Health | Research Methods | February 4, 2015

“Visiting Disneyland Causes Measles”… Um, No

Visiting Disneyland causes measles. Huh? Something doesn’t quite add up…. In this post, Stephanie Medley-Rath illustrates how visiting Disneyland has recently become correlated with contracting measles and uncovers the true culprit behind outbreak.

One hundred cases of measles have been reported in the United States in 2015.  News reports vary and the Center for Disease Control (CDC) data is a few days old, but anywhere from one-half to a majority of these cases are linked to the outbreak at Disneyland. Therefore, going to Disneyland causes measles.

You’re thinking, “no it doesn’t.”

But, these people would not have contracted measles if they had not visited Disneyland (or came into contact with someone who went to Disneyland). Therefore, Disneyland causes measles.

Ridiculous, right?

Visiting Disneyland does not cause measles. Visiting Disneyland in the past couple of weeks, however, is correlated with risk of contracting measles. Always remember, correlation does not equal causation.

A correlation means that a relationship exists between two or more variables. When you hear the word correlation think “co-” meaning shared and “-relation” meaning relationship. In this scenario, contracting measles and visiting Disneyland are correlated with one another. Further, January and 2015 are also correlated with contracting measles. What this means is that a person who visited Disneyland in January of 2015 is at a higher risk of contracting measles than someone who did not visit Disneyland during this time period.

How to Know if This Causes That

To establish causation, three conditions have to met:

  1. Correlation
  2. Time-order
  3. Non-spurious explanations

We’ve already established that a correlation exists, so next we have to check that the two variables are in the correct time order. That is, if we are going to argue that thing X caused thing Y, then X has to happen before Y. In this case a person visited Disneyland (X) and then contracted Measles (Y). So let’s move on to condition three, non-spurious explanations.

If we’re to say that going to Disneyland causes measles, then we have to rule out alternative explanations. In statistics, a spurious relationship exists when it looks like X might cause Y, but really it is some other variable Z that causes Y to change. For instance, violent crime increases in the summer and so too does ice cream sales, but the relationship between ice cream and violence is spurious. Violence increases in the summer because people are outdoors more often and interactions between people everywhere increases.

If Not Disneyland, Then What?

What non-spurious explanations could account for this measles outbreak? Experts point to declines in vaccination rates. Vaccines work when a critical mass of people are vaccinated. A critical mass matters as not all folks are able to be vaccinated due to underdeveloped immune systems (e.g., infants) or compromised immune systems (e.g., someone undergoing chemotherapy). A critical mass of vaccinated people protects all of those folks who are vaccinated and those who are physically unable to get vaccinated themselves (i.e., herd immunity).

Smallpox Vaccine

According to the CDC (2014), one in 12 children in the U.S. do not receive their first dose of MMR (measles, mumps, rubella) vaccine on time, though 91.9% of Americans have received the MMR vaccine.

Measles was declared eliminated in the U.S. in 2000. A record number of cases of measles were reported in 2014, with 593 cases reported as of August 8, 2014. Based on last year’s outbreaks, most cases of measles “have occurred in persons who were unvaccinated or had unknown vaccination status.”

Declining vaccination rates is correlated with increasing rates of measles. One could argue that declining vaccination rates caused this outbreak. We’ve established that declining vaccination rates is correlated with the measles outbreak. We’ve established time-order because declining vaccine rates occurred before the measles outbreak, not after. We’ve established that a spurious relationship does not exist-that is, a third variable is not the culprit. It’s perfectly reasonable to make the claim that declining vaccination rates caused an increase in measle rates.

Dig Deeper:

  1. What is the difference between correlation and causation?
  2. What factors are necessary to estabish causation?
  3. Visit Spurious Correlations. Select one of the graphs. Explain why a spurious relationship exists between the two variables.
  4. Why are vaccination rates declining in the U.S.? Read “With Measles Cases Surging, Why are Vaccination Rates Dropping?” Write a 4-6 sentence argument on how confusing correlation and causation contributes to one of the explanations offered in the article.

Read More: