|
ooh I spotted the witty part...
|
#
?
Sep 29, 2021 19:08
|
|
|
|
| # ? May 27, 2024 19:18 |
|
|
Ew Jersey.
|
|
#
?
Sep 29, 2021 19:26
|
|
|
Nited States of America because the power infrastructure is gonna break and leave everything dark.
|
|
#
?
Sep 29, 2021 19:33
|
|
|
All my exes live in exas.
|
|
#
?
Sep 29, 2021 19:39
|
|
|
Phanatic posted:And researchers have spent entire careers producing useless output by mashing together enormous datasets and looking for the statistically-significant noise that will inevitably be found if you do enough of that. Oh dang. I've referenced the soup bowl thing positively before, so it's a bummer to find out this dude sucks.
|
|
#
?
Sep 29, 2021 21:04
|
|
|
Assachusetts checking in. (Though currently in Ew York).
|
|
#
?
Sep 29, 2021 21:46
|
|
|
BonHair posted:"you should always do a linear axis plot" doesn't sound like particular good advice honestly. Especially when given a specific hypothesis to test that doesn't even require plotting. I don't know, given two numerical values, a quick scatterplot is a quick and easy sanity check on your data. I always want to get a feel for what the numbers I've been given look like before I start running anything more in-depth.
|
|
#
?
Sep 30, 2021 13:17
|
|
|
If you are given a bivariate dataset and your first idea isn't to plot it you have broken instincts about data.
|
|
#
?
Sep 30, 2021 14:27
|
|
|
Chin Strap posted:If you are given a bivariate dataset and your first idea isn't to plot it you have broken instincts about data.
|
|
#
?
Sep 30, 2021 14:34
|
|
|
DarkHorse posted:I don't think that's what they were driving at, more just "play around with the data some first instead of going straight for your goal" And every time you look to check for significance you do you need to shrink your target probability. Quick and dirty method (Bonferroni) is to just divide by the number of tests (so if you're using 5% and do two tests, you actually need 2.5%), but there's some nuance in what actually constitutes a 'test', and it's generally considered to be erring on the side of too-harsh. The other advice I got was that if you eyeball a graph aimlessly you should count that as running a couple statistical tests, and that if this ended up being a problem for your confidence interval, you should design a better experiment.
|
|
#
?
Sep 30, 2021 17:00
|
|
|
Tunicate posted:The other advice I got was that if you eyeball a graph aimlessly you should count that as running a couple statistical tests, and that if this ended up being a problem for your confidence interval, you should design a better experiment. This is bad advice. Either you're doing exploratory analysis, in which case you shouldn't be running tests, or you're doing confirmatory data analysis, in which case you should only be plotting your data for quality control. If you're using the same data set for exploration and confirmation, you're doing something wrong.
|
|
#
?
Sep 30, 2021 17:06
|
|
|
eh, insofar as p-values are useful, you should still pay attention to them when doing a pilot study. You probably shouldn't get excited and set up a next experiment if you try 20 hypotheses and one of them beats the 0.05 mark.
|
|
#
?
Sep 30, 2021 17:52
|
|
|
 I can't tell if Poland is indecisive or coming straight for you 
|
|
#
?
Sep 30, 2021 21:35
|
|
|
So, wait ... That's what the pedestrian crossing signs look like, isn't it? I was so certain it was some sort of political stuff, at first.
|
|
#
?
Sep 30, 2021 22:14
|
|
|
The UK and Ireland guys walking away from each other seems about right.
|
|
#
?
Sep 30, 2021 22:15
|
|
|
No AmpelMann chart is useless 
|
|
#
?
Sep 30, 2021 22:28
|
|
|
DarkHorse posted:No AmpelMann chart is useless Yep they are called Ampelmännchen! The Germans named it first again. https://en.wikipedia.org/wiki/Ampelm%C3%A4nnchen
|
|
#
?
Sep 30, 2021 23:46
|
|
|
The former USSR should really see a doctor about that growth.
|
|
#
?
Oct 1, 2021 00:33
|
|
|
Captain Hygiene posted:
Russia’s is doing the Gimli–Patterson Bigfoot walk.
|
|
#
?
Oct 1, 2021 00:53
|
|
|
Platystemon posted:Russia’s is doing the Gimli–Patterson Bigfoot walk. I thought the same thing about the UK.
|
|
#
?
Oct 1, 2021 00:56
|
|
|
https://twitter.com/TerribleMaps/status/1443628578483183623?s=19
|
|
#
?
Oct 1, 2021 05:48
|
|
|
Chin Strap posted:If you are given a bivariate dataset and your first idea isn't to plot it you have broken instincts about data. 
|
|
#
?
