The analysis of psychometric properties of Student Opinion Scale

This article presents psychometric analysis of Polish version of Student Opinion Scale (SOS) by Donna Sundre (2007), which measures low stakes assessment test motivation. Adjusting by motivation the results of cognitive tests, which do not bear personal consequences to examinee is necessary to make valid inferences about students’ achievement. The analysis was performed using data collected from random samples of students from three different age groups - at the end of primary school (N = 1551), lower secondary school (N = 1724) and upper secondary schools (N = 16 271). Although one-level CFA results show the best fit of two-factor solution (as proposed in original version of SOS), low correlation of SOS score with the result obtained in the cognitive tests shows that SOS validity can be questioned. As social desirability and response style can influence the results - they need to be purified. Further analysis of validity, e.g multitraitmultimethod matrix is also needed. peerReviewed publishedVersion