Stats Behind the Belief System

Factor Analysis of Pro-Death Penalty Opinions

Factor analysis is an advanced statistical technique for grouping or "clustering" opinions, based on how people replied to questions about their opinions. The left column shows 15 opinions that people in the Death Penalty Study either agreed or disagreed with. The column labeled 'Death Penalty Moral & Just' reflects a cluster of pro-death penalty opinions that make up a belief system. People who agree with any one of these 7 opinions (salmon background) are also likely to agree with the other 6.

The numbers in the column are called factor coefficients. These numbers range from .00 to .99. They can be positive or negative. For example, the opinion that 'The death penalty is the only just punishment for murder' is the 'best' measure for this system of beliefs.


Note the 2 negative numbers (white background) in the column. The negative numbers mean that those agreeing with the 7 pro-death penalty opinions tend to disagree with these 2 opinions. This makes sense, since these 2 opinions are anti-death penalty opinions.


The number at the bottom of the column is Cronbach's reliability coefficient. This number ranges from .00 (very unreliable) to 1.00 (perfectly reliable). The .80 here indicates that these 7 items are generally reliable measures of pro-death penalty beliefs. For more technical information about the full factor analysis conducted in the study, click here.


NEXT>