TABLE 6
FROM:
'Do the Vote Test': The Electoral Effects of a Popular Vote Advice Application at the 2004 Belgian Elections
Stefaan Walgrave, Peter van Aelst and Michiel Nuytemans
BACK TO ARTICLETable 6. Multivariate analyses of the effect of DVT use on votes gain and loss for Belgian parties (cartels) between waves 2 and 4
| Gain | Loss | |||||
|---|---|---|---|---|---|---|
| Exp(B) of DVT use | R2 (of the model) | N | Exp(B) of DVT use | R2 (of the model) | N | |
| CD&V–N-VA | 1.61** | 0.080 | 2,029 | 0.64** | 0.068 | 2,039 |
| Groen! | 1.33* | 0.082 | 1,339 | 0.92 | 0.070 | 988 |
| Sp.a–Spirit | 1.25 | 0.126 | 1,552 | 0.81 | 0.053 | 1,848 |
| Vlaams Belang | 0.94 | 0.106 | 998 | 0.59* | 0.136 | 1,076 |
| VLD–Vivant | 2.04*** | 0.077 | 976 | 0.62* | 0.088 | 923 |
| Other parties | 0.80 | 0.123 | 99 | 0.58 | 0.234 | 119 |
Note: The coefficients represent standardized Beta's in logistic regression analysis models predicting joining or leaving a party. Other independent variables in the model (but not in the table): age, sex, education, interest in politics, left-right placement and the extent to which the campaign had been followed on television and in the newspapers. Sig. ***=0.001, **=0.01, *=0.05.
Source: UA Internet Panel 2004 (N=6,985).
