Archive

Tag Archives: Ira Glass

Megan McArdle’s The Up Side of Down is good survey of literature about the science of failing, resilience, and success. Books of this sort, written for popular consumption, generally suffers from the three ring binder effect; it’s a loose collection of research and interviews, organized by themes. In some cases, the research has been presented in other contexts, both by the researchers themselves (Daniel Gilbert and Jonathan Haidt) and by other popularizers of behaviorial science.

Luckily, Ms. McArdle’s approach is disarming and charmingly self-deprecating. Her binder, as it were, ties together her own failures to the research she presents. Her failure to find a job, her inability to move past a relationship, and her experience combating 9/11 Truthers provide a human face to the statistics of neuropsychology research. Most importantly, she demonstrates the power inherent in recognizing when a path is failing and taking action to shut it down. Loss aversion supplies a  motive in maintaining status quo, and variations on this theme are explored.

As with such popular science books, there is a hint of the prescriptive in her book. Ms. McArdle supports a more generous approach to mistakes and wishes that political forces would stop moving towards harsher punishment for any mistakes.

Despite the compelling theme, and one that I tend to agree with it, I find these books shallow. To Ms. McArdle’s credit, I would absolutely love for her to expand on just about every chapter. As it is, she combines general lessons learned from both investigators and from her life. It is effective. Take it for what you will; if you want more, follow up on her bibliography. The book is compelling.

I found useful lessons, especially with emphasizing the need to give kids a safe place to fail. Ever since I became aware of research regarding the contradictory effect of praising intelligence rather than effort (actually, pointing out anything aside from effort), I’ve focused on the process. (There’s actually new research suggesting that merely visualizing directions – up versus down, flying versus digging – might affect cognitive tasks due to the emotionality of the visualization.)  It’s actually nicer and easier in some ways, because it gives adults cues to talk about specific things about the child’s project (“Oooh! I like how you did the trees and arranged them according to perspective!”).

Ms. McArdle’s book reminds us that it is not only OK, but necessary to identify faults. Especially when younger and with lower stakes; the kids can immediately see where they went wrong and they can correct it. The key is to be gentle enough to call attention to the mistake but not dwell on it. Make it feel like a bump; comment and move on.

Although I wish Ms. McArdle spent more time on developing the idea and presenting more research, I agree with her that the ability to remain calm and not focus on the emotional sting of shame and feelings of failing is absolutely crucial to moving on. Perhaps becoming accustomed to the iterative process of failing/identify/improve will help desensitize kids to the emotional turmoil of being wrong so they eventually focus on the substance of criticisms.

I happen to think there’s a lot to learn from Ms. McArdle’s book, and I can draw many parallels to the process of science. My colleagues and I have joked that we are in an asymmetric relationship: the science has all the power. We work, but our feedback is generally negative. Our advisors and supervisors simply give comments for improvement (ask anyone about the process of writing a grant or manuscript), only to receive more feedback upon submission – the paper is rejected/won’t fit our journal. If accepted provisionally, we will get more feedback from reviewers. Grants also get scored and we receive comments.

But we all understand this is the process. The worse comment for a grant is no comment at all. The grant being so bad that it was not worth the reviewer’s time to improve on it.

And of course, a lot of our time is spent on dealing with no or opposite results: no change where change is expected. Change were stasis is expected. The effect is too small or opposite what you predicted. And things break and stop working all the time. A lot of these errors come down the the experiments and analysis (perhaps an incorrect baselining or normalization.)

But when experiments start pulling together and a paper is eventually accepted, it is exactly like the first sunlight after an arctic winter. The rest of the time, it’s that arctic darkness.

Sorry; do I sound bitter?

I’m sure authors/writers/reporters all have analogous stories. The point is that success is more about attrition and self-selection. The people who thrive and have careers all continue to produce and deal with failures as if they are minor. They integrate criticism, iterate, and improve. So yes, I pretty much buy into Ms. McArdle’s thesis.

One thing I like about the book is that she tackles the issue of normative errors and accidents. The distinction is important to make, even if the definitions are not necessarily clear cut. Accidents are events that occur and couldn’t really be accounted for in the planning and execution. The operative word is could. Many things can and do happen, but the definition of those accidents happening is that it is coincidental, with the unfortunate victim falling prey to a low probability event.

Normative errors arise during process and execution, due to missed steps. The word here is should. Generally, there are a few things that should have been done, but weren’t. The two seem separated by degree; I suppose if you find yourself linking a series of events – if only I had walked a few steps quicker or slower, I would have turned the corner and seen the the guys backing out with the large pane of glass instead of walking into the glass – this probably is an accident.

