| Effect Summary (χ² and Cramér’s V) | |||||||||
| Full vs Treated vs Matched | |||||||||
| Scale | Chi-squared (χ²) | df | P-value | Cramér’s V | N | Min exp. | Cells <5 (exp.) | Δ Minimal/No (n, %) | Significance |
|---|---|---|---|---|---|---|---|---|---|
| GAD-7 (full) |
8.305 | 3 | 0.0401 | 0.119 | 589 | 15.210 | 0.0% | 27 (16.3%) | * |
| CES-D 10 (full) |
11.471 | 3 | 0.0094 | 0.140 | 589 | 6.870 | 0.0% | 31 (27.2%) | ** |
| GAD-7 (treated) |
8.294 | 3 | 0.0403 | 0.118 | 591 | 15.210 | 0.0% | 27 (16.2%) | * |
| CES-D 10 (treated) |
11.412 | 3 | 0.0097 | 0.139 | 591 | 6.870 | 0.0% | 31 (27.2%) | ** |
| GAD-7 (matched) |
10.579 | 3 | 0.0142 | 0.143 | 516 | 13.500 | 0.0% | 36 (26.3%) | * |
| CES-D 10 (matched) |
11.614 | 3 | 0.0088 | 0.150 | 516 | 7.000 | 0.0% | 34 (37.8%) | ** |
| Significance codes: *** p<0.001, ** p<0.01, * p<0.05, † p<0.10 (marginal), ns = not significant. | |||||||||
| Cramér’s V = strength of before–after association in the 2×4 table (0 = no shift; higher = stronger). Heuristics in health/social data: ~0.10 small, ~0.30 medium, ~0.50 large. Min exp. / Cells <5 check χ² assumptions; a large share of cells <5 may weaken the χ² approximation. | |||||||||
| Δ Minimal/No (n, %) shows the net flow into the best category (After − Before), as counts and percentages. | |||||||||