"Data suggests 9 months after a Super Bowl victory, winning cities see a rise in births." That's the claim in a new NFL commercial. But how does this assertion hold up from an EVERYDATA standpoint? Here are a few questions to ask:
- How much of a rise in births was there? The commercial doesn't tell us if it was 1 extra kid or 1,000 (although from the number of kids in each scene it's implied there were at least a dozen or so per city). Specifically, we'd want to know if there were enough extra kids born per year to be statistically significant.
- Was the supposed rise in births caused by a winning Super Bowl team, or is it simply a correlation? In situations like this you might want to look for omitted variables. For example, there's data to suggest that "low-severity hurricanes and tropical storms" can increase the birth rate 9 months later. Were there weather issues (or other factors) in the winning cities that could have contributed to the supposed rise?
- Was the increase the same each year? Were there even more babies born when teams won by a larger margin? Was there more of an increase in cities that revolve around football (we're looking at you Green Bay)?
- How precise is the data? The most common length of a pregnancy is 39-40 weeks (it actually decreased in recent years due to medical advances in fetal monitoring). But more than 27% of pregnancies are either shorter or longer. Even for those that fall within the 39-40 week range, this level of precision would be difficult (if not impossible) to achieve—which means we probably don't know for sure that these kids were conceived the same day as the game. Which leads us to our next question...
- Can we rely on self-reported data? Even if the NFL checked birth certificates for the Super Bowl babies (and we're not sure that they did), given the imprecise nature of the data we're forced to trust the parents who said they also scored the night of the big game. Self-reported data, however, is not always the most reliable.
- Was there also a rise in births in the losing team's city? What about cities that didn't have any affiliation with the teams? Is the NFL simply cherry picking data from winning cities to make a fun commercial?
- Finally, how do you define the city? Did the NFL only look at parents who lived within the city limits? Was the data only from hospitals within each city? Or did they expand it to include the surrounding suburbs, which (we presume) are also full of fans?
What other questions do you have about the NFL's data? Let us know in the comments!