Wednesday, April 06, 2022

Policing Your Own

Interesting study about media reporting, (twitter thread by one of the authors, with followup here) and it ties into the lack of media scrutiny in Sweden over covid. I have a discussion link for that too, though it is from an Australian woman who I know nothing about, and thus am uncertain how much she is shading or cherry-picking her information.  I will wait until I have a better. As we are used to twitter threads being used for rants, sometimes quite clever ones, we lose how much actual research scientists are using the form to talk with each other now.  This is especially true as here, when the study is in preprint and not yet reviewed, and they use twitter as something of a group editing function.  Did we miss anything?  Does this mean what we think it does?  Are there downstream implications we haven't been thinking of?  Is there other research that bears on this? I was surprised a few years ago listening to Razib and Spencer Wells talk about genetics research coming out this way. One of them laughingly said it is so common now that they were suspicious of research that showed up in a journal which no one they knew had heard of before. It seems to be the new version of peer review, a system that worked well until a few decades ago but has gradually broken down because of volume and specialisation.

The study paid regular Fox viewers to watch CNN for 7 days, then asked questions. They tried to get funding to do the same with regular CNN viewers (who they openly stated they thought would be as bad or worse) but were unable to.  What they found - oversimplified - was that it didn't change their votes, but it did affect their willingness to call their own side to account. Being exposed to another POV made you more willing to call the worst of your own side out. This strikes me as a very good thing in itself.  

WRT Swedish media reporting on covid triage, reporting basically wasn't there. Even months later it was unreferenced and Swedes had little idea that this had happened. Even the government commission appointed to look at the whole response, while it did include some critical information, did not include the switch from active treatment to palliative care as a matter of common practice. Parents were encouraged not to tell others if the children had covid as it would "spread fear."

The thread I was reading also noted how discouraging it is that our initial estimates of what was going to be required for herd immunity were too low. There was some prediction that 60-70% was going to be enough to at least greatly reduce spread. That turns out to not be close, and estimates of what will be necessary are now more likely to be 90%.  This is territory where I have no expertise, so I can't get a sense of whose answers are better.  Because one of the things I rely on in such matters is noticing who is fighting fair, I am hampered when no one in the discussion seems to be accusing, posturing, or ignoring the others. I also don't always have a clear idea of the value of various credentials on the issue. Is a epidemiologist better than a statistician here?

Yet back to the point of the study. I had not predicted any such thing, but once heard it makes a kind of intuitive sense. When you see what the other side is actually getting exercised about, you find that it is somewhat different from what you have been told they are exercised about.  I have long complained that liberals do not often refute conservative arguments because they don't even know what those arguments are. Working among liberals, and within a media universe that is default liberal, it has been frustrating to keep hearing "well, the conservatives think..." and have it be something that I knew a few people who thought that, but mostly, no, it was a fringe opinion.  Over the last few years, and not just around covid, I have seen the same thing from conservatives, and conservative sources. The usual tactic is to find particularly egregious examples of something an opponent has said and blasting that out as major news. 

When one reads or listens to the actual source, one finds what they are more often saying.  You might still think it is wrong or stupid, but at least you have the data right.  This in turn gives you into an insight into your preferred sources.  You start to notice that a few of them are linking to article after article of outrage about what those terrible Others are teaching our children or trying to pass legislation about in distant states, but you now know that this is not really happening all that much. You immediately notice that many of those articles contain exaggeration or supposition, or do not record possible responses from the players in question.  When it happens often enough, that source drops out of your rotation of places to check. With your newly freed-up mental space, you find that one of your less-used sources is actually a better one. You get some idea what sources - and what public figures - are more reliable.

One younger friend who has a liberal brother asked him what was a reliable indicator of what liberals are actually saying and he chose Vox. That was a couple of years ago and these things change, but it's a possibility. I used to be able to go to The Nation, but I gave up on that almost a decade ago, as I found that they were publishing those on the left who were merely dissenting voices, not ones who had solid work or represented at least some others.

No comments: