Jumping to conclusions with CDC data
I have seen so much misinformation about the Center for Disease Control’s latest Youth Risk Behavior Surveillance System (YRBS) data that it’s time for me to speak up.
First, take a look at the numbers for yourself.
Now for some quick background:
Drug-legalization proponents and journalists who could use some lessons in interpreting scientific data have publicized claims the 2011 YRBS data don’t support. Among them is that adolescent marijuana use rates in Colorado have fallen since the start of the 2009 boom in dispensaries in this state. This is evidence, legalization supporters say, that regulated storefronts help keep marijuana out of kids’ hands.
Not so fast — and not by a long shot. Here is a short list of reasons why the CDC data do not support claims that adolescent marijuana use in Colorado is on the decline, much less that dispensaries have contributed to a fall in teens’ use of the drug:
- The data are statistically insignificant. If you understand what a “margin of error” is in political polling, then you understand statistical insignificance. As my wife, Christine Tatum, wrote about this CDC data on another post: “Folks who refer to these numbers as a ‘drop,’ or a ‘decrease,’ or a ‘fall,’ or a ‘decline’ demonstrate they know very little about how to analyze and interpret scientific data. Then again, these people may know exactly what they’re doing — and are just hoping the general public will fall for their deceptive ‘analyses.'” Similarly, Kevin Sabet, a world expert in drug policy, examined the numbers and wrote: “Almost all of the indicators for ALL drugs show an insignificant decline, meaning there are other reasons for the ‘no difference’ between 2009 and 2011 than marijuana dispensaries — unless legalization proponents claim dispensaries also make kids not drink, not smoke, not do cocaine, not do prescription drugs and not do other drugs. The bottom line is this: there is NO difference, AND the NO difference can be found in ALL drugs. The data then do not tell us anything.”
- The data were collected only a year from the start of the rise of marijuana dispensaries in Colorado. In prevention medicine, there are well known lag times between watershed moments and widespread public impact. Because the data in question were collected only a year from the start of the boom of marijuana dispensaries in Colorado, it is too soon to say how dispensaries have affected marijuana use rates (but if drug legalization supporters insist, they should also review a study I co-authored in the July edition of the Journal of the American Academy of Child & Adolescent Psychiatry.). Given the “prediction windows” that are fairly common and standard in prevention medicine, Colorado won’t begin to have a good idea of how dispensaries have affected adolescent use rates until at least 2013 or 2014.
- The data do not include teens who were not in school. It is important to note that Colorado has reported record drug-related school expulsions in the last four years. These expulsions are overwhelmingly attributed to marijuana use, officials at the Colorado Department of Education say. It is highly possible that teens who have substance problems did not participate in the CDC’s research.
- The CDC’s 2011 numbers could be a flatline because 85 Colorado communities voted to ban dispensaries the previous year, 2010. We could be looking at steady numbers precisely because Colorado voters shut down pot shops, making marijuana less accessible and socially acceptable.