Intellect no shelter from cognitive bias

Choose you evidence carefully by rocket ship

In a recent post I shared my observation that, as the usefulness of many psychotropics have been cast into doubt by a growing body of evidence, many people who used to chide questioners to “follow the evidence” now resort to anecdotes to defend their advocacy for these drugs. (And, denigration of other methods.)

One story about about ability to rationalize from an interview in Wired:

Wired: You write that people find it easier to rationalize stealing when they’re taking things rather than actual cash. You did an experiment where you left Coca-Colas in a dorm refrigerator along with a pile of dollar bills. People took the Cokes but left the cash. What’s going on there?

 This, I think, is one of the most worrisome experiments we’ve ever conducted, and it’s again about rationalization. There’s a story about a kid who gets in trouble at school for stealing a pencil from another kid, and the father comes home and says, ‘Johnny, that’s terrible, you never steal, and besides, if you need a pencil, let me know and I’ll bring you a box from the office.’

Why is that slightly amusing? Because we recognize that if we were taking the pencil from the office we would not have to confront that we are being immoral, in the way that we would if we took $10 from the petty cash box (even if we used that cash to buy pencils).

Another about how intellect is no protection from cognitive bias in The New Yorker [emphasis mine]:

But West and colleagues weren’t simply interested in reconfirming the known biases of the human mind. Rather, they wanted to understand how these biases correlated with human intelligence. As a result, they interspersed their tests of bias with various cognitive measurements, including the S.A.T. and the Need for Cognition Scale, which measures “the tendency for an individual to engage in and enjoy thinking.”

The results were quite disturbing. For one thing, self-awareness was not particularly useful: as the scientists note, “people who were aware of their own biases were not better able to overcome them.” This finding wouldn’t surprise Kahneman, who admits in “Thinking, Fast and Slow” that his decades of groundbreaking research have failed to significantly improve his own mental performance. “My intuitive thinking is just as prone to overconfidence, extreme predictions, and the planning fallacy”—a tendency to underestimate how long it will take to complete a task—“as it was before I made a study of these issues,” he writes.

Perhaps our most dangerous bias is that we naturally assume that everyone else is more susceptible to thinking errors, a tendency known as the “bias blind spot.” This “meta-bias” is rooted in our ability to spot systematic mistakes in the decisions of others—we excel at noticing the flaws of friends—and inability to spot those same mistakes in ourselves. Although the bias blind spot itself isn’t a new concept, West’s latest paper demonstrates that it applies to every single bias under consideration, from anchoring to so-called “framing effects.” In each instance, we readily forgive our own minds but look harshly upon the minds of other people.

And here’s the upsetting punch line: intelligence seems to make things worse. The scientists gave the students four measures of “cognitive sophistication.” As they report in the paper, all four of the measures showed positive correlations, “indicating that more cognitively sophisticated participants showed larger bias blind spots.”

Keep this in mind when you read headlines like, Addiction Treatment in America: Not Based in Science, Not Truly ‘Medical’. Ask yourself why she’d give the impression that Twelve Step Facilitation is not an evidence-based practice? Or, why she’d fail to mention that the best informed and culturally empowered addicts do not seek the treatments she advocates? They receive the kind of treatment she sneers at (Albeit reliably high quality versions of that treatment.) and they enjoy stellar outcomes.

This isn’t to say that there’s no truth in what she or other critics have to say. (That would be an example of a cognitive error called the argument from fallacy.) Just be suspicious of people who don’t acknowledge their bias, particularly when they say their only agenda is evidence or science.

Med Shortages Not Limited to Adderall

Over the last year, there have been several articles and posts howling that drug warriors (a group I have no interest in defending) are harming patients by creating shortages of Adderall. It struck me as possible, but also as an awfully simple and convenient narrative for writers with a particular point of view on drug policy matters.

The Detroit Free Press recently published an article that offers a little more context. Medication shortages aren’t limited to Adderall or drugs with misuse potential.

