This project use classroom.dta dataset provided by Professor Marc Scott.
The classroom dataset has three levels of nesting: schools, classrooms within schools and students within those classrooms. In this sample, there are 107 schools, a total of 312 classrooms across all schools, and 1190 students total. Within a school, there are between 2 and 31 students sampled.
| Variable Name | Label | --- | ---| --- | Sex | Student Gender (0/1) | | Minority | minority (0/1) | | Mathkind | math score in spring of kindergarten| | Mathgain | math score in spring of first grade | | ses | ses | | yearstea | Teachers' years of teaching | | mathknow | Teachers' math knowledge | | housepov | Average household poverty | | mathprep | Teachers' math prepartion (#courses)| | classid | Class ID | | schoolid | School ID | | childid | Child ID |
This project starts with fitting an unconditional means model with school-specific random effects, which can be simply expressed as: