Friday links: can you identify any ecosystem on sight, why papers get rejected, your state bird is lame, and more

From Jeremy:

Do not click this link if you ever want to be productive again: Geoguessr is a new free online game from Google. You’re shown a random location somewhere in the world, as a 3D rotatable, zoomable panorama. Your job is to guess where it is. Or rather, guess where you are, since the experience is as if you were standing in that location and able to look and walk around (within limits; the game won’t let you wander more than a short distance from your starting point). You get points based on how close you guess is to the true location, as the crow flies. Each game consists of 5 different locations. It’s totally addictive. I’m sharing this here because it really helps if you’re a good botanist. Many of the locations are wilderness, so that your main clue is the local vegetation. One nice feature is that, after you finish, you get a URL you can share so that others can try to beat your score on the same five locations. Probably every reader of this blog will be better at Geoguesser than me, since I am a crap botanist. So if you want to prove your botantical chops by beating my pathetic score from my first try at the game, go here. Share your score in the comments! And no cheating and googling words you read on roadsigns or anything like that! You’re only allowed to use the information provided to you by Geoguesser, and your own knowledge. Oh, and I’ve since scored much better (my high so far is 19,210)–it’s possible to nail urban locations within meters, which really jacks up your score.

A while back we discussed how the Ecolog-L listserv, pioneering as it was originally and as valuable as it’s been to ecologists over the years, may be in need of updating or even replacement. An alternative has now been proposed. The popular Stack Exchange family of sites, where you can post questions and get answers from a knowledgeable user community, is testing out an Ecology Stack Exchange. In my admittedly-limited experience, Stack Exchange is a wonderful resource–there are already very popular Stack Exchanges for academia, biology, and statistics, among many others. And Stack Exchange has a bunch of modern features–such as tagging and up-voting of answers–that make it way more usable, and far more troll-resistant, than Ecolog-L. Plus, it’s just as easy to use as Ecolog-L, and just like Ecolog-L it’s totally free and open to anyone to ask and answer questions. Really, it’s what Ecolog-L would be if Ecolog-L were invented today. If you’d like to see an Ecology Stack Exchange happen, go over there and propose some sample questions, and participate in the discussions as to what an Ecology Stack Exchange ought to cover. (HT Noam Ross, via the comments)

From Nature this week: six red flags for irreproducible papers. The authors previously showed that the majority (!) of papers on preclinical cancer in top-tier biomedical journals cannot be reproduced. Here, they reveal key differences between the irreproducible papers and the reproducible ones. The irreproducible papers tended to lack blinding, had only been done once, inappropriately chose not to show or report all the results, didn’t show results from crucial positive and negative controls, didn’t validate their reagents, and used inappropriate statistics. Some of these issues (like reagent validation and blinding) aren’t relevant to ecology. But others–like selectivity in which data to analyze and which results to report–absolutely are. This is something I’ve talked about before. And here’s one possible solution: a registry of planned studies, or disclosure requirements.

The program for the 2013 Ecological Society of America meeting is now up. Time to start getting excited–seriously, I love the ESA meeting!

A slickly-delivered lecture or talk doesn’t improve the listener’s understanding or memory of the material any more than a badly-delivered one–but listeners think it does. (HT Felix Salmon)

Every state in the US has an official state bird. Most of them are lame choices, and many states have made the same choices, as if they were all copying one another. Here’s what the state birds should be, courtesy of The Birdist. It’s very funny; here’s a sample:

I am finishing this post the next day because I had to go buy a new computer after I threw my last one out the window when I read that Florida’s state bird was the Northern Mockingbird.  I cannot think of a lamer choice.  What’s their state beverage, A Half Glass of Warm Tapwater?

Someone should do a follow-up post for the state flowers!

And finally, stay safe in the field by consulting this handy, and quite funny, snakebite FAQ:

Q: No, what I really meant was what if I’m way the heck out there, and I don’t have anybody else with me, and it’s a really severe bite, and I’m going into shock, and there are hungry bears all around me?

A: You may get eaten by bears.

(HT Jeremy Yoder, via Twitter)

From Brian:

Here is a nice link to seven common mistakes in paper writing, which lead to papers getting rejected.

From Meg:

I think I’ve only been asked once to comment on a new paper published by another lab. In that case, I was traveling and didn’t get the request in time. And, really, I was kind of relieved about that, because the thought of doing so intimidated me. So I found this post from Ed Yong entitled “A Guide for Scientists on Giving Comments to Journalists” interesting.

Over at Small Pond Science, Terry McGlynn asks what your students call you. I’m interested in others’ answers to this. As I say in a comment there, I am considering addressing this at the start of my Intro Bio course the next time I teach. I think a lot of freshmen don’t realize that I have a doctorate, and that “Dr. Duffy” (or “Professor Duffy”) is the appropriate address, not “Mrs. Duffy” (which I get pretty regularly).

Hoisted from the comments:

Brian’s dad (!) on what it takes to do policy-relevant science. It would seem that Brian’s recent post should’ve given a HT to his pop.😉

Lots of useful career-development stuff in the comments on Carla Davidson’s guest post on non-academic careers. Commenter Mike points out a great resource for those thinking of switching to a non-academic career track: Branching Points. Includes lots of practical advice about what sorts of careers are out there, how to gather information and decide what path would work for you, recognizing your transferable skills, and much more. Frequent commenter Arne Schroder notes this old but still useful article on recognizing your transferable skills. And commenter Sandra Chung is co-organizing a panel session at the Ecological Society of America meeting on non-academic careers for ecologists.

14 thoughts on “Friday links: can you identify any ecosystem on sight, why papers get rejected, your state bird is lame, and more

  1. Blinding is relevant to ecology and all science. A huge component of the published results is biased because the data were analyzed unblinded – that is with knowledge of “assignment” both in experimental and observational studies. Data should be analyzed “blinded” although I’ve never seen anyone do this, nor have I really done it myself. Randomly assign the real group names to fake group names. Write up the results. Then reveal the mapping between fake group and real group and substitute in the results.

  2. This post is SO not helping me finish my dissertation (except maybe the seven common mistakes in paper writing; thanks, Brian)…

    My score was 8919:
    I actually found that road markings and signs were more useful than plants. For example, the first one reminds me of northern New England vegetationally, but we don’t paint our roads like that, nor do we have little blue signs like that. Two others had signs on the left side of the road, rather than the right, which really narrows down the possible countries…

    • Oh yeah, road markings and the style and placement of road signs (even if the signs aren’t readable) are definitely helpful, at least in eliminating some countries. Unfortunately, it’s not uncommon to get totally unmarked dirt roads in the middle of nowhere, which means vegetation and the general terrain are the only things you have to go on.

      I do wish there was a way to move down the road more quickly (or is there a way I haven’t found?) In remote areas, where I can see that there’s no road sign on the visible portion of the road, I’d like to be able to skip to the horizon rather than have to slowly “walk” down the road by repeated mouse clicking.

    • I’m claiming the lead with 10573 points (here) and hoping this allows me to go back to doing real work! (although I may have cheated by practicing on some other scenarios first)

      Extremely interesting – language, cars, roadsigns, architecture, people all come into it. Its a real lesson in convergent evolution too. Getting the biome is easy, but with the level of fuzziness in the pictures it is sometimes impossible to get the continent (I counted 2 out of 5 where both Jeremy & I got the biome right but we both got the continent wrong).

      • Glad to hear I’m not the only one who’s gotten the wrong continent more than once. My biggest miss so far is by 18,800 km. It wouldn’t have occurred to me that New Zealand would have even a passing resemblance to any part of Northern Ireland…🙂

        And I’ve subsequently built up a bit of pride by scoring 19,000+ a couple of times, so I won’t even hassle you about wimping out and practicing a few times first before taking me on.😉

      • Yes, although Tobias got lucky and got a location close to where he grew up.

        Based on the locations I tend to get (which are by no means a random sample of all terrestrial locales; clearly there are many countries yet to be visited by Google’s camera cars…), people who grew up in New Zealand, the Australian outback, the US southeast, and northern Canada should be awesome at Geoguessr.🙂

      • A couple of hard deserted roads in this set. I’m annoyed about the Canadian location – I choose between Newfoundland and north Canada towards Beufort sea, and selected the wrong area. Otherwise the score could have been really good.

      • Yep. If you get a deserted dirt road in northern Canada, you pretty much just have to guess and hope to get lucky. Because northern Canada is a big place, large swathes of it all look the same, and there are no location-specific clues like signage.

        Score is nonlinearly related to how close you get. The difference between between missing by half the globe and a few hundred km is only 2000-2500 points, but the difference between missing by a few hundred km and nailing it to within meters is 4000+ points. My really high scores have come when I’ve gotten more than one urban location I was able to nail within a few meters. If you don’t have at least one of those, then scoring much above 11,000 or so requires a massive amount of luck.

    • A location in my first round happened to be where I did my Masters research (the Missouri Ozarks). I looked at it in disbelief: did Google seriously just send me to a few miles from one of my field sites? Yup. I have since performed much worse…

      • And we have a winner! I was waiting for the first person to recognize their own field site on GeoGuessr–but I didn’t think it would happen so quickly!

      • Well, I worked in Ozark rivers and was a few miles from one of my stream access points – so not an exact match, but awfully close. I had to convince myself I was right: that’s just outside of West Plains, isn’t it? Am I crazy? I was not crazy.

Leave a Comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s