Designing a micro instrument for the measurement of learning climate useful to develop context questionnaires

Abstract

We work with an initial context questionnaire to measure the learning climate construct in two school samples corresponding to the 6th level of primary education and the 4th level of secondary education. The objective was to design a parsimonious micro-instrument which would be valid, reliable, and with enough capacity to discriminate among subjects and with substantive significance. For this reason we designed a shorter version of the original scale, applying the RAVACO reduction procedure. In the first stage of this procedure those key items with higher quality for the proposed objectives are selected. Analysis of missing values, exploratory analysis, cluster analysis and reliability analysis are used. In a second stage a global validation of the microinstrument is performed. As a result, two shorter versions are obtained and fitted to the small samples, noting the importance of missing values and merging the results of several statistical analyses carried out in the preceding five steps on R-AVACO.