Here’s their methodology :
Strayhorn compiled data from various data sets. The religiosity information came from a sample of nearly 36,000 participants who were part of the U.S. Religious Landscapes Survey by the Pew Forum on Religion and Public Life conducted in 2007, while the teen birth and abortion statistics came from the Centers for Disease Control and Prevention.
For religiosity, the researchers averaged the percentage of respondents who agreed with conservative responses to eight statements, including: ‘’There is only one way to interpret the teachings of my religion,” and ‘’Scripture should be taken literally, word for word.”
They found a strong correlation between statewide conservative religiousness and statewide teen birth rate even when they accounted for income and abortion rates.
And the questions they asked DO NOT pertain to Christianity ( as such) most of those statements for the “very relgious” would apply to Muslims as well. And it is known that the ROP encourages young marriage and lots of babies
This is a wonderful example of a meaningless correlation. To have any significance, we would need to know the birth/abortion rates for those people within each state who are defined as "religious" according the Pew Survey. We also need to know the ages of the teens and their marital status when they gave birth. No rational conclusion can be made from the correlation without this additional information.