Nerds
I’m going to present two claims, and you can decide for yourself which is more compelling:
- Differentiation of species occurred over millions of years through natural selection of hereditary traits.
- Differentiation of species occurred over millions of years through natural selection of hereditary traits, you prick.
The second one just sounds truer, doesn’t it? That is the odd finding of this study, helpfully summarized by one of the authors in last weekend’s New York Times. First of all, I think we’re all glad that there is a Journal of Computer-Mediated Communication, and it is printed. Second, the study focused specifically on online comments sections, finding that comments which contained epithets, profanity and ad hominem attacks affected readers’ viewpoints more powerfully than equivalent comments without those attacks. Civilization is doomed.
Maybe it’s just doomed to develop the internet, collapse into mindless belligerence, revert to agrarianism and then develop the internet again. That wouldn’t be so bad, provided you weren’t born near either end of the cycle. Sucks to be us, though. Perhaps the most alarming aspect of the study is that the documented effect is particularly pronounced in the area of science, where we are likely to evaluate new information without firm expertise.
Science is also maybe the area were we most need to be scrupulous in our thinking. It’s one thing for everyone to believe crude and fallacious arguments about which congressional candidate to vote for, and quite another for us to convince one another with ad hominem attacks about, say, global warming. If you vote for the wrong guy because I called supporters of his health care plan commie traitors, the political system suffers. That’s a problem. But if my use of the word “idiot” convinces you that man-made carbon emissions do not affect the temperature of Earth, you face a problem of a different order.
Like a lot of contemporary issues, climate change is a collective action problem. If we’re going to do something about it, we have to do it en masse. In that kind of situation, the question of how we can ensure people know what’s happening and come up with useful ideas by talking about it moves from a theoretical underpinning of democracy to an existential need. You cannot have a functioning republic of 350 million people unless a substantial number of them know what the fudge is going on.
You also cannot have a functioning republic of et cetera etc. in which everyone discusses things wisely and without recourse to fallacy, period. At least it’s historically unprecedented. The ordinary person regularly takes recourse in ad hominem attacks or whatever, because the ordinary person knows he is right first and fashions his argument second. That’s why “scientist” is a specific job: almost no one else thinks that way.
Granted, it is extremely debatable whether internet comments reflect the ordinary person. There’s good reason to think they don’t; Combat! blog, for example, has an unusually smart Comments section, but the average number of comments on a given post—three or four—constitutes a little less than 1% of average daily unique visitors. Still, even if internet comments only reflect the ideas of a small subset of cranks, they affect our perceptions. How differently would you read Gibbon’s Decline and Fall of the Roman Empire if it ended with 20 pages of anonymous quotations about how it is fake and gay?
So what do we do about that? As Brossard and Scheufele point out in the Times, the internet is supposed to be this rad democratizing force that gives voice to all of us, regardless of place and with only indirect regard to status. What happens when that voice says the president is not an American citizen and you’re retarded if you think global warming is real?
Here we are forced to confront an ugly aspect of democracy: it is the most ethical system, but often it does not produce the best results. People have acknowledged this problem from Aristotle to Alexander Hamilton, but the internet allows us to stare at it more directly than ever before. It could be that the marketplace of ideas functions better with certain borderline unjust controls. NASA does not open its discussions of jet propulsion research to whatever opinionated amateur has built a model rocket in his garage. Maybe, as Paul Krugman sometimes does with his columns, we should not open certain discussions to commentary from the general public.
That seems unsatisfying, though. As a person who leans toward the equality side of the liberty/equality tension in American democracy, I want everyone to be a responsible arguer. It’s why I shouldn’t read comments sections, and why I so easily get sucked in. I want people to be immune to ad hominem attacks and able to interpret an article about new science on the strengths of the information therein. Thirty seconds reading almost any comments section on the internet is enough to prove that what I want isn’t how things are, though. As my grandfather used to say, wish in one hand, shit in the other, and see which fills up first.
I still don’t know what possesses people to hop on a thread to an article with 500+ comments and counting. Probably the same force that possesses people to begin their comments with “Dear Mr. President …” as though he were just waiting for you to call his name.
It is strange to me that the structure of online commentary hasn’t evolved much past the hyperlink, the flame war, the nested reply chain and the wiki-editor dispute. After 20 years of Internet we’ve just bred more powerful trolls (Breitbart, et al) and sterile gated communities like Facebook. Any referee just gets blamed for bias by the loser. Still. Seems there should be a space/app for cybernetic dialectics, for a free battle of ideas, for … Combat?!