Oct 1, 2021 11:13
|
|
|
Captain Hygiene posted:
I know Poland is always a weird outlier on these maps but I think this one is fake news? The Polish “no pedestrians” sign does look like this, but both the crosswalk sign and the pictogram on traffic lights have a standard walking man. Source: am Polish.
|
|
#
?
Oct 1, 2021 11:47
|
|
|
DashingGentleman posted:I know Poland is always a weird outlier on these maps but I think this one is fake news? The Polish “no pedestrians” sign does look like this, but both the crosswalk sign and the pictogram on traffic lights have a standard walking man. Source: am Polish.
|
|
#
?
Oct 1, 2021 13:08
|
|
|
wait a minute, that's supposed to be "Labama"  that being said, "ew jersey" is totally accurate
|
|
#
?
Oct 1, 2021 19:17
|
|
|
DashingGentleman posted:I know Poland is always a weird outlier on these maps but I think this one is fake news? The Polish “no pedestrians” sign does look like this, but both the crosswalk sign and the pictogram on traffic lights have a standard walking man. Source: am Polish. That makes a lot more sense but I'm also sad to learn it.
|
|
#
?
Oct 1, 2021 19:31
|
|
|
I like the idea that at least one student plotted the data, identified the gorilla, assumed the professor just threw it in there as an easter egg and did the analysis anyway because they assumed the point was to show you could analyze a dataset and not identify something so stupid.
|
|
#
?
Oct 1, 2021 23:56
|
|
|
World War Mammories posted:wait a minute, that's supposed to be "Labama" Yeah I saw that and rolled my eyes. I love how Indiana is now... Ndiana so there's no real change in pronunciation. And Utah to just "Tah" makes me giggle.
|
|
#
?
Oct 2, 2021 05:58
|
|
|
|
|
#
?
Oct 2, 2021 15:00
|
|
|
I once had the pleasure of seeing a biology prof go on a frustrated rant at a 300-level class on invertebrates that unless you saw the spider bite you it probably wasn't a spider because spiders almost never bite, and it definitely wasn't a brown recluse because we live in Canada, no he doesn't care what your cousin in Saskatchewan says, we don't have them, and a dedicated group of students just refusing to accept this and arguing with him, and then it escalating to the prof emailing all of us, twice, with various sources about the non-biteyness of spiders and the normal range of the brown recluse. He'd appreciate this but it would probably be weird to email him this. That was the class I was saddest about when COVID made us go online mid-semester.
|
|
#
?
Oct 2, 2021 15:50
|
|
|
Spread a rumour that the library is infested with Mediterranean recluse spiders, as was the University of Michigan Ann Arbor’s. The Finnish Museum of Natural History has Chilean recluses and those things are even worse.
|
|
#
?
Oct 2, 2021 16:06
|
|
|
RoboRodent posted:I once had the pleasure of seeing a biology prof go on a frustrated rant at a 300-level class on invertebrates that unless you saw the spider bite you it probably wasn't a spider because spiders almost never bite, and it definitely wasn't a brown recluse because we live in Canada, no he doesn't care what your cousin in Saskatchewan says, we don't have them, and a dedicated group of students just refusing to accept this and arguing with him, and then it escalating to the prof emailing all of us, twice, with various sources about the non-biteyness of spiders and the normal range of the brown recluse. You should definitely email him, best case it's hilarious to hear that story, worst case he just ignores you 
|
|
#
?
Oct 2, 2021 18:24
|
|
Yes, it's like a lava lamp.
|
Did you find a spider? YES --> Squish it NO --> Then who cares
|
|
#
?
Oct 2, 2021 19:04
|
|
|
spiders are your friends, they just really want to make you a nice silky blanket to roll up in
|
|
#
?
Oct 2, 2021 20:50
|
|
|
Captain Hygiene posted:Did you find a spider? Rude. Spiders are friends.
|
|
#
?
Oct 2, 2021 21:48
|
|
|
Captain DIEgiene posted:Did you find a spider? Spiders eat the actual annoying bugs while keeping to themselves. Do not squish them please. Also always tell the nice teacher that they made an impression.
|
|
#
?
Oct 2, 2021 21:57
|
|
|
I'm gonna squish any spider I find, I'd rather deal with the bugs they eat
|
|
#
?
Oct 2, 2021 21:59
|
|
|
Spiders are stored in the Georg.
|
|
#
?
Oct 2, 2021 22:13
|
|
|
Spiders are friends and can live in my apartment. They just can't touch me or spin webs or they get put on guard duty outside
|
|
#
?
Oct 2, 2021 22:47
|
|
|
|
| # ? May 27, 2024 19:18 |
|
|
Captain DIEgiene posted:I'm gonna squish any spider I find, I'd rather deal with the bugs they eat The mosquito and housefly defender has logged on, I see.
|
|
#
?
Oct 2, 2021 22:59
|
|
