Exploring school students' views of the nature of science

This study aimed to develop, validate, and use a fixed-response test to assess in a quick manner the views of the nature of science (NoS) of groups of secondary school students and to explore the issues involved in developing such tests. The Nature of Science Test (NoST) used episodes from the history of science as contexts for questions. For each aspect of the NoS probed, three options were presented, using a cartoon format: a "desired" position and more positivist and relativist alternatives. The NoST was validated by an international expert panel and trialled with students in England (n=168). Only 50-60% of respondents gave a consistent response to the same question presented in different contexts.

To explore this further, parallel forms of the test - with different contexts - were administered several weeks apart to English students (n=169), while a test-retest trial using the same form of the NoST twice was conducted with a smaller sample (n=49). A sample of Mexican students (n=185) also completed one form of the test, to explore culture and language effects. Two focus groups in England and twelve in Mexico (n=6 and 36 students, respectively) probed the reasons behind students' views and checked the interpretation of their written responses.

Almost all students had mixed profiles, where desired views of some aspects of the NoS coexisted with alternative ones. Again, only 50-60% of respondents gave a consistent response to the same question across contexts, and test-retest variability was similar. In the focus groups, most students engaged adequately with the central issue raised by each question, justifying their responses coherently. English students selected slightly more "desirable" views than Mexican students, but differences in reasons for choices were negligible. Together, these findings suggest that students may posses an incoherent and unstable understanding of the NoS. A single administration of one form of the NoST does not appear reliable enough for the summative assessment of individuals, but can provide more reliable data at the population level. The quality of focus group discussions suggests that this technique could be used effectively for formative assessment within the classroom.