Nah, that’s for gaytards.
What if you’re wishing for shit? And that’s pretty much what one should do, considering human nature hasn’t changed at all in the last 20-years of the Internet age.
Being human, our half-reptilian brains respond more efficiently to fear and anger. This is obviously why blunt illogical statements have a deeper impact and are therefore more “real” and effective on people.
It is pretty (fucking) interesting that you (and these “scientists”) assume elaborate “intellectual” arguments are more correct merely because they are presented without insults or cuss words. I don’t see any reason why the exact opposite could be true. Maybe there is an inherent truth to simplicity and brute force?
Many “intellectual” arguments merely carry the illusion of logic. A great example of that is exactly what you mentioned… “anthropogenic climate change.” Even a cursory glance at the record shows that Earth’s climate changes so radically and unpredictably that it would be naive to expect it to remain static. The Vikings grew barley and tomatoes in Greenland and the English grew wine grapes for hundreds of years during a (relatively) warmer medieval period that existed long before the industrial revolution. This kind of temperate farming is, of course, impossible now as the northern hemisphere is (relatively) much colder.
But, these facts will likely be instantly erased by your preexisting beliefs and the psychological brawn of herd mentality, which you refer to as “collective action.”
So I could…
1) Attempt to create a coherent “intellectual” argument disproving the neoliberal propaganda of anthropogenic climate change by using academic words, citations, and bullet points, to generate the illusion of intellectual domination. In this framework I could easily dismantle every myth of global warming with valid data, and yet it would probably not even dent the dogmas of “true believers.”
2) I could just say “Climate change is fucking gay.” To appeal to people’s moral degeneracy and overall apathy in order to cause them to not-give-a-fuck. – This would be easier and probably have a wider impact.
3) Ignore people’s inability to think independently altogether and accept the cruel illogical universe we live in as a joke, democracy as the science of fools, and Internet comments as merely obnoxious.
This last options is, of course, impossible for a “blogger,” because comments are the coin of their realm and their entire Internet addiction revolves around generating, critiquing, cataloging, and mediating….. comments.
Reed Perry has now surpassed Sarah’s Abusive Husband as the greatest commenter in Combat! history.
“It is pretty (fucking) interesting that you (and these “scientists”) assume elaborate “intellectual” arguments are more correct merely because they are presented without insults or cuss words.”
That is the line at which I knew I was reading something from Mr. Perry. He satirizes like no other.
Hey Mr. Perry, I suggest you attempt #1. If you’re wrong, you will dismantle every myth about global warming and save billions of dollars in mitigation costs. If you’re right, you will be ironically demonstrating dogma, which I also would enjoy.
——
Now for my next post, I will talk about statistics. If you see no post below, that is because, I, through solely the use of beginning HTML tags, sent it to an entirely different area of the blog, one filled with comments offering penis-enhancing pharmaceuticals, which unsurprisingly, only Dan reads.
I read that study, and it was awesome. A rainbow dropped out of my ass into the toilet while I gleefully read. Here are some thoughts, you may find them valuable.
1.
“But when it comes to reading and understanding news stories online—like this one, for example—the medium can have a surprisingly potent effect on the message.”
Surprisingly potent? We already expect the comments to have a polarizing effect on people, so I wouldn’t call it surprising. That word suggests the study found something unexpected.
2.
“Half of our sample was exposed to civil reader comments and the other half to rude ones — though the actual content, length and intensity of the comments, _which varied from being supportive of the new technology to being wary of the risks_, were consistent across both groups. The only difference was that the rude ones contained epithets or curse words, as in: “If you don’t see the benefits of using nanotechnology in these kinds of products, you’re an idiot” and “You’re stupid if you’re not thinking of the risks for the fish and other plants and animals in water tainted with silver.””
Help me out, am I wrong or does this study have two independent variables? From the NYtimes article:
Therefore the study exposed people to.
1)civil comments against,
2)civil comments for,
3)incivil comments against,
4)incivil comments for.
There’s no variable for for/againstness. Seems strange they didn’t want to assess whether the for/againstness of a thread. Reading support for and against stuff affects me, at least.
