Table 1 Phi Coefficients for 2x2 Crosstabulations
|
|
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
1. Evangelical Bush voter |
--- |
|
|
|
|
|
|
|
2. Secure God's blessing |
.349 |
--- |
|
|
|
|
|
|
3. Increase belief in God |
.322 |
.454 |
--- |
|
|
|
|
|
4. Improve moral values |
.308 |
.407 |
.532 |
--- |
|
|
|
|
5. Ban gay marriage |
.271 |
.287 |
.296 |
.223 |
--- |
|
|
|
6. End limits on Ten Comm. |
.217 |
.279 |
.381 |
.248 |
.287 |
--- |
|
|
7. Ban abortion |
.137 |
.215 |
.253 |
.107* |
.386 |
.253 |
--- |
|
8. End school prayer limits |
.113* |
.189 |
.322 |
.226 |
.212 |
.375 |
.242 |
--- |
*Nonsignificant (p > .002)
The phi coefficient was chosen to express the strength of the association in each crosstabulation, both because the coefficient can be compared across tables and because, for two-by-two tables, it has the property of registering a negative value for negative associations. The coefficient's absolute value ranges from 0 to 1, with 0 indicating no association and 1 indicating perfect association. Table 1 presents the results. To hold the overall tolerable error of the analysis to .05 or less, the maximum p value for any one crosstabulation was set at .002. The phi coefficient for each crosstabulation was significant except for two: the crosstabulation of "ending school prayer limits" and being an evangelical Bush supporter, and the crosstabulation of "ban abortion" and "improve moral values." |