Context questionnaire design for the evaluation of educational systems: optimization of complex constructs measurement

Abstract

This paper is a review of a possible model for design and development of context questionnaires for the educational systems evaluation. General stages are reviewed and identify the composition of possible indicators (simple and complex) of these instruments. In the design of questionnaires of context one must consider various aspects, some of which the most important are: a) the selection of variables and indicators to be considered as part of the model, and b) the way in which complex indicators, i.e., those that are the result of a scale or a combination of simple indicators (represented by a single item) are evaluated. To evaluate complex indicators (for example, the social climate of the classroom) there are good construct validity scales; however, they cannot be used completely, since they are made up of a large number of items. In the latter case, we include a procedure that we have developed to design micro-instruments that are composed of the best items of the scale (i.e., that best predict the total score, keeping the dimensional structure of the scale). These micro-instruments must be able to represent complex indicators so that they can be included in the questionnaires of context. It is a procedure of reduction of scales, aimed at identifying the best elements or items that best predict the overall score of a scale, while maintaining its theoretical structure.