Saturday, November 1, 2014

New York Times Attacks Marijuana Legalization--Again

What is it with these establishment newspapers? It was only a fortnight ago that I responded to the Daily Telegraph's  diatribe against marijuana legalization. The Telegraph published some stale data from Wayne Hall and presented it as a new, 20-year study of the effects of marijuana. The "new" study came to the sam conclusions as Hall came to 20 years ago.

New studies on marijuana have not been influenced by old prejudices. The Big Lie is that, since marijuana is illegal, it must be harmful. Since marijuana was made illegal to control African-American recreational users. The first head of the Federal Bureau of Narcotics, Harry Anslinger, was a racist who used racial slurs to get marijuana criminalize in 1937. Anslinger argued that drugs should be outlawed based on the "undesirable" people who use them:
"There are 100,000 total marijuana smokers in the U.S., and most are Negroes, Hispanics, Filipinos and entertainers. Their Satanic music, jazz and swing result from marijuana use. This marijuana causes white women to seek sexual relations with Negroes, entertainers and any others."
Despite the overwhelming evidence that Marijuana is not a gateway drug and does not cause insanity, here comes Abigail Sullivan Moore, writing in the New York Times, that a new study has shown that marijuana use causes changes in the brain.

Moore entitles her article, This is your brain on drugs, apparently unaware that the "Just say no" campaign from which this slogan is taken was simply relying on ingrained prejudices rather than actual scientific evidence. The advertisement that used this slogan was an attempt to scare people by showing an egg frying in a frying pan and implying that somehow, evidently magically, marijuana turns your brain into a fried egg.

What about this new study, so ominously accompanied by pictures of a human brain on which yellow and red circles, strangely reminiscent of the old advertisement, are superimposed. The pictures themselves show nothing, but the circles are certainly scary. The study was published in the Journal of Neuroscience on April 16.

Criticism of the Study predates the Times article by six months

On the next day, April 17, Lior Pachter, a professor at UC Berkeley with a PhD in mathematics, published a post on his blog calling this study, "quite possibly the worst study I have read all year."

Pachter went on to make multiple criticisms of the study:

  • While the author of the study, Hans Breiter, claimed to be a mathematician, in reality he had only taken a couple of courses in mathematics.
  • Breiter lied when he claimed that some of the study subjects used marijuana only once or twice a week. In fact, the majority were heavy users of between 10 and 30 joints per week.
  • The number of users studied was 20 (a very small sampling, considering the millions of marijuana users in the US alone) and there were no long-term studies done to verify Breiter's conclusions.
  • There was no evidence to back up Breiter's claim that casual users were having problems.
  • The experiment consisted of a single MRI scan for each subject, which Pachter considered inadequate, given the variability of MRI scans.
  • The study did not calculate separate p-values for each of 4 tests, but simply used one value for all. A p-value is used to determine whether the hypothesis of the experiment--in this case, that marijuana smoking induces changes in the brain--is correct. In other words, the study invented a p-value that did not exist. Since the p-value determines whether the hypothesis is true, or even likely, this carelessness with statistical methods should invalidate the entire study. Pachter calls the methods of the study, "unbelievable".
Pachter was not alone in his criticism of this study. Ryan M. Smith, PhD in Neuroscience, Ohio State University, criticized the study for its repeated statements that there is a causative relationship between cannabis use and anatomic changes in the brain, when they state clearly at the end of the article that the experiment does not prove any causative link.

Nidia J. Melendez, Research Assistant at Columbia University, makes the obvious point that there are no measures of cognitive behaviors or any other behaviors in the report. The authors make no attempt to connect their observations of changes in the brain to any real-world effects. Also, she says, there are indications that the marijuana users have used other drugs, but the study makes no attempt to distinguish the effects of marijuana and other, unknown drugs. In other words, this supposed scientific study draw no conclusions about marijuana use because it did not exclude other drugs from the study.

What is the New York Times Thinking of?

The article, This is your mind on drugs, has nothing of value in it. All of its conclusions are based on a single study. Its author, Abigail Sullivan Moore, has no credentials for writing a review of a scientific article. She has primarily worked as a public relations writer for an insurance company. It is not necessary for a reporter to have extensive experience in the field she writes about, but she should at least pay close attention to the subjects she writes about.

In particular, Moore writes this article for the Times six months after the study was published. Since that time, experts in statistical analysis and physiology have severely criticized the conclusions of the article. But Moore apparently did not bother to consult other sources before writing. Although she is not a scientist, she should have some understanding of common practices in the scientific field, including the practice of other scientists reviewing publications.

She should also know that a single study cannot be widely accepted until other studies have been conducted that corroborate its findings. Nevertheless, Moore proceeds as if this study proves something that it does not.

The fault her is not Moore's, however. No doubt she is doing the best she can with limited training and only a layman's understanding of her subject matter. The fault belongs to the editorial staff of the New York Times. As the newspaper of record, the Times should take pains to keep from printing the results of unverified experiments. This article will undoubtedly spawn others like it, each one pointing to the Times as a reliable source of information.

It should be reliable, but it is not.

No comments: