Talk:1877: Eclipse Science

Explain xkcd: It's 'cause you're dumb.
Jump to: navigation, search

I'm pretty sure his point was actually that in the modern day eclipses are pretty well understood, easy enough to travel to, and provide very little actually unique circumstances, so in reality there is very little scientific value at all. Just a really cool thing to see 15:56, 16 August 2017 (UTC)

Yet still there are some aspects of eclipses such as shadow bands which are unexplained. There are some theories, but eclipses are rare enough, plus shadow bands don't occur with every one, so there is no definitive explanation yet.
RChandra (talk) 18:56, 16 August 2017 (UTC)
The explanation of shadow bands on wikipedia seems pretty solid to me. Zmatt (talk) 11:51, 17 August 2017 (UTC)

To me the comic is poking fun at the idea the media (and by extension laypeople in general) perceive eclipses to be of great scientific importance and that scientists are excited about it for that reason. Note how his almost every utterance contains "science". Megan deftly deflects his attempts to put words in her mouth and remains resolute in her stance that eclipses are interesting to everyone on their face value but not necessarily more so to scientists that others. 19:35, 16 August 2017 (UTC)Pat

I guess Randall is sad that people outside the band seem to have less interest in the eclipse, like shown in the previous comic, and that they might not wish to travel a few hours to see a once in a life time spectacle, or as he also feared in the last comic, try too late to get there and get stuck in the trafic jam outside the totality zone. Being in the 99.9999% dark part is nothing compared to being inside the zone being able to see the corona (the ring in the sky). Go and see it if you have any chance of doing so! --Kynde (talk) 21:13, 16 August 2017 (UTC)

Should we make a new category for eclipse comics? 11:53, 17 August 2017 (UTC)

If he continues tomorrow and Monday maybe. There are so far only really two. The other two comics mentioned in the previous comic only briefly mentions the eclipse. So it is really only two comics so far. And they are not related in their individual subject. --Kynde (talk) 12:30, 17 August 2017 (UTC)
While I agree an Eclipse category would be uncalled for, this is the third comic I can think of without looking back further: 1868, 1876 and now 1877. If Randall continues the theme tomorrow / today and Monday, that's a total of 5. Maybe a post-eclipse comic on Wednesday, then I expect he'll be done on the subject unless something big happens (like a plane crash by the eclipse being in their eyes or something. we can blame Randall for that example coming to my mind, LOL!). 6 and done doesn't seem to warrant a category. :) NiceGuy1 (talk) 06:52, 18 August 2017 (UTC)
"Puts on sunglasses" got a category after three comics. 09:31, 19 August 2017 (UTC)

The explanation claims, "While some astronomers might be testing elaborate hypotheses during an eclipse, for other scientists (eg. organic chemists and herpetologists) it is just a once in a long time (maybe even once in a lifetime) event which is visually interesting". This is at odds with the existence of several citizen science projects encompassing not just astronomy but also atmospheric sciences and animal behaviour. 13:00, 17 August 2017 (UTC)

There's also a decent number of science experiments that are worth repeating mostly to take advantage of improved instruments and so people get to see them--which can also be considered under 'improved instruments,' if part of what you're doing this time is recording the entire procedure on video. 22:38, 17 August 2017 (UTC)

While the eclipse may not provide very many more opportunities to advance scientific research and knowledge, it is a terrific opportunity for science teachers at the elementary level to demonstrate the basic science of orbits and positions of the Sun, Moon, Planets and Stars. As well as an opportunity at the secondary level to talk about the laws of gravity and motion and how predictable it all is that we can calculate the exact path of an eclipse hundreds of years in advance. Too bad this is happening during summer vacation. It is also the kind of event that excites children enough that they might consider pursuing a career in science. So while astronomers may find it old news and not providing new research opportunities, there is still a lot of science teaching opportunity. Rtanenbaum (talk) 12:36, 18 August 2017 (UTC)

So there's a couple points here:

  • Public demonstration of previously performed science is not really the same thing as actual experimentation where the outcome is in doubt, the latter of which advances our understanding. So yes, you can show something to people who haven't observed it directly before, but then you're not using the word "science" in the same way as Megan.
  • There actually are new experiments being performed, not just old experiments with a new audience or improved equipment.
  • There is a tendency among non-scientists to underestimate the certainty of scientific or statistical results. Some examples are "you can't teach evolution as settled science because it's only a theory" and more recently "this poll only sampled a thousand adults, which is 0.03% of Americans, so it's meaningless because we have no idea what the other 99.97% think". So scientists are expected to keep verifying and re-verifying fairly basic results. D5xtgr (talk) 16:29, 21 August 2017 (UTC)
That last one is hilarious, for reasons I doubt you realize. Some of us here know how to get a proper population sample--and, well, with that poll? I don't even know where to start on the potential sources of error. 'Too small' should appear on the list several times, though, since a lot of the sources of error can be controlled for by simply adding to your sample. (Yes, yes, changing your sampling techniques will also do it, but it's just plain easier to add more people, so you start there.) This is actually part of why being able to replicate results matter--plus, the training in how to do it properly does give you a pretty good idea how to rig your poll to produce the desired results, and what to watch for when reading others' research. 21:39, 24 August 2017 (UTC)
Reasons you doubt I realise? What are you on about? The point is, the fraction of the population appearing in your sample doesn't need to be very large to have a margin of error below five percent. Margin of error for a representative sample of 1000 from an infinite population is 3.2 percent - the hard part is to get a representative sample of adults rather than a biased one. And then you go on about "that poll"... your meaning is very unclear. D5xtgr (talk) 23:17, 29 August 2017 (UTC)