Based on 1988 census data for the 50 States in the United States, the correlation between the number of churches per state and the number of violent crimes per state was 0.85. What can we conclude?
A) The presence of a lot of churches in a state causes the number of violent crimes in the state to increase.
B) The correlation is spurious because of the confounding variable of population size: both number of churches and number of violent crimes are related to the state's population size.
C) Since the data comes from a census, or nearly complete enumeration of the United States, there must be a causal relationship between the number of churches and the number of violent crimes.
D) The relationship is not causal because only correlations of +1 or -1 show causal relationships.

Question

Quizplus · Accepted Answer

The correlation of 0.85 between the number of churches per state and the number of violent crimes per state does not imply causation. Instead, it suggests that both variables are likely related to a third variable, which is the population size of the state. Larger populations tend to have more churches and also report more violent crimes, making population size a confounding variable that explains the observed correlation.

Based on 1988 Census Data for the 50 States in the United