| Title |
Hierarchical modelling of mathematical achievements / |
| Translation of Title |
Matematinių Pasiekimų Hierarchinis Modeliavimas. |
| Authors |
Šimonėlis, Edvardas |
| Full Text |
|
| Pages |
40 |
| Keywords [eng] |
Hierarchical data Structures, Trends in Mathematics and Science Study, hierarchical linear models, HLM, hierarchical logistic models, HGLM, TIMSS 2015, TIMSS 2019, HLM 8, socioeconomic status, urban vs rural schools, mathematics achievement. |
| Abstract [eng] |
In education, students tend to be more homogenous and share characteristics within their respected hierarchies such as schools, than the entire student population. Hierarchical modeling becomes the best solution in dealing with hierarchical data structures to account for the group effects that ordinary regressions tend to over/underestimate. This paper aimed to construct and analyze hierarchical linear and logistic regression models for mathematics achievement scores from Lithuanian 4th-grade data provided by IEA’s TIMSS 2019 using HLM 8 software. The analysis included socio-economic status and other significant independent variables (including school location, early skills and education, school climate and safety, and student attitude and instructional clarity). TIMSS 2019 models were compared with TIMSS 2015 data tested with the same variables. In the TIMSS 2019 models, we found that the mean socio-economic status of schools, school location, and school’s emphasis on success variables were significant and explained a large amount of variance for mathematics achievement at the school level. |
| Dissertation Institution |
Vilniaus universitetas. |
| Type |
Master thesis |
| Language |
English |
| Publication date |
2022 |
