作者: Patricio Troncoso , Maria Pampaka , Wendy Olsen
DOI: 10.1080/09243453.2015.1084010
关键词:
摘要: ABSTRACTSchool value-added studies have largely demonstrated the effects of socioeconomic and demographic characteristics schools pupils on performance in standardised tests. Traditionally, these assessed variation coming only from pupils. However, recent shown that analysis academic could significantly benefit additional complexity model structure, incorporating non-hierarchical unexplored levels variation. Using data secondary students Chilean National Pupil Database (2004–2006), this study shows how traditional models fall short addressing complex phenomenon performance, because they overestimate school effects. A 4-level contextualised for progress Mathematics was implemented to avoid masking classroom locality found models. We also analyse important structural fa...