Analytical Bias
Analytical Bias
The world is made up of systems. Our body is a system, or in fact a system of systems. What we call “society” is another system of systems, as is the natural environment. And all these meta-systems are themselves elements in even more encompassing systems that interconnect them.
But these systems are very complex, difficult to explain or predict. One successful strategy, which has had a revolutionary impact on how we live, is analysis. This approach segments complex entities into smaller parts in order to study them individually. Medical researchers don’t study the body as such, but perhaps kidney function or particular blood cells. Social scientists may specialize in the effect of lobbyists on legislation, labor market patterns among immigrant communities, or changing child-rearing norms. Natural scientists study the viability of artificial wetlands, upwelling cycles in certain coastal zones or changes in the carbon balance of a set of tropical forest plots. By biting off chewable portions of a much larger world, science makes it possible to achieve progress in our understanding of how things work: testable hypotheses, demonstrable evidence, causal explanation. Analysis is the art of taking things apart, studying the pieces, and then putting them back together.
But this approach, for all its benefits, fails to capture most of the interactive effects that make a system a system. It leads us to overstate the separateness of the things we study and observe and to understate their connectedness. This is not an argument against thinking analytically, but for not being surprised by what this thinking fails to see so we can at least somewhat compensate for its shortcomings.
I’d like to introduce the term “analytical bias” to refer to this tendency to overlook the interconnectedness of things. Of course, many thinkers, going back centuries, have recognized this problem; it’s the guilty conscience of analysis itself. I’m just giving it a name.
@Peter,
Reductionist thinking is a powerful tool in biology. It is at the center of biochemistry, which has yielded profound insights. But all biochemists know that while you can run many enyzmatic reactions backwards if you supply excess product, that doesn’t mean it happens in nature.
One familiar pitfall of analytical bias is eugenics. Eugenicists focus on a trait or small group of traits, overlooking both the genetic complexity underlying those traits and the fact that genes contributing to that trait also contribute to others.
The good news, for science at least, is that the self-correcting peer review system, together with better experimental design, can eventually uncover analytical bias.
A Definition of Systems Thinking: A Systems Approach
“Systems Thinking’ has been around for a few decades as a middle-school curriculum, at least where I live. It influenced one of my kids (the brilliant one) to obtain an engineering degree in ‘systems engineering’. The link above is to an article in Procedia Computer Science:
… This paper proposes a definition of systems thinking for use in a wide variety of disciplines, with particular emphasis on the development and assessment of systems thinking educational efforts. The definition was derived from a review of the systems thinking literature combined with the application of systems thinking to itself. …
Just to be clear, I didn’t want to write a massive post describing all the different approaches that have been taken to ameliorate analytical bias. For myself, I first became acquainted with the problem when I read a portion of Hegel’s Logic over 50 years ago. There’s a lot to be said about the pluses and minuses of different antidotes….
Peter:
I hope you have a good holiday tomorrow.
One of the things Dan and I have tried to do is have a blend of people at AB who would approach topics from different directions or disciplines. I think we are somewhat successful in doing so.
I am not your equal by any stretch of the imagination. Just a simple supply-chain guy who had a great experience globally working for different cultures and bringing my knowledge to bear in improving the movement of material, improving costs of manufacturing, and solving issues. Since 1974 I am been at this. It has been a fun go.
With all the tooting of my horn, I find your topics extremely interesting. We are very happy to post your material and commentary at AB.
Thanks — that’s such a nice thing to say. Of course, we’re all more knowledgeable about some things than others. There’s only so much attention we have to give the zillions of things we might be interested in.
You open up a topic, and who knows what will become of it?
‘Systems thinking’ is all about interconnectedness, and
has become an important field of study in its own right.
Fred,
Yep, but the most difficult part about systems thinking is actually being any good at systems thinking. When academics learn their method of epistemological formalization then that framework is acquired with its own set of biases and limiting boundaries. There are always limits to knowledge such that how we know what we know may in the end matter less than how we do not know what we do not know.
The unknowable unknown unknowns were the high point of the unknowing Donald Rumsfeld’s career.
Well, i first saw the “systems” offered in the college catalogue about fifty years ago, and I thought “what else would you do?”
Turns out I may have been wrong. I don’t know what they actually teach, but I have learned from my own “analysis” that NO ONE, genius or not, can think of everything. And everyone, genius or not, can make himself stupid by forgetting that he does not know everything.
I think it’s the reason big corporations have “meetings” but from what I have heard those are usually a waste of time, with everyone who wants a promotion agreeing with the boss.
But I also know that the people who won WWII worked in groups and communicated between groups very well and very efficiently. They not only understood the concept, they had bosses who very much understood it.
Now, of course, that doesn’t mean that geniuses who invented game theory, or “fundamentals of mathematics” or much of the computer industry, can understand why it might not be a good idea to atom bomb the Russians while we had one and they didn’t.
Some day the Rosenbergs will be recognized as great patriots.
YEP!
Dorman…of Grist?
i’d be very interested to read why you think environmental groups may be wrong to chase “green” infrastuctue dollars. I think I agree with you but won’t play the cokie game with Grist.
I have enough worry having given Market WAtch my credit card info to pay a dollar to comment on Social Security. Only to have them tell me they will automatically bill my credit card 20 dollars a month after the first month.
For a brief version of my take on the GND, see the article I wrote for Challenge last year. A longer version will appear with my climate book next year, Alligators in the Arctic and How to Avoid Them. I had a more specific critique of the strategy of some green and social justice groups to bypass elected representatives and assure themselves of allocations from carbon taxes, as attempted a few years back in WA State. I wrote a number of blog posts on that.
thanks, peter.
dorman,
all i found was the abstract and what appeared to be a price of $45 for access. i am not in that income bracket. but i did notice in the abstract that you seem to minimize forest loss. even without climate change, forest loss will kill us, or at least make life not worth living.
i agree with you, i think, that the green new deal crowd don’t seem to really know what they are doing.
Sorry the article was paywalled. Try this.
Peter
Thank you
When I saw the post title I was thinking of the bias in choosing what to simplify based on the result you like rather than in choosing what to simplify based on what sub-system you think you understand. What you call analytical bias is understandable if the boundary conditions you can use allow your simple model to produce useful results. It becomes a problem when you try to extrapolate beyond those boundary conditions – which is almost always going to be the case when you try to combine subsystems.
Dorman
Thank you. Apologies ifthe following misses your point, or just repeats it:
We are not going to solve the globl warming poblem by making more electric cars that are faster with longer range. We could solve the global warming problem if people would learn to use less energy. That is doable (using less energy), but “people learning” does not seem to be,
electric cars here are just an example. We need to do with a much less energy intensive economy. We are trying to save the earth without it costing us any money or inconvenience or even any change in the way we value our time.
in other words, we will go the way of Ozymandius (yes, a fictional character, much like the human race).