Drugs in short supply more than tripled from 61 in 2005 to more than 200 this year, according to industry groups and the FDA, which lists drugs with shortages on its website —

The most critical shortages involve cancer, antibiotic, nutrition and electrolyte imbalance medicines, according to a fall report from the FDA.

Those used for neuromuscular conditions, anesthesia in surgery, pain and antiviral conditions also have had bad shortages, the report said. Some hospitals have had to postpone surgery because they didn’t have the right anesthesia drug, pharmacy groups say.

A U-M study found three different drugs are hardest to get: succinylcholine injections, a muscle relaxant used in surgery; dextrose 50% syringe medicine, to restore blood glucose levels, and epinephrine injections, used in cardiac surgery.

Further, there are multiple causes of these shortages:

The most common causes are manufacturing violations, production delays, shipping problems or ingredient shortages. The problem — called a national health crisis by federal regulators and leading industry groups — has worsened because of complex legal, regulatory, economic and other factors, according to a 2011 report from the federal Food and Drug Administration.

Shortages grow as generic manufacturers have consolidated and fewer plants are left making certain drugs, the FDA and others say.

The DEA is frequently blamed for the Adderall shortage, but this article suggests that there may be another side to the problem:

The DEA decides how much of the drugs can be manufactured each year based on the previous year’s prescriptions. It allots a percentage of that total production to each manufacturer and those firms are free to decide whether to manufacture generic or brand name forms of the drugs. The FDA blames overly strict manufacturing quotas for the shortages, but the DEA suggests pharmaceutical firms are choosing to make more of the expensive brand name pills than their generic forms.

Pharma and the DEA both have credibility problems, so I don’t pretend to know the truth, but it’s noteworthy that so many drug policy bloggers and reporters failed to provide this context. It would appear that their frustration with drug warrior hysteria and bias has led many of them to push their own biases in their writing.


Everybody worships

I just read David Foster Wallace’s 2005 commencement address at Kenyon College. It’s a really, really great read.

A lot of it spoke to me, but, in the context of my recent posts on research and the values of harm reduction, this lept out to me:

…in the day-to day trenches of adult life, there is actually no such thing as atheism. There is no such thing as not worshipping. Everybody worships. The only choice we get is what to worship.

Seems a little strange to post something about this speech without commenting on his suicide, but I don’t know what to say.

I don’t believe his insight is diminished in any way by his suicide. It shows that insight often doesn’t lead to relief from emotional torment. It’s just very sad.

Addicted to discovery

A new article in Addictive Behaviors suggests that some research may be focused on meeting the needs of the researcher and media rather than clinicians or patients:

I’d like to suggest that perhaps the scientific community may be addicted to discovery, with too little consideration of the consequences. Indeed, Stange and Phillips (2007) contend that the “[h]unger to produce and sell dramatic research results is created and fed by the laymedia and medical journals’ need for news, researcher career pressures and egos, funders’ eagerness to prove their value, and a public made hungry for the next big thing” (p. 98).

A lot of the research looks less like the researchers are trying to produce something dramatic than just get something else published.

The writer offers the US Department of Agriculture Cooperative Extension Service as a potential model to improve the research, dissemination and implementation process:

The structure of the extension service is comprised of three discrete but highly interrelated components: academically-based agriculture researchers; county extension agents who work as change agents at the local level; and state extension specialists who link the researchers to the local county agents. Interestingly, the extension service is funded by federal, state, and county governments, a mixed funding model which coincidentally mirrors the tripartite streams of funding in the addiction prevention and treatment system.

Rogers (2003) identifies critical elements in the evolution of the extension service that have contributed to its success. Several of these elements speak directly to the inestimable value of collaborative relationships between producers and consumers of locally relevant research. First, the farmers (clients) participate fully in identifying their own local needs and communicate them to the county (and state level) agents. Next, the state agents work (and live) in close professional (and social) contact with the academically-based researchers, affording the opportunity to develop highly relevant research questions. Next, the research that the academics design and conduct is oriented to utilization right from the start. Finally, this utilization orientation is institutionalized — and valued — across the entire extension model structure.