The number of participants within the sample was 1546. The average age of the respondent is 39.01 years with a standard deviation of 13.472. The mean income of the sample is $8,000-$14,000 with a standard deviation of 3.40. (Descriptive Statistics can be found in Appendix A)

Within the regression model numerous variables were thrown out due to lack of significance. Among those variables were BlackRace, OtherRace, Age, Male, Employed, Unemployed, MOFA, FASMO, MOSFA, FaOnly, MoOnly, and Relative. All of the above variables are not significant enough to explain the variance within the model. The only three variables that remained in the model because they were significant were the respondent’s religiosity, level of education and income level. Under the Model Summary table, the respondent’s religiosity (ReligiosityIndex) is actually the variable that is the strongest predictor of the variance at 12.8%. The variances predictability actually increases to 16.5% when the respondent’s level of education is added with ReligiosityIndex. When the last significant variable, respondent’s income, is added into the model the predictability of variance increases to 16.6%. Looking at the adjusted R squared in the Model Summary table, which is adjusted to account for more complex models like what we have, the difference in percentage of predictable variance is still very close to the regular R squared. If you look at the Model Summary table under the R square change column you can see what percentage of predictability each variable adds to the total as it’s put into the model. For example when adding level of education (which accounts for 3.7% of predictability) to the 12.8% of predictability of the variable ReligiosityIndex, you get the total predictability of 16.6%. Continuing down the column will give you the amount of predictability that each variable adds to the model. As you can see adding in the last variable, respondent’s income, only accounts for .2% of the predictability of the index. While the income level is significant it doesn’t add much to the predictability.

Continuing onto the Coefficients table using a stepwise regression it’s found that ReligiosityIndex is still the most important variable accounting for the most variance while the variable Rincome accounts for very little. Something from this table that is particularly noteworthy is the negative relationship that both ReligiosityIndex and Educ have with our dependent variable. This being the case as respondent’s religiosity increases their likelihood of approving of suicide actually decreases. And as the respondent’s education level increases their likelihood of approving of suicide actually increases as well.