We have some great ideas but none of us can draw. We need your help with designs and art for any or all of these projects:
1. “Gone Fishing” T-shirt
A person is standing in a boat, fishing. The lake is full, not of fish but of little numbers: “.14”, “.31”, “.08”, etc etc. And also a few shiny glittery numbers like “.04”, “.01”, “.03”. The fisherman is reeling in a glittering “.05”. Then, next to him on the boat is a bucket with a few shiny “.04″‘s and “.03″‘s.
Possible titles:
– Teach the Controversy (as in here).
– This is Science?
– Psychological Science
– Or we could play it straight and give it the title, We Can Do Better Than This.
In any case, all this would be on the back, and then the front would just have a tasteful Stan logo over the pocket. Or something like that.
2. “Garden of Forking Paths” T-shirt
Something similar with a “forking paths” design. I’m not quite sure how to do it.
3. “Tabloid Science” T-shirt
Some sort of collage of all the “power = .06” studies we talk about here: himmicanes and hurricanes, beauty and sex ratio, ovulation and just about anything, MRI’s, etc etc, we must have dozens of examples so we could just cram them in to a single overstuffed image.
Or maybe do it as a mock-up of a tabloid newspaper, with the title “Psychological Science” (in that newspaper-title font like they have for the New York Times) and then various tabloid-style headlines laid out, newspaper style).
Then on the front: “Stan. We can do better.”
Or something like that. Use your creativity!
4. Handy Statistical Lexicon
I think this could work in the form of an illustrated book. This is for someone who really likes to draw!
5. “Second Chance U” and “The New Dirty Dozen” movie posters
I think these could be awesome. See here for the descriptions of the movies, to get a sense of what you’d be working with.
What I really want to see are the movies themselves, but that would take a lot of work. Maybe I can throw together a script or two sometime. Or at least a Gong Show act.
I’m serious!
I’m serious! Anyone who wants to draw any of the above, please let me know. There’s a free T-shirt in it for you. Also glory.
P.S. The submissions (from Ed Witt) are here.
Do you remember the poster that hung for years (maybe still does) in the window of Top Dog in Berkeley, in the food court on Hearst? It was titled “The Path to Success” or something like that, and it showed a road climbing from the base of a mountain to the peak. People are walking up the road, but almost all of them are being diverted into areas along the wayside: “Mutual Admiration Society”, “Bohemianism”, etc.
I’m thinking you could do the same kind of thing with The Road to Statistical Significance….not the exact same thing, but the same idea. “Thought of the Question Before Looking at the Data”, “Didn’t Cherry-Pick Enough”, “Didn’t Try Enough Likelihood Functions”, etc., each could divert people from reaching the shining city on the hill labeled “Statistical Significance”. Actually this wouldn’t work as a tee shirt, it would have to be a big poster like the one at Top Dog.
Or, hmm, how about a cherry orchard (sign that says “Fisher’s Cherries” or something) , with guys wearing lab coats out collecting cherries in baskets, and a caption: “Scientists selecting data for their new study.”
Phil:
These are some good ideas. But I want to avoid the term (and concept) of “cherry-picking” because it’s so negative, and I want to convey the “garden of forking paths” idea that these problems with null hypothesis significance testing arise even when the researcher is not trying to cheat.
That said, with the “fishing” idea, I’m clearly implying that researchers are looking around for statistical significance. Somehow, to me, “fishing” does not sound as bad as “cherry picking.”
I think of fishing and cherry picking as different things: Fishing is where you stop when you get something good enough; cherry picking is where you get lots of possibilities and choose the one(s) you like best.
Yes, I like this way of describing the distinction.
1-3 sound extremely detailed for a T-shirt. Not that I’m an expert on T-shirt design, but I would think that T-shirt images should be simple, clear, and readable from a few yards away.
Raghuveer:
Maybe a good artist/designer could capture the idea without requiring too much detail. You could think of my descriptions above as background that could be adapted by the artist to make an effective image.
Using isotopyes can be effective for the forking garden metaphor.http://thenounproject.com/term/network/27303/ or for the main site http://thenounproject.com/
Using a rod and reel always leaves the chance that the fish won’t bite. Should the fisherman be using dynamite?
I think I’ve pre-emptively stolen idea #1 for my book cover. We were tossing around ideas and “fishing for significance” came up, resulting in this:
http://www.nostarch.com/sites/default/files/imagecache/product_full/statsdone-wrong.png
Not quite the same — we decided that putting 0.05 on the end of the hook wouldn’t be understood by some of the audience, so we just used a plot. But it’s close. I guess the fishing metaphor is universal.
OK, I should be working right now but
Statistics done conventionally right, no guarantee of a catch that’s certain to replicate.
(Perhaps fishing to stock one’s pond or private lake.)
Is the person in that picture intended to be RA Fisher? If so then I think that you should change it. He is hardly a model of how to do statistics wrongly, and if you think that he is then you have not read his work.
No, the person is just a generic scientist.
“Teach a scientist to fish and they’ll be fed for life.”
something along those lines?
Have you considered some form of ‘crowd sourcing’, if that’s the right term…or a competition for graphic design students, with the prize being a free copy of STAN (that’ll get them going!)
David:
That’s what this post is, no?
A Google Image search for
Borges forking paths
brings up these pictures that might give you some inspiration;
https://www.google.com/search?q=borges+%22forking+paths%22&num=30&safe=off&rlz=1C5CHFA_enUS503US503&espv=2&source=lnms&tbm=isch&sa=X&ei=vMq1VMr9HsmcgwSMj4SIAw&ved=0CAgQ_AUoAQ&biw=1895&bih=959
The apparent credibility of a highly cited publication is not truly more credible than the assurance that the statistics were carried out to the usual academic standards, which can easily advocate the most incredible.
https://www.google.ca/search?q=borges+%22forking+paths%22&num=30&safe=off&rlz=1C5CHFA_enUS503US503&espv=2&source=lnms&tbm=isch&sa=X&ei=vMq1VMr9HsmcgwSMj4SIAw&ved=0CAgQ_AUoAQ&biw=1895&bih=959&gws_rd=cr#imgdii=_&imgrc=c8cAlAkJsJXLzM%253A%3BN7P1a8U-k_-VGM%3Bhttp%253A%252F%252Fwww.notable-quotes.com%252Fb%252Fjorge_luis_borges_quote_2.jpg%3Bhttp%253A%252F%252Fwww.notable-quotes.com%252Fb%252Fborges_jorge_luis.html%3B400%3B280
I think titles for t-shirts 1 and 2 should simply be “Gone Fishing” and “Garden of Forking Paths”.
Pingback: Some art so far - Statistical Modeling, Causal Inference, and Social Science Statistical Modeling, Causal Inference, and Social Science
Pingback: Art Featured on Andrew Gelman's Blog - Kant or Wont