Diagnosing conceptions about the epistemology of science: Contributions of a quantitative assessment methodology

Abstract

This study applies a new quantitative methodological approach to diagnose epistemology conceptions in a large sample. The analyses use seven multiple-rating items on the epistemology of science drawn from the item pool Views on Science-Technology-Society (VOSTS). The bases of the new methodological diagnostic approach are the empirical psychometric scaling of the item sentences in accordance with experts’ criteria, the implementation of a multiple-rating model of answering (the respondent appraises each item sentence), and a scoring procedure that computes standardized indices from the multiple-rating responses and the sentence’s scaling. These standardized indices represent the extent to which the respondent's conceptions about the epistemology of science are informed (the higher the index, the more informed the conception). The approach is applied to evaluate the conceptions of a large and diverse national sample through its standardized indices, which provide statistical hypothesis-testing across individual sentences, items, groups, or studies, computation of reliability indices, the correlations between items, and an exploratory factor analysis that may complement qualitative analysis. Finally, the most relevant features of the new approach, its potential for applications to teacher training and curricular development in the science classroom and the method's power to make easy, quick, and economic evaluations of conceptions about the nature of science, are discussed.