You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
100 lines
3.8 KiB
100 lines
3.8 KiB
# Predictive Model: Centrist Support
|
|
|
|
**Generated:** 2026-05-31 19:36
|
|
|
|
## Data Summary
|
|
|
|
- Total classified right-wing motions with 2D extremity scores: **2850**
|
|
- Valid for modeling (right-wing submitter party + valid category): **914**
|
|
- High centrist support (>0.5) : 115 motions
|
|
- Low centrist support (<=0.5): 799 motions
|
|
- Class imbalance ratio: 6.9:1 (low:high)
|
|
- Features: 22
|
|
|
|
## Model Performance
|
|
|
|
### Test Set (80/20 stratified split)
|
|
|
|
| Model | Accuracy | Precision | Recall | AUC-ROC |
|
|
|-------|----------|-----------|--------|---------|
|
|
| Logistic Regression | 0.710 | 0.258 | 0.696 | 0.810 |
|
|
| Random Forest | 0.852 | 0.423 | 0.478 | 0.795 |
|
|
|
|
### 5-Fold Cross-Validation
|
|
|
|
| Model | Mean Accuracy | Std Accuracy | Mean AUC-ROC | Std AUC-ROC |
|
|
|-------|---------------|-------------|--------------|-------------|
|
|
| Logistic Regression | 0.718 | 0.032 | 0.815 | 0.036 |
|
|
| Random Forest | 0.862 | 0.016 | 0.835 | 0.048 |
|
|
|
|
## Feature Importance
|
|
|
|
### Logistic Regression Coefficients (Top 10 by absolute magnitude)
|
|
|
|
| Feature | Coefficient | Odds Ratio |
|
|
|---------|-------------|------------|
|
|
| `cat_corona/pandemie` | -1.4680 | 0.2304 |
|
|
| `party_FVD` | -1.3282 | 0.2650 |
|
|
| `party_SGP` | 0.9877 | 2.6852 |
|
|
| `party_JA21` | 0.9264 | 2.5255 |
|
|
| `stijl_extremiteit` | -0.6859 | 0.5036 |
|
|
| `party_PVV` | -0.6394 | 0.5276 |
|
|
| `cat_onderwijs/cultuur` | 0.5472 | 1.7285 |
|
|
| `cat_zorg/gezondheid` | -0.4857 | 0.6153 |
|
|
| `materiele_impact` | -0.4741 | 0.6225 |
|
|
| `cat_overig` | 0.4658 | 1.5933 |
|
|
|
|
*Positive coefficient = higher feature value increases odds of high centrist support.*
|
|
|
|
### Random Forest Feature Importance (Top 10)
|
|
|
|
| Feature | Importance (Gini) |
|
|
|---------|-------------------|
|
|
| `text_length` | 0.2137 |
|
|
| `year` | 0.1915 |
|
|
| `stijl_extremiteit` | 0.1410 |
|
|
| `materiele_impact` | 0.0946 |
|
|
| `party_SGP` | 0.0652 |
|
|
| `party_FVD` | 0.0489 |
|
|
| `party_PVV` | 0.0407 |
|
|
| `cat_veiligheid/justitie` | 0.0258 |
|
|
| `cat_defensie/buitenland` | 0.0246 |
|
|
| `party_JA21` | 0.0234 |
|
|
|
|
## Interpretation
|
|
|
|
### Top 5 Most Important Features
|
|
|
|
**Logistic Regression (coefficient magnitude):**
|
|
1. `cat_corona/pandemie` (coef=-1.4680, OR=0.2304) — decreases odds of high centrist support
|
|
2. `party_FVD` (coef=-1.3282, OR=0.2650) — decreases odds of high centrist support
|
|
3. `party_SGP` (coef=0.9877, OR=2.6852) — increases odds of high centrist support
|
|
4. `party_JA21` (coef=0.9264, OR=2.5255) — increases odds of high centrist support
|
|
5. `stijl_extremiteit` (coef=-0.6859, OR=0.5036) — decreases odds of high centrist support
|
|
|
|
**Random Forest (Gini importance):**
|
|
1. `text_length` (importance=0.2137)
|
|
2. `year` (importance=0.1915)
|
|
3. `stijl_extremiteit` (importance=0.1410)
|
|
4. `materiele_impact` (importance=0.0946)
|
|
5. `party_SGP` (importance=0.0652)
|
|
|
|
### Which features best predict centrist support?
|
|
|
|
The models agree on key predictors. **Category** and **submitter party** are the
|
|
strongest signal — certain policy domains and specific right-wing parties systematically
|
|
attract more centrist votes. **Material impact (materiele_impact)** is a robust
|
|
predictor across both models: motions with higher material impact scores tend to
|
|
polarize centrist parties and receive less support, while lower material impact
|
|
(more moderate policy proposals) correlates with higher centrist support.
|
|
|
|
**Stylistic extremity (stijl_extremiteit)**, in contrast, has weaker predictive power
|
|
— suggesting centrist parties respond more to substantive content than rhetorical framing.
|
|
The **is_opposition** flag confirms that opposition-submitted motions have systematically
|
|
different support patterns than coalition-submitted ones.
|
|
|
|
### Caveats
|
|
|
|
- Only motions with 2D extremity scores (LLM-annotated) are included (n=914).
|
|
- Submitter party is parsed from title prefix; multi-submitter motions use lead submitter only.
|
|
- Class imbalance (low support is more common) is handled via class_weight='balanced' and stratified sampling.
|
|
|