A mistake can probably be traced to something one did or didn’t do, and a compounded mistake just means many people failed down the line. I can see how some readers might want clearer explanations.

But the point of the book is not explicitly about mistakes, but how we recover from them.

Ms. McArdle put together a rather compelling book. She connects threads in research on attention, motivation, and economics and drew new observations. I especially liked her chapter on tunnel vision (“inattentional blindness”). She starts with the description of Daniel Simons’s and Christopher Chabris’s experiment with having students score the number of times a basketball team, in a video. Afterwards, they ask the students about the number of passes – and whether they saw a gorilla mascot run threw the middle of the court, between the players. She seques into an analysis on the Dan Rather/President G.W. Bush National Guard story that cost Mr. Rather his job. Dan Rather made the mistake of defending his decision, rather than simply working to figure out whether something went wrong.

There were apparently a whole chain of mistakes, but the point is that there is power to simply acknowledge he could have been at fault. The proper play would be along the lines of Ira Glass’s signing off on Mike Daisey’s Apple story, where Mr. Glass admitted he was wrong and then spent a subsequent hour on analyzing the mistakes he and his team made – while rectifying the original story. A hot-of-the-press example is in how Bill Simmons dealt with the Dr. V’s Magical Putter story.

I do hope people read Ms. McArdle’s book. I think she has a talent for providing proper context and tackling the best and most relevant arguments between opposing views (see her chapters on bankruptcy, welfare reform, and moral hazard.) For the short length of time reading the book, I think readers will gain an immeasurable sense of well-being as they learn to love mistakes.

Advertisements

I recently heard a fun episode of This American Life, called “Kid Politics”. Ira Glass presented three stories about children being forced to make grown-up choices. The second story is an in-studio interview of Dr. Roberta Johnson, geophysicist and Executive Director of the National Earth Science Teachers Association, and Erin Gustafson, a high-school student. The two represented a meeting of minds, between a scientist who is presenting the best evidence demonstrating human induced climate change and a student who, in her words, does not believe in climate change.

It is worth listening to; Ms. Gustafson is certainly articulate, and she is entitled to think what she wants. I simply emphasize that, Ms. Gustafson uses language that suggests she is engaged in a defense of beliefs rather than an exploration of scientific ideas.

Ira Glass, near the end of the interview, asks Dr. Johnson to present the best pieces of evidence arguing in favor of anthropogenic climate change. Dr. Johnson speaks of the analysis of ice cores, where carbon dioxide levels can be detected. This can be correlated to evidence of temperature. Ms. Gustafson points out that apparently, in the 1200s, there was human record of a warm spell – I gathered it was somewhere in Europe, although the precise location and the extent of this unseasonably hot weather was not mentioned –  where low CO2 levels at the time.

Clearly, Ms. Gustafson has shown enough interest in the topic to find some facts or observations to counter a scientific conclusion. She then calls for scientists to show her all evidence, after which she herself will get to decide. I suppose at this point, I’m going to trespass into Kruger-Dunning territory and speak about expertise, evidence, and the use of logic.

In general, I do not think it is a good approach for scientists to simply argue from authority. I admit, this comes from a bias in my interests in writing about science to a lay audience. I focus on the methods and experiment design, rather than the conclusions; my hope is that by doing that, the authority inherent in the concept of “expertise” will be self-evident. That is, I show you (not just tell you) what others (or I) have done in thinking and investigating a problem. By extension, I hope I informed myself sufficiently before I prepare some thoughts on the matter, shooting specifically for fresh metaphors and presentation. (As an aside, I suppose that this might be a mug’s game, given the findings of Kruger and Dunning.)

If a scientist has done his or her job, one is left with a set of “facts”. These facts populate any school textbook. But the facts are more than that: they can act as, with a bit of thought and elaboration, as models. I dislike the distinction people make when they argue that we need to teach kids how to think and not a set of facts. I argue that learning “how to think” depends crucially on how well a student had been taught to deal with facts. These skills include how to deal with facts by using them as assumptions in deductive reasoning, weighing whether a fact has solid evidence behind it, and using facts as if they were models.

Here’s my issue with how Ms. Gustafson, and other anti-science proponents (like anti-evolutionists), argue. Let’s say we were told that gas expands upon heating. One might take this as a given and immediately think of consequences. If these consequences are testable, then you’ve just made up an experiment. Inflate a balloon and tie it off. If temperature increases lead to volume increases, one might immerse the balloon in hot water to see if it grows larger. One might choose to examine the basis of thermal expansion of gas, and he’ll find that the experiments have been well documented since the 1700’s (Charles’s Law). A reasonable extrapolation of this fact is that, if heating gas increases its volume, then perhaps cooling gas will lead to a contraction? One might have seen a filled balloon placed in liquid nitrogen (at – 196 deg C) solidify, but it also shrivels up.