Or maybe entire threads weren’t for/against, maybe each person got a potpourri of comments, some for, some against! The study doesn’t specify. The study also doesn’t specify whether all the comments in a thread had the same civility, or if civility varied from comment to comment while was for/againstness was controlled.
But both civility and for/againstness they can’t both be varying, can they?
3.
Table 1 shows predictors of nanotechnology risk perception, meaning, as variables with high beta coefficients vary, so does the risk perception. When variables with low beta coefficients vary, they don’t correspond to changes in risk perception much at all.
You might notice that Civility has a really low beta coefficient (-.04), suggesting it doesn’t change people’s perception of risk very much. It is so slight it’s not even statistically significant. Shit. Shut the computers down. Go home. Study’s over.
Well, our researchers aren’t done yet. The analysis necessarily gets a little more sophisticated at the bottom of the table, with interactor variables composed of the product of two other variables.* Two of the three ones they looked at show significant findings, but they’re still quite minor. .09 for Support across Civility and -.07 for Religiosity across Civility. I might say something at .25 was moderate, but anything below it that is mild. And these figures are like, half as important to perceptions of risk as something moderately important.
If your stats are rusty, a beta coefficient 1.0 means two variables are the same thing, and 0.0 indicates a completely random relationship. In real-world statistical analysis, I think .35 would be a strong relationship. It’s probably the kind of thing you wouldn’t be surprised to find out, like how the square footage of a house relates to the amount of natural gas used annually to heat a house. Duh. But even that is no .6, which would be the coefficient of climate as a predictor of gas used annually. That’s way more important.
Your religiosity, when also exposed to civil/incivil comments, accounts for .07 of how you assess risk? Your support for nanotech going into an article (and comments) accounts for .09 of how you assess its risks? Pish posh. I’m guessing now, but I would imagine the presence of sunshine outside a window in the room when someone reads an incivil comment has a beta of .15. That matters, but it’s not controlling. Making a sound argument, or using emotional rhetoric probably could pull more than .15, if we could design valid tests for them.
Stated another way, .85 of the factors influencing risk assessment are not sunshine. .81 of other stuff (some other variables in the study and mostly ones unstudied) is accounting for your risk assessment, not your levels of religiosity, when you’re also reading comments.
4.
“they often changed a participant’s interpretation of the news story itself.”
That’s not in the study. That doesn’t mean it is irrelevant, but it does mean we can’t generalize it to the broader population with confidence, i.e., not being an idiot. We can generalize most of their other findings (anything with an asterisks).
5.
“Until then, beware the nasty effect.”
I have to say, both the NYtimes article and the study do a good job of not exceeding the findings. This final line doesn’t really exceed them either. I just highlight it to add the unspoken additions which more fully reflect the findings:
Until then, beware the nasty effect. Also getting older and having an opinion. Those are more influential to your risk assessments.
Reed, I imagine you may have experienced some shock seeing your name on a blog last week. Shock quickly becomes defensiveness and defensiveness quickly becomes aggression. One can quickly find himself at a loss, hurt and uncertain. At least I can. And we are all human.
I encourage you now to move on to brighter things to come. It sounds like you have a move in your future. I hope it is a good one.
Okay….
The reason why no “collective action” is being taken on the so-called “climate change” crisis is because no such crisis exists. The oceans have not risen. The patterns of nature remain as we have known them. This is (for some reason) difficult for people to observe. The “climate change” non-disaster is the apocalyptic scenario of a certain class of people, similar to Evangelist Christians or neo-confederate survivalists who are preparing for civil war.
Anthropogenic climate change (ACC) is a joke first because the climate of the Earth is constantly and rapidly changing. Digging into the climate record, as anyone can do, will reveal a bizarre, tumultuous history of extreme weather. It’s impossible to choose a “baseline” or climate “status quo” with which to judge some kind of place we “ought to” be at. Accurate measurements of global climate have only been around for some 60 years (if that) and even those portray a shifting and uncertain environment.
As for pseudo-scientific councils that have authoritative names (such as the IPCC), my best recourse in denouncing them would be either the Wikileaks papers that were hacked out of the “climatologists” emails, or the (BBC funded) doc called “The Global Warming Swindle” (http://www.youtube.com/watch?v=YtevF4B4RtQ) in which global warming ideology is exposed for it’s dogmatic absurdity.
As I mentioned, eras such as the Medieval Warm Period prove that modern notions of “climatology” are delusional. If the Jet Stream were to shift today, allowing Britons to once again farm wine grapes, then ACC lunatics would still blame fossil fuel emissions. The fact is that climate played a major role in the collapse of Rome, the Anasazi, and many other major civilizations who blossomed during stable periods. Our world is in a constant flux. I would be more afraid if it were a goldfish bowl of static nothingness. That would really be evidence of some sick all-powerful God who maintained some kind of isolated environment.
A basic course in biology will learn you up with the fact that increased amounts of carbon dioxide increase plant growth and oxygen output. Horticulturists and pot growers know that this (http://en.wikipedia.org/wiki/Carbon_dioxide_generator) will not only augment plant growth, but ultimately reverse carbon saturation. Of course, the most prolific greenhouse gas is water, but most people put undo focus on the minor – carbon dioxide, which actually draws other heavy gasses out of the air in the form of rain, which can be acid rain when it includes sulfur or other chemically active elements.
Further, I would argue ACC is a non-issue due to the fact that there is a limited amount of oil (see: Peak Oil) and therefore any real or imagined damage caused by burning fossil fuels will have a “naturally” limited scope of impact.
In conclusion, so-called climate change is a massive environmental distraction that not only may or may not exist, but detracts from the truly devastating aspects of industrial civilization. Awhile the herd obsesses over “carbon taxes” or whatever, the rainforests of Africa and South America continue to be defiled, nuclear monstrosities cause incalculable meltdowns/mutations, and the oceans fill up with garbage. Paying into any government, corporate, or “non-profit,” endeavor that supports “global warming” is supporting a non-issue that has become a political toy for wealthy politicians.
“Global Warming” is a fucking joke.
And no, Mark. I Actually believe this shit and did not even know about this blog until I was made aware of it by a random, due to the fact my (stranger) neighbor was venting, dropping full names and personal info, doxxing me. If I had known it existed prior I would have been happy to challenge it’s shallowness.
Good Night.
Is it satire?
I mean, it matches this climate change denial flowchart–that the damn twitter informs me is the most shared piece on slate–so closely as to indicate possible intentionality…plus the cherry on top of peak oil, a canard as old as the hole in the ozone layer.
http://www.slate.com/articles/health_and_science/climate_desk/2013/03/climate_change_flow_chart_how_to_win_any_global_warming_argument.html
It’s not a very interesting or authoritative image, but it’s what’s bouncing around the echo chamber. Though, I agree that climate gate revealed a conspiracy so wide it inculcates 95% of climate scientists, I’m not aware you also reject authoritative information from the energy establishment. So behold, your peak oil position undermined in the Wall Street Journal. http://online.wsj.com/article/SB10001424053111904060604576572552998674340.html
If you want, I could probably find peak oil debunked in just about any major newspaper, journal, or energy industry publication. I want to accommodate your high standards for evidence.
I wonder, does Reed see the irony of posting angry, curse-laden conspiracy theory rants on a blog post commenting on…the use of angry, curse-laden comments on Internet posts? I mean, it’s just too perfect. So, maybe it is satire.
“Reed Perry has now surpassed Sarah’s Abusive Husband as the greatest commenter in Combat! history.”
You don’t appreciate anything, you know that?!?
Also, I was beginning to wonder if Dan had been unfair to Reed Perry with his description of him as a generally jerkish and inconsiderate neighbor. Then he showed up here and started commenting and, well, let’s just say my opinion of him has not improved.
Back to article… a couple things based on my limited stats knowledge:
1. Beta coefficients tell you how much one standard deviation change in the independent variable is associated with (or predicts) a change in the SD of the dependent variable.
2. Many researchers don’t report betas for interaction and binary (dummy) variables because any interpretation of them is highly dubious.
3. If the betas were interpretable, however, I don’t agree that a beta of 0.07 or .09 would be unimportant. This is sociology — a totally different level of complexity than predicting fuel use based on house size — and there are bound to be many factors that are important predictors.
4. The beta does not tell you how much of the variance in risk perception is accounted for by the independent variable. There is no single number for a particular independent variable that tells you that in a multivariate regression. The R-squared tells you how much variance of the DV the total model can account for. As the table shows, the introduction of the interaction variables increases the R-squared by about 10%, which is highly significant. Changes in R-squared are a much better way of telling how important the interaction variables are to risk perception.
