Data Dictionary: teacher_adoption_survey.csv
Binary outcome dataset for Bayesian logistic regression.
Use this dataset for binary-outcome teaching. It keeps the substantive focus on predicted probabilities rather than log-odds alone.
Research Scenario
A researcher studies whether teachers adopt a new instructional practice. Adoption may depend on years of experience, prior training, school support, and province.
Variables
| Variable | Type | Description |
|---|---|---|
teacher_id |
string | Unique teacher identifier. |
province |
categorical | Province code: A, B, C, or D. |
years_experience |
integer | Years of teaching experience. |
prior_training |
binary | 1 if teacher previously received related training, 0 otherwise. |
school_support |
integer | Perceived school support from 1 to 5. |
adopted |
binary | 1 if teacher adopted the practice, 0 otherwise. |
Main Questions
- How does prior training change the probability of adoption?
- How should log-odds, odds ratios, and predicted probabilities be interpreted?
- What is the posterior probability that training has a positive association with adoption?