course. It's always a good idea to make friends with someone in science or mathematics education. A conversation with them about some of their experimental data would help with data analysis.

Commonly used statistical packages:

• Instructors may have an unpleasant surprise in discovering that their students' views are quite different, perhaps in direct opposition, to their own views or those of experts in the field.
• Some students may fear that their responses will not be anonymous and therefore be less candid than they might wish to be.
• The discovery of how students really feel about a given course can be depressing to instructors.
• The preferences and needs expressed by students may not fit well with the instructor's plans for a course. In this case, the instructor must either ignore students' expressed preferences or change components of the course.
• Surveys usually provide a broader range, but less detailed data than can be obtained from qualitative data such as individual interviews or focus groups.