Data Dictionary: education_intervention.csv
Main teaching dataset for regression, prior specification, posterior interpretation, and hierarchical modeling.
Use this as the main workshop dataset. It supports the prior, regression, posterior interpretation, hierarchical modeling, and reporting modules.
Research Scenario
An education researcher evaluates whether a learning intervention improves student endline achievement after accounting for baseline achievement, socioeconomic status, attendance, and school-level clustering.
Variables
| Variable | Type | Description |
|---|---|---|
student_id |
string | Unique student identifier. |
school_id |
string | School identifier. Used for multilevel modeling. |
school_context |
categorical | School context: urban, peri_urban, or rural. |
intervention |
binary | 1 if the student received the intervention, 0 otherwise. |
gender |
categorical | Student gender in the simulated data. |
ses_index |
numeric | Standardized socioeconomic status index. Higher values indicate higher SES. |
baseline_score |
numeric | Pre-intervention achievement score from 0 to 100. |
attendance_rate |
numeric | Attendance percentage. |
endline_score |
numeric | Post-intervention achievement score from 0 to 100. |
Main Questions
- What is the posterior distribution of the intervention effect?
- What is the probability that the intervention improves scores?
- What is the probability that the intervention effect exceeds a practical threshold, such as 3 points?
- How much do school-level estimates vary?
- How does partial pooling change school-level conclusions?