Depending on how well facts are presented, they can be organized within a coherent framework, as textbooks, scientific reviews, and  introductions in peer-reviewed articles already do. My graduate advisor characterized this context fitting as  “provenance.” No idea is truly novel; even if one does arrive at an idea through inspiration and no obvious antecedents, it is expected that this idea have a context. It isn’t that the idea has to follow from previous ideas. The point is to draw threads together and if necessary,  make new links to old ideas. The end point is a coherent framework for thinking about the new idea.

Of course, logic and internal consistency is no guarantee of truth; that is why a scientist does the experiment. What hasn’t been really emphasized about science is that it is as much about communication as it is about designing repeatable experiments. Although scientists tend to say, “Show me,” it turns out that they also like a story. It helps make the pill easier to swallow. The most successful scientists write  convincingly; the art is choosing the right arguments and precedents to pave the way for the acceptance of empirical results. This is especially important if the results are controversial.

The error Ms. Gustafson makes is that she thinks by refuting one fact, she can refute an entire tapestry of scientific evidence and best models (i.e. “theory”). She points to one instance where carbon dioxide levels do not track with the expected temperature change. But in what context? Is it just the one time out of 1000 such points? I would hazard a guess that the frequency of divergence is probably higher than that, but unless the number of divergences is too high, one might reasonably suppose that the two correlate more often than not. (Causation is a different matter;  correlation is not causation.)

But let us move on from that; a more elemental disagreement I have with Ms. Gustafson’s point is that, let’s say that one agrees that carbon dioxide is a greenhouse gas. A simple model is that this gas (and other greenhouse gases such as water vapor, methane, nitrous oxide) absorbs heat in the form of infrared radiation. Some of this energy is transferred into non-radiative processes. Eventually, light is re-emitted (also as infrared radiation) to bring the greenhouse molecule to a less energetic state. Whereas the infrared light had a distinct unidirectional vector, radiation by the greenhouse molecule will occur in all direction. Thus some fraction of light is reflected back towards the source while some other light essentially continues on its original path. If infrared light approaches earth from space, then these gases act as a barrier, reflecting some light back into space. Absorption properties of molecules can be identified in a lab. We can extend these findings to ask, what would happen to infrared heat that is emitted from the surface of the planet?

A reasonable deduction might be that just as out near the edge of the atmosphere, greenhouse gases near the Earth surface also absorb and reflect  a  fraction of heat. Only this time, the heat remains near the Earth’s surface. One logical question is, how does this heat affect the bulk flow of air through the atmosphere? (An answer is that the heat may be absorbed by water, contributing to melting of icebergs. Another related answer is that the heat may drive evaporation and increasing kinetic energic of water vapor, providing energy to atmospheric air flows and ultimately to weather patterns.

For someone who ignores greenhouse gas induced global warming, dismissing the contribution of carbon dioxide isn’t just a simple erasure of a variable in some model. What the global warming denier is really asking that the known physical property of carbon dioxide be explained away or modified. Again, the point is that carbon dioxide has measurable properties. For it not to contribute in some way to “heat retention” is to say that we must ask why the same molecule won’t absorb infrared radiation and re-emit infrared radiation in the atmosphere, in the same way that was observed in the lab. In other words, simply eliminating the variable would require us to explain why there are two different sets of physical laws that apply to carbon dioxide. In turn, this would require a lot of work to provide context, or, the provenance to the idea.

Yes, one might argue that scientists took a reductionist approach that somehow removed some other effector molecule if they measured carbon dioxide properties using pure samples. Interestingly enough, the composition of the atmosphere is well known. Not only that, one can easily obtain the actual “real-world” sample and measure its ability to absorb unidirectional infrared and radiate in all directions. This isn’t to say that thermodynamics of gases and their effects on the climate of Earth is simple. But it is going to take more than a simplistic type of question, such as to posit that there is some synergistic effect between carbon dioxide and some other greenhouse gas or some as-yet unidentified compound, so that we actually modify the working model physicists and chemists have about absorption and transfer of energy.

