School composition on student proficiencies The effect of student background and school composition on the static scores and gain scores of cognitive and non-cognitive proficiency tests in Dutch primary schools

Schools are often evaluated based on student proficiency. Evaluating schools, undistorted by external factors is preferable. We investigate whether student background characteristics and school composition have a different effects on cognitive and non-cognitive student proficiencies in value-added models using gain scores compared to static scores. To investigate which models are preferable static- and gain scores in mathematics, reading comprehension, and task motivation and self-efficacy data of 4,918 students at 216 primary schools in the Netherlands were used. Student background and school composition had a larger effect on static scores than gain scores, and are stronger for cognitive than for non-cognitive proficiencies, with student background having a larger effect than school composition. The results offer new arguments to use gain scores.