Comparative evaluation of different methods for the assessment of attitudes to science

Abstract

In the field of science education a considerable decree of confussion has existed over specifying appropriate constructs and assessment techniques to ascertain measures of a pupil's attitude towards science. In a conceptual analysis of the items from a representative sample of attitude assessment instruments ten attitude dimensions were identified. These dimensions reflected personal aspects of science attitude and interest, difficulties with science, the social implications of science and the nature and working of science and scientists. These were examined empirically on a range of attitude measurement techniques identified from the review of the literature. These techniques included the Likert^Semantic Differential and Forced Choice formats with newly developed measurement instruments, in the form of a Free Response Structured and Open Response Situation questionnaires. A teacher - pupil assessment instrument based on a repertory grid technique was also developed for comparative purposes.
The empirical study involved 1200 pupils selected from the second year of secondary education. A precise and detailed programme of analyses was prescribed to allow comparative data to be reviewed. Each questionnaire was initially analysed independently. The overall analysis of tho results indicates that three perceptions are prominent to pupils of this age:
(a) Personal attitude toward science and interest in science related activities and careers,
(b) Personal characteristics of a scientist and
(c) The Social implications of science on society.
In the comparative evaluation of the different assessment methods the Likert questionnaire produced the most reliable and valid scales to assess these perceptions. The performance of tho Free Response technique indicated considerable potential for further development.