5. The overall R-squared is 17%, which is not amazing as the authors note. They suggest that what is important is that change in R-square from the initial model to the final model. It’s true as I’ve said that that’s a highly significant change (overall change in .13 for R-square). But the implication is that this significant change is independent of any other variables you might subsequently add to this model and that’s not the case. Once you start adding more variables to make a more complete model, the earlier effects could disappear completely.
Thanks for your post Leif. It’s quite likely your limited stats knowledge exceeds mine, and I appreciate you sharing it.
“The beta does not tell you how much of the variance in risk perception is accounted for by the independent variable.”
…Because it’s telling you how changes in standard deviation are related? An while the SD of a variable and the value of the variable are obviously related, they aren’t strictly the same thing so I am overdoing in my interpretation? Is that right?
What, I’m a jerk because I have an opinion that deviates from your herd-mind, or because I respond when someone is insulting me and doxxing me online? Is it a surprise to you that someone would respond, “Mike” or “Mark” or whatever the fuck?
As this fucking ridiculous article proves, my comments will be read more and taken more seriously, so continue blipping around the Democrat blogosphere wasting your life posting comments that are so fucking boring they’re virtually unreadable.
Combat!blog is now my favorite place to troll.
Grr, spamcatcher trying to prevent me from posting. Allow me to rewrite my post so it does not detect it as a duplicate comment.
Thank you Leif. Your post demonstrates that however limited your stats knowledge, it likely exceeds mine. I appreciate the explanations.
You said the variance in risk perception is not accounted for by the beta of the independent variable (#4). Is that because the beta is really about how changes in standard deviation are related between two variables? And while the values of the variable are obviously related to the SD of a variable, they aren’t strictly the same thing? Is that right?
The beta is the regression coefficient (the slope) of the variable expressed in standardized units. That means you’ve subtracted the mean from all variables and divided by the SD of that variable. So your mean is now 0, and your units are now in SDs. If your variable was money, for example, with a standard deviation of $20, an increase in $1 now means and increase in 1/20 of SD ($ units cancel out). If you do this for your outcome variable (DV), then you get an equation that predicts SD change in DV with SD change in IV. So they are really are sort of the same thing. The reason for doing it is mainly that you can now compare coefficients of different IVs, which you can’t do when the IVs are in different (unstandardized) units.
This also indicates part of the reason why it’s such a problem to transform binary variables into betas: What does it mean to talk about a SD change in courtesy, if the SD of courtesy is, say, .13 but you’ve only coded values 0 and 1 for courteous/not-courteous? Not much.
The betas (or the unstandardized regression coefficients) are your predictors of the outcome. They are not the outcome itself, which is much more variable. It’s easiest to think about this as a bivariate regression, with one IV and one DV. Your actual DVs will be scattered all over the graph. Your regression line (line of best fit) will be running through these. The coefficient for the IV (plus an intercept) gives you that regression line. The coefficient could be big or could be small, but that wouldn’t tell you how much variance there is in the actual DV. That’s what your R-squared tells you: if it is equal to 1, then your DV’s line up exactly on the predicted line (the regression line). The more they vary from that (roughly speaking), the lower the R-squared will be. So you might have betas for two IVs of .1 and .9. That tells you the latter IV is a stronger predictor than the former. But you could have those same betas in different models with different R-squares. Could be 1 or could .0001. The first model accounts for every bit of the variance in the outcome variable and the latter accounts for almost nothing.
* should say you subtracted the mean from all values, not variables.
I don’t normally comment on anything on the internet, even if I really like it, just because I’m one of those weird people who’s averse to having conflict with people I’ve never met (or people I have met).
But I wrote a blog piece on Kitty Werthmann, the Austrian woman who lived through Nazi occupation only to spend the last five years shilling for the Tea Party and comparing Obama to Hitler. And holy crap, the comments just keep coming.
I posted it two months ago, and it’s close to 400 now. A few of them are thoughtful and reasonable, but most of them are totally insane. Some are manifestos about school shootings or bailouts or inflation or mind control chips. One lady posted ten times shilling her own book. And at least 100 are variations on “wake up” or “open your eyes” or “New World Order.”
The only thing I’ve learned from them is that if you write anything about Obama that doesn’t out and out say he’s Antichrist Muslim Hitler, the crazies will find you.
Read at your own peril: http://skeptoid.com/blog/2013/01/07/kitty-werthmann-history-distorted/