Link roundup for April 2020

We have a poster webinar from Shiz Aoki at BioRender:


There were over 1,000 people attending this, I think I heard her say. The five topics covered within:

1. Consistent margins / padding
2. Arrange “sections” in a grid (Chicago, not Paris)
3. Pick one section to highlight
4. Colour, background, gradients,
5. Poster-specific figure formatting

• • • • •

With the COVID-19 crisis, almost no conferences this summer are planning on going forward. A few organizations are exploring some sort of scaled back online conference, which is one of seven recommendations in this article about softening the blow from conference cancellations.

The article points to a couple of examples of academic societies doing interesting things in the online poster space.

The Cognitive Neuroscience Society will have a virtual poster session.

More than a thousand posters (and growing) will be presented in individual, customizable virtual booths that include optional video presentations by the authors. New poster submissions will be accepted through April 14, 2020.

The Society of Toxicology has browsable archive of conference posters that normally would have been presented at their cancelled March meeting.

The Federation of European Neuroscience is also meeting virtually, and they say their format will include posters.

The Genetics Society of America says it is doing their virual poster session by uploading PDFs to Figshare, allowing an option video walkthrough, and an option Q&A session on Zoom.

If anyone has examples of online poster sessions in the next months, please email the link to BetterPosters@gmail.com!

• • • • •

The COVID-19 pandemic has shown how challenging it is to convey uncertainty. We humans crave it. We want to know, “When it this going to end? How many people might die? Is my community likely to be hard hit?”

This great image compiles ways that we can try to show uncertainty visually:

This is taken from a book chapter that is just about ready to be published. The link contains an almost complete version of the chapter.

Ironically, the first way to show uncertainty in the illustration – error bars – is often bungled. When I view scientific scientitic documents, I look for error bars and very frequently find that they are uninterpretable. The authors don’t say what they represent.

Similarly, the predicted path of hurricanes are often shown with ensemble plots. People tended to not understand those, either. They saw the path for the central path of the eye of the storm, and didn’t realize that most of the damage is far from the eye.

It’s indicative of how badly we show uncertainty in scientific graphics.

• • • • • 

An interview with Elissa Schloesser of My Visual Voice, who is a graphic designer specializing in information design.

• • • • •

Alissa Eckert and Dan Higgings are interviewed in the New York Times about creating the CDC illustration of the SARS-CoV-2 coronavirus.

But for the coronavirus illustration, they went with what professional medical artists call a “beauty shot”: a detailed, solo close-up.

“We just call attention to the one virus,” she said.

This one gets into a little more nitty-gritty detail about the software (Autodesk 3ds Max) and choices made in the process than the article that was in last month’s roundup.

• • • • •

Many people retweeted (may be started in this Reddit thread?) this amazingly bad graph from a television station. The source seems to be Fox31 KDVR in Colorado, according to Manny Equerra.

Look closely at the Y axis.

Reddit user ke1v3y helpfully created a correct graph.

Intuitively, the two graphs show the same trend, so it’s not at all clear why the original is so completely bizarre.

• • • • •

Another great document from MyFonts on using colour with typography. I particularly like point #5 (my emphasis):

Regardless of its brightness, color is paler than black and contrasts less with a light background. As a result, type printed in color will not have the power of simple black. You can compensate for this paleness by increasing the amount of color used. Increase the impact of type up by bumping it up by one weight (book to regular, regular to semi-bold, semi-bold to bold etc.) – or set it a size larger.

• • • • •

This is a nice little post using cupcakes and wedding cakes as an example of a creative process.

You get to a complex project by making a lot of small but complete projects, not by completing parts. Hat tip to Justin Kiggins.

• • • • •

Well, these tweets were sitting in my notes for months. Ioana Marinescu asked:

We’d all love to learn & connect but approaching new people can be awkward. What is the best question to start a conversation with at a conference? Other conversation starters?
What are you working on? 39.1%
What’s your research area 22.4%
What’s exciting in econ? 10.8%
Best papers you saw? 27.8%
832 votes

And Kording Lab asked the same question for neuroscience:

We all love to learn & connect but approaching new people can be awkward. What is the best question to start a conversation with at a conference? Other conversation starters?

What are you working on? 55.6%
What's your research area 20.8%
What's exciting in neuro 8.2%
Best papers you saw? 15.5%
466 votes

• • • • •

This plot looks like a tornado but was made like a duck.

The plot was from this article in The Conversation UK. The Twitter account for the website called it “useful,” (now deleted) prompting a lot of people to say, “This is not useful.” My own thoughts are in this Twitter thread.

Non-standard data displays are tricky. A big problem with this graph is that it breaks a very common expectation. “Time” on graphs is almost always shown the X axis. Here, it's at arbitrary points along each coloured line.

The other question is: What does this graph show that a bar or line graph does not? Nothing that I can see. In a standard bar graph, "deaths per day" is shown by height of bar. "Increase or decreases in death" is seