Fast food and grades: correlation is not causation

30 Dec

Please repeat after me “correlation is not causation,” “correlation is not causation.” This mantra is one of the first things we teach in university statistics courses. Yet, it appears to be the first thing people forget when they graduate, especially journalists and advocates.

Take a look at this headline: “Eating Fast Food Hurts Children’s Ability To Learn”  or this one “The food that drags down American kids’ test scores.” Now, I think a fast food diet is terrible for adults and children alike, but how good is the evidence for this claim? It is based on a paper published in Clinical Pediatrics. Here is the abstract of that paper:

“Objective. The objective of this study is to examine the associations between fast food consumption and the academic growth of 8544 fifth-grade children in reading, math, and science. Method. This study uses direct assessments of academic achievement and child-reported fast food consumption from a nationally representative sample of kindergartners followed through eighth grade. Results. More than two thirds of the sample reported some fast food consumption; 20% reported consuming at least 4 fast food meals in the prior week. Fast food consumption during fifth grade predicted lower levels of academic achievement in all 3 subjects in eighth grade, even when fifth grade academic scores and numerous potential confounding variables, including socioeconomic indicators, physical activity, and TV watching, were controlled for in the models. Conclusion. These results provide initial evidence that high levels of fast food consumption are predictive of slower growth in academic skills in a nationally representative sample of children.”

In other words, the researchers found a negative correlation between fast food consumption and academic performance. But this does not mean that eating fast food causes low grades. For example, fast food consumption might be a proxy for some other variable, such as socioeconomic status. Perhaps children who are poor have both terrible diets and do poorly in school.

But, you may protest, the study controlled for those confounding variables! Well actually, they didn’t, they used statistical techniques to try to control for confounding variables, but these techniques rests on certain assumptions that probably are not met in the real world. Everyone who uses research should read this paper “Risk factors, confounding, and the illusion of statistical control”

I do not fault the researchers, they have alerted us to a possible relationship between fast food consumption and academic achievement, but additional research will be needed to either verify or reject this hypothesis.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: