In Defense of the Right to Life: Analyzing Factors Affecting Filipino Opinion About Death Penalty Table 4 in the annex presents the results of the logistic regression models to determine the factors that predict support for the death penalty. Model 1 examined the influence of demographic factors, model 2 examined the influence of TV news access and internet use, model 3 examined the influence of the influence of instrumental factors, model 4 examined the influence of symbolic factors, and model 5 examined the influence of attitudes to alternatives to the death penalty. Overall, the strongest predictors of support for the death penalty were urbanicity, sex, location, level of education, feeling of safety with the death penalty, worry of being victimized by a crime, following the law over religious principles, opinion that the death penalty serves and provides speedy justice, and preference for the death penalty over other forms of punishment. Across all models, support for the death penalty was associated with being from urban areas and being male. In terms of location, support for the death penalty was found to be significantly lower in Balance Luzon and Visayas than in Mindanao. The level of education likewise has a significant direct effect on support for the death penalty – support was lower for those who only reached or completed elementary of high school than those who are college-educated. Contrary to expectations, religion and religiosity did not influence one’s likelihood to support the death penalty; this was observed for all models. Results from model 3 shows that the knowledge-based instrumental predictors, rather than experiential, were the main predictors of support for the death. Those who would feel safe with the death penalty were 4.3 times more likely than those who do not feel safe to support the death penalty. Those worried of being victimized by a crime were 1.9 times more likely than those not worried to also support the death penalty. The influence of feeling safe with the death penalty was reduced in models 4 and 5 (2.7 times more likely) when the set of symbolic attitudes were added. TV news media usage was only found to be a significant predictor of support in model 4, when the symbolic factors were added to the equation. This raises the possibility that TV news media usage mediated attitudes related to the justice system. Internet usage did not have an influence, and this was observed for all models. 14