If you think that it seems rather pat for a scientist to sit and basically discriminate among all these various counter-arguments, I am sorry to disabuse you of the notion that scientists weigh all facts equally. Ideally, the background of the debaters ought not to matter. Hence, you will get scientists to weigh your criticisms more heavily if you show the context of the idea. The more relevant and cohesive your argument, the more seriously you will be taken. Otherwise, your presentation may do you the disservice of giving the appearance that you are simply guessing. That’s one problem with anti-science claimants: all too often it sounds like they are trying to throw as many criticism as possible, hoping that they will get lucky and have one stick.

Take evolution: if one suggests that mankind is not descended from primates, then one is saying that mankind was in fact created de novo. That is fine, in and of itself, but let’s fill out the context. Let’s not focus on the religious texts, but instead consider all the observations we have to explain away.

If we were to go on and to try and explain mankind as a special creation, how would we go about explaining mankind’s exceptionalism? Can we even show that we are exceptional? Our physiology is similar to mammals. We even share physical features as primates. Sure, we have large brains, among the largest brain mass to body mass ratios in the animal kingdom. Yet we differ in about 4% of our genome compared to chimpanzees. Further, at a molecular level, we are hard pressed to find great differences. We simply work the same way as a lot of other creatures. We have the same proteins, despite the obvious differences between man and mouse, a weak similarity between our proteins mean that we have only 70% sequence homology. It seems to me that at multiple levels, at a physiological level, at the level of physical appearances, and at a genomic level, we are of the same mettle as other life on earth. Yes, the fact is that we do differ from these other lifeforms, but it seems to be more logical to suggest that mankind is one type of life in a continuum of the possible lifeforms that can exist on Earth. It just seems likely that by whatever process that led to such a variety of creatures, man must also have been “created” from such a process.

 I hate to harp on this, but a fellow grad student and I had such arguments, while we were both doing our thesis work. My friend is a smart guy, but he still makes the same mistake that anti-evolutionists make: by disproving natural selection, one  therefore has provided some support for creationism. We argued about Darwin’s theory and whether it can be properly extended from a microscopic domain. He was willing to concede evolution occurs at a microbiotic level – such as for “simple” organisms, evolution makes sense, since fewer genes mean less complexity and therefore changes can be just as likely to be beneficial and deleterious.

I thought the opposite. If an organism is “simpler” – namely because it contains a smaller genome – it is even more crucial for a given organism that a mutation be beneficial. A larger genome, from empirical data, generally contains more variants of a given protein. While this in itself reflects the appropriation of existing genes and their products for new functions. Perhaps one possibility is that   an increase in isoforms of a protein also suggests how mutations can occur without the organism suffering ill effects directly. There is a redundancy of protein and function. Also, my friend seems to regard fitness as a “winner takes all” sort of game – as in the organism lives. I merely saw the “win” as an increase in probability that the animal will have a chance to mate, not organismal longevity. Sure, this is a just so story; I think his argument is better than the usual creationist claptrap, but only in the  trivial sense that, yes we need to take care not to over interpret our data or models and yes,  scientific theories – althoughswa they are our best models –  are temporary in the sense that we can revise them when better evidence comes along.

To go back to the way Ms. Gustafson and my friend argue, it behooves them to explain the exceptional circumstances by which we, or carbon dioxide, can act differently from our best model (i.e. theory) and yet conform to it most of the time.

Thus, despite Ms. Gustafson’s call for “all the evidence”, I somehow was left thinking no amount of evidence will persuade her. Part of the problem is that, like the religious who misapply ideas of meaning found in their bibles to the physical evidence generated by scientists, she misapplies her political views to provide the context through which she views scientific evidence about global warming. Whereas she should have used logic to deduce that global climate does not predict local weather and scientific principles  to determine whether global warming is part of a normal cycle for the Earth or is in fact due to circumstances like an increase in greenhouse gases, she probably thought of global warming in terms of regulations and taxes pushed, generally in the United States, by Democrats. Thus, Ms. Gustafson speaks, in Stephen Jay Gould’s term, from the magisteria of meaning (as defined by her political and religious beliefs) and not from the magisteria of science. In this case, she isn’t defending her theory about how the world works; her motivation is to fit the observations to her political and religious ideals.

Can we really separate the political from the scientific? If some scientist argues that there is a problem, it seems difficult to find ways to argue against them. My only suggestion is that Ms. Gustafson and others like her consider their arguments more carefully. Nitpicking specific examples is counter-productive. All theories can be criticized in this way. However, integrating the counter-example is not a straight-forward process, especially if simplistic criticism is at odds with some other firmer, more fundamental observation that even Ms. Gustafson has no problems accepting.

 

%d bloggers like this: