Beware the confirmation bias
Recently, while alone and unsupervised, I allowed myself to be goaded into an ill-advised Twitter debate. With someone who calls himself Ducks Dodger, no less. (For those of you who didn’t see it, you can go to my Twitter account, scroll back to April 8-9 and check out this tar baby I got caught up in.)
I soon realized the entire affair was an exercise in futility, so extracted myself from it. It was an exercise in futility because it was a blatant case of the confirmation bias writ large. But my aggravation hasn’t gone to waste because I’ve been waiting for an excuse to write a little essay on the confirmation bias that plagues us all. Who knew fate, in the guise of Ducks Dodger, would give me the prod I needed.
Taktu cleaning fat from seal skin with an ulu
Cape Dorset (Kinngait), Nunavut, Canada, August, 1960
In a bit, I’ll describe the details of this Twitter fiasco and how the confirmation bias reared its head. But first, let’s look at provoker-in-chief of the confirmation bias: cognitive dissonance.
Cognitive dissonance is the uncomfortable feeling we all get when trying to hold two opposing or clashing ideas in our brains at the same time.
For example, let’s say there exists a politician you hold in great esteem. You’ve actually met him, and he’s a warm, friendly charismatic guy, who embodies everything you and your party hold dear. Then a credible news report surfaces, fingering the guy as a pedophile.
If this accusation were made against a much-despised politician from the other party, you would have no trouble swallowing it whole. But since it is against your guy, you refuse to believe it. Even as more evidence stacks up, you just can’t bring yourself believe this wonderful person could do such a thing. The uncomfortable feeling you are experiencing is cognitive dissonance. You cannot hold the idea of this warm wonderful guy who exemplifies your beliefs dandling little boys.
The example above is a bit over the top, but similar, though less extreme, events happen constantly in politics and elsewhere. It is illuminating to see how they are dealt with. How those in power use the resolution of cognitive dissonance to get out of a jam.
Politicians seem to be incapable of staying out of trouble. Of course, they live under the magnifying glasses focused on them by the press and the opposing party, so there is little room for error. When one does screw up, a little drama unfolds that always follows the same script.
A news report pops up pointing out a political snafu. The opposing party and its news organs make hay. The offending party is rattled, confused, and lies low until it gets a few plausible talking points put together either denying or minimizing guilt. In rare cases – John Edwards’ impregnating his mistress while his wife, dying of breast cancer, was campaigning at his side springs to mind – the party will sacrifice one of its own. But most of the time, the creation of talking points is the common course of action.
The press and other politicians, of course, all see through this smoke screen. But the smoke screen wasn’t created for them – it was created for us, the great unwashed masses of voters.
When we hear of one of our own screwing up, we instantly are overwhelmed with cognitive dissonance. It’s an uncomfortable feeling. We like this guy, he is in our party, yet it appears he did bad. How can we hold those opposing thoughts in our brains? We cannot, at least not without considerable discomfort.
But, we don’t have to endure for long because our leaders’ talking points come to our rescue. Our cognitive dissonance dissolves in the solvent of these soothing words. We don’t have to think. Our man is blameless after all. Life is once again good.
The confirmation bias is what allows us to uncritically accept these soothing words as the truth. We’re being told what we want to hear. And, thus, we believe it blindly, willingly, because it confirms whatever we already believed.
To show how the confirmation bias is built brick by brick, let us turn to politics once again.
Assuming we come to politics a tabula rasa (a major assumption because most of us follow in our parents footsteps or openly rebel against our parents), we start at zero.
Imagine yourself at the very top of the pyramid of knowledge and belief. Right at the apex, there is no knowledge or belief. You are a political newborn, so to speak.
At the base of the pyramid, the knowledge level is deep and wide. On the right side of the base is where all the conservative ideals, knowledge, and insights lie. On the left side live the liberal ones.
When you start at the top, you get tipped down one side or another. Maybe it’s a column you read, or a talking head on TV, or a parent, teacher, or friend. Doesn’t really matter, but somehow you get tipped to one side or the other and start rolling down that side of the pyramid, gathering ideology as you go.
You establish your rudimentary political views, and, as you start rolling down, you read more, you engage in discussion, you watch cable channels that mirror your views, and you, in general, become more entrenched in your ideology. All the way down you continue to confirm your ever-growing bias. Once you have reached the bottom, you have marinated so long in your particular political sauce that you can’t possibly understand how anyone could not believe the same way you do. In fact, you are certain that anyone who doesn’t is a completely misguided idiot.
It never occurs to you that others may have tipped and rolled down the other side of the pyramid. They, too, have reached their side of the bottom and are completely infused with the righteousness of their own beliefs and cannot imagine how someone could be so stupid as to believe any differently.
What is even worse is that many of us who have rolled down our own side of the pyramid refuse to even read anything written by one who has rolled down the other side. You don’t want to learn anything that might throw you into cognitive dissonance, so you renounce it as trash, unworthy of your reading, and move on. We’ve all done this at some time or other.
We work hard never to let opposing views penetrate our consciousness in an effort to avoid the unpleasant sensation of cognitive dissonance. The confirmation bias is the tool we use.
Conservatives don’t watch MSNBC; liberals don’t watch Fox. We always subscribe to magazines and newsletters that reflect our own underlying views. We read blogs that confirm our biases. If we stumble on one that doesn’t, we often leave snide comments and never return. The wide variety of material available on the internet and cable TV today is, in my view, why politics is so vicious. Not all that long ago, most of us had access to a daily newspaper or two and three TV channels, all of which were moderately liberal. People could argue with one another and drift toward the extremes, but were almost invariably brought back toward the middle by the staid mainstream media. Now people can indulge their confirmation biases 24/7, which doubtless leads to more edging toward the fringes and less trust of those on the other side of the debate.
Whenever someone comments for the first time on this blog and has a website (which can be discerned by their name being hyperlinked), I will often click on to take a look. If the commenter turns out to be a liberal, his/her website will be crawling with countless links to other liberal blogs, newsletters, articles, etc. Same if a commenter is conservative. I never find a blog – of a political nature (save one, see below) – that has links to both conservative and liberal sources.
You can test your own confirmation bias by going to one of my favorite sites, RealClearPolitics.com (RCP), the blog mentioned above that does link to both liberal and conservative sites. Go to the site and read down the list of hyperlinked articles. Some days it’s a little more conservative than liberal, and other days it is just the opposite. But on any given day, the site serves up a pretty even mix.
As you let your eye run down the list, notice which ones you want to open and read versus the ones you want to avoid. If you are a liberal, you will doubtless be drawn to those articles that confirm your liberal bias, and you will avoid opening and reading the obviously conservative ones. Same thing if you are a conservative.
And, God forbid, if you are a liberal (or conservative) and you happen to click on a link you think will confirm your bias only to discover it does not; chances are you will call BS and back out of that turkey in a heartbeat.
As an exercise in exorcizing my own political confirmation bias, I several years ago made the commitment that if I went to RCP, I would read every article. Not pick and choose. It has been quite enlightening.
We should all do this, not just with politics but with everything. Many aspects of life plunge us into cognitive dissonance, and our good friend and enabler, the confirmation bias, always stands ready to help us out of it. It takes real work to hold the confirmation bias at bay and think critically. Critical thinking efforts are difficult and do not always provide us with the answer we want or sometimes even the correct answer, but they are, as Freud said of women, “the best thing of their kind that we have.”
The confirmation bias is everywhere
Sadly, our tendency to succumb to the confirmation bias is not limited to politics. Even scientists are all too prone to fall victim.
I would say at least half, if not three fourths, of papers published in academic journals are exercises in employment of the confirmation bias. It’s even worse with non-academicians, bloggers and the like, who truly (if sometimes mistakenly) believe they are practicing good science and indulging in heavy duty critical thinking.
I want to use my recent Twitter dustup with Ducks Dodger as an example of what I mean.
Here is the set up.
A blog post appeared, crediting Ducks Dodger for most of the content, stating in no uncertain terms that the Inuit were not in ketosis, and so therefore the Inuit didn’t really eat a high-fat, low-carb, ketogenic diet.
The basis for this unequivocal pronouncement were three old papers published between 40 and 85 years ago. The authors of these papers had not been able to find measurable ketones in their Inuit subjects under normal conditions. After a day or two of fasting, however, ketones were present just as they are in most of the rest of us. This finding set the authors to speculating as to why a group of people who ate primarily fatty meats would not be in ketosis all the time, because it just didn’t make sense that they wouldn’t be throwing off ketones like crazy.
One proffered solution to this conundrum was that the glycogen in the meat providing the bulk of the Inuit diet was providing enough carbohydrate to prevent their going into ketosis. It was known then that glycogen (the storage form of glucose) was present in muscle, and it made sense that there might be enough glycogen therein to prevent ketone formation.
There is about 6 grams of glycogen per pound of meat, so if the Inuit consumed from five to eight pounds of meat per day, as the articles claimed they did, the 30 to 48 grams of glycogen could conceivably halt ketone production. Especially if there were a bit of plant or vegetable matter thrown into the mix. That was the speculation at any rate.
The fly in the ointment here, however, is that glycogen rapidly degrades into lactic acid upon death. So unless this meat came from living animals that were still breathing while being eaten – highly unlikely – there wasn’t any glycogen to be had.
Scientists working today use liquid nitrogen ( ?320F) to flash freeze the livers and muscles of sacrificed lab animals immediately after death in order to determine the glycogen content before it degrades. Scientists working 80 years ago didn’t have this option, nor, apparently, did they understand how quickly glycogen degrades.
Glycogen degrades to lactic acid post mortem
What happens is this: when an animal meets a sudden end, it doesn’t die all at once. It’s kind of like a modern car. You pull it into the garage and turn off the key, but the lights stay on, the fan keeps running, the windows can still go up and down for a bit. The car is dead in that it has no way to actually run, but parts of it are still alive.
If, God forbid, you were to be shot in the heart with a high-powered rifle, you would doubtless be dead. But, like the car in the garage, parts of you would live on for a while. The muscles in your legs, for example, wouldn’t know you were really dead, and they would keep on keeping on.
In order for muscles, even relaxing muscles, to survive they need to burn fuel. Burning fuel requires oxygen. If your heart isn’t pumping because a bullet went through it, the muscles in your legs get no oxygen. When no oxygen is available, muscles do what muscles do when they exercise intensely and can’t keep up with the oxygen demand: they switch to anaerobic metabolism and go after the stored glycogen.
Burning the glycogen anaerobically results in a buildup of lactic acid. When your muscles are exercising, the blood washes away the lactic acid, but even then, if you exercise intensely enough, the lactic acid build up outpaces the blood’s ability to carry it away, leading to lactic acidosis. Your exercising muscles begin to burn and if you don’t slack off and let the accumulated lactic acid get washed away, your muscles will begin to cramp.
Same thing happens when you are dead. Since there is no blood pumping oxygen to the muscles, they burn glycogen anaerobically, producing lactic acid, which accumulates. Ultimately, the glycogen is depleted, the tissues are acidic, and no more ATP is produced. No more ATP means no power to run the calcium pumps, the cells lose calcium and the muscles begin to contract, leading to the condition called rigor mortis.
The upshot of all this is that when animals die, the glycogen in their muscles quickly degrades to lactic acid. The only way this process can be halted is to immediately flash freeze the tissue with liquid nitrogen to halt the glycogen to lactic acid conversion. Since that happens mainly in the lab and in special flash freezing facilities, any meat you purchase at the store or from your local farmer or that you kill yourself will not contain any glycogen to speak of.
The glycogen to lactic acid conversion upon death is all really basic science, not in dispute by anyone.
Then along comes the Duck claiming the glycogen survives death and is really there, which, he believes, is why the Inuit, under normal circumstances, don’t have measurable ketones. They’re eating all this meat, which is full of glycogen. What he fails to understand is that the Inuit are keto-adapted. Their lifelong diet of high-fat meat has gotten their ketone-producing-and-consuming systems working in precisely controlled fashion. Like, dare I say it, a well-oiled machine.
The Inuit burn ketones as they make them, so it stands to reason that they might not have measurable ketones under normal circumstances. During the time these old studies were done, the non-Inuit, white-bread-eating, white man was the standard. Feed a non-keto-adapted bread eater the Inuit diet, and he goes into ketosis in no time. Researchers back then figured the Inuit should do the same thing, and when they didn’t – because of keto-adaptation, which was not understood at the time – these scientist thought it worthy of publishing.
I gently pointed this out via Twitter and generated a barrage of Tweets from the Duck, consisting primarily of a series of links to obscure documents showing there was indeed glycogen in meat. Especially meat that had been quickly frozen. There were anecdotes about fish, having been jerked from holes in the ice and exposed to air, freezing instantly. I responded that it didn’t matter how cold it was at the Arctic Circle, it wasn’t ?320F, which is about the temp of the liquid nitrogen required to freeze tissue quickly enough to preserve glycogen.
