2003 - The Effect of Schools and Classes on Language Achievement

Attribution: De Fraine, Bieke, Van Damme, Jan, Van Landeghem, Georges, & Opdenakker, Marie-Christine
Researchers: Bieke De FraineGeorges Van LandeghemJan Van DammeMarie-Christine Opdenakker
University Affiliation: K.U. Leuven
Email: bieke.defraine@ped.kuleuven.be
Research Question:
The effect of schools and classes upon literacy in the second grade of secondary education.
Published: 1
Journal Name or Institutional Affiliation: British Educational Research Journal
Journal Entry: Vol. 29, No. 6, pp. 841-859
Year: 2003
Findings:
  • The results of a three-level analysis (students within classes within schools) indicate that the group composition at the class level is very important.
  • In classes with a high average initial cognitive ability or a large proportion of girls, the language achievement is higher.
  • The analyses show that group composition is more important than the educational practices in accounting for differences in language achievement.
  • The learning climate is more positive in classes with a higher cognitive level indicating that a cognitively strong class group is mostly a calm, cohesive and study-oriented group.
  • When individual student characteristics are taken into account, it does not matter if teachers focus on individual development , pay attention to high and low achievers, give feedback or consult colleagues. Even the opportunity to learn cannot be seen as an effectiveness-related characteristic.
  • Girls perform better than boys and it is beneficial for all students when the percentage of girls in their classroom is high.
  • By exploring the school-level variables, we concluded that the difference between public schools and other schools is the most informative.
  • Looking at the unadjusted school effects, Catholic schools do better than public schools. This is because Catholic schools recruit higher-achieving students.
  • The class an the teacher are especially important for low achievers.
  • Results indicate that school effects do exist and that the differences between schools are large. The between school variance for language achievement is 29%. And when variations in the characteristics of students are taken into account, substantial school differences remain.
  • The student background characteristics and prior attainment explain more than 70% of the variance at the two higher levels. Teacher and classroom account for more or less the same amount of variance in language achievement as the school, even after adjusting for intake.
  • The class and teacher have a strong impact upon the students’ achievement, probably because the class is a grouping structure that is closer to the student and the learning process than the school.
  • Students benefit from learning Dutch in classes with a high average initial cognitive ability and a large proportion of girls.
  • School composition characteristics are strongly related to school process variables such as cooperation between teaching staff and orderly learning environment at school (Opdenakker & Van Damme, 2001). Schools with high social class or high ability intakes are more likely to have greater support and involvement from parents, fewer disciplinary problems and an atmosphere conducive to learning (Wills, 1992; Grisay, 1996).
  • In educational effectiveness research, neglecting group composition may lead to incorrect conclusions. Incorporating group composition variables yields more valid to incorrect conclusions. Incorporating group composition variables yields more valid results about differences between classes and between schools and about correlates of effectiveness.
  • It is reasonable to conclude that the language achievement is affected by the proportion of girls and the average cognitive ability in the class group.
  • Language achievement is slightly better in public schools than in private schools, at least when a correction is made for differences in student intake and class composition. The positive climate in Catholic schools may be largely due to a favorable student population. Research into the differences between public and private education should take into account differences in group composition.
Keywords: Academic AchievementCognitive AbilityGenderLanguagePrivate SchoolsPublic SchoolsTeachersTrackingRegions: InternationalMethodologies: QuantitativeResearch Designs: Secondary Survey DataAnalysis Methods: Multilevel Models Sampling Frame:Students in Private and Public schools in Belgium
Sampling Types: NonrandomAnalysis Units: ClassroomSchoolStudentData Types: Quantitative-Longitudinal
Data Description:
  • The LOSO-project (Longitudinal Research in Secondary Education).
  • The analysis of the language achievement was performed using a subsample of 2569 students in 152 Dutch classes in 55 schools.
  • In this data set, there are 16 public schools, 38 Catholic schools and one city school.
  • In the large sample, there are 155 Dutch teachers and 275 second grade classes, so some teachers taught more than one class. Therefore the data have in fact a four-level structure: individuals within classes within teachers within schools.
  • DV: Score on the language test at the end of the second grade when the students were approximately 14 years old.
  • IV: Student characteristics (initial cognitive ability, socioeconomic status of the family, achievement motivation, immunity to stress, gender, and language spoken at home., Class variables (mean initial cognitive ability, mean SES, mean achievement motivation, mean immunity to stress, proportion of girls in the class, and proportion o f students who speak Dutch at home), teaching practices (learning climate, special attention to low achievers, special attention to high achievers, and feedback on study results), school variables (school averages of the six student variables to describe the student population of a school, etc.).
Theoretical Framework:
Relevance:
Archives: K-12 Integration, Desegregation, and Segregation Abstracts