Stefánsson makes an appearance
I proposed a reading of some of the works of the arctic explorer Vilhjálmur Stefánsson, who had lived among the Inuit for years and knew their ways well. He had described their eating habits in many of his books and articles. This suggestion inspired another over-the-top post, with content credited to the Duck, describing Stefánsson as a charlatan, a humbug, and a poltroon, who couldn’t be trusted to get anything right. (Stefánsson was involved in a couple of controversial events – the loss of the Karluk and the Wrangle Island expedition, both of which resulted in the loss of life. If Stefansson had fault in these, it arose from his overestimating the skills of those others involved. He, himself, was an intrepid Arctic explorer, who was never daunted by cold, barrenness or distance. And, as far as I know, no one ever doubted his observational ability in terms of the ways of the Inuit.)
I responded with a link to Stefánsson’s obituary appearing on the front page of the New York Times on August 27, 1962, which, though it did make reference to the Wrangle Island fiasco, made no mention of any skullduggery on his part.
I was then accused of not using science to argue my point, but resorting to nonscientific articles in “news rags” to bolster my case. I pointed out there was no science involved in repeating accusations of Stefánsson’s alleged humbuggery, and all I was doing was noting the most prestigious newspaper of the time, a newspaper never afraid to state the bad along with the good in an obit, ran a front page obituary without any mention of these accusations, leading me to believe they were not widespread and in common knowledge. Everyone in the public eye has detractors – for example, Christopher Hitchens wrote an entire book damning Mother Teresa. Had Stef been the vile lying swine the Duck was making him out to be, I doubt the NY Times would have treated him so kindly.
After a little back and forth, I withdrew from the debate, because it was becoming a bit harsh. Twitter is kind of the internet Wild West today, and many less restrained Twitterers were jumping into the fray, and the attacks were becoming too ad hominem for me.
Soon, the Duck bailed out, too, but not without recasting his views in a long farewell comment.
Parts of a couple of paragraphs provide a perfect example of what I mean by the confirmation bias and is emblematic of the way most people argue a point today:
I hope you realize that Dr. Eades isn’t an expert in everything. I did a lot of research and simply showed Dr. Eades numerous studies that showed he was mistaken. I don’t see what the big deal is. [Italics mine]
I put a lot of research into this article, citing over a dozen scientific studies to support my case. When he outright dismissed the article, I continued to show him real world evidence on Twitter and scientific data that explained why he was mistaken. Dr. Eades never once showed any scientific evidence to refute anything I put forward. [Italics mine]
I am absolutely certain the Duck put a lot of time in on his research, but he went about it in the wrong way, and the confirmation bias jumped up and bit him in his tail feathers.
I know this to be true from one of his Twitter links. Take a look at this screen shot:
If you look at this link purporting to be real science bolstering the Duck’s case, you’ll see the Google search terms he used to track down this article were “freezing” “stops” “glycogen.” He wasn’t making a serious scientific inquiry – he had already gone public with his theories, now he was simply looking for anything to confirm them. The sad thing is, in Googling those three words, Google presented Duck many pages of material, most of which confirmed the basic science view of the degradation of glucose to lactic acid immediately post mortem, before he ever got to the page pictured above confirming his bias. Try it yourself. Google those three words and see what you find.
That’s not how science works. In real science, which is, sadly, not practiced all that often, researchers attempt to discredit their hypotheses, not confirm them. Only after repeated efforts to prove their own theories incorrect, do true scientists start to consider that they may be on to something.
Most researchers, however, are like the Duck. They come up with an hypothesis, then construct an experiment designed to confirm what they think they know. And they usually do.
But let me end this interminable post on a conciliatory note. Let us grant the Duck the glycogen he swears the Inuit get from eating their five to eight pounds of raw meat per day. There is about 6 grams of glycogen per pound of raw meat, so if you want to eat five pounds of raw meat per day to get your 30 grams of carbs, go for it. I still say it is a ketogenic diet, and I’ll bet good money that until you become keto-adapted, you can eat all the carbs you’ll find in your daily ration of five to eight pounds of raw meat and still turn your ketostix purple most of the time.
If only this were the kind of ‘high-carb’ diet most people followed, we would have a lot less disease and a lot fewer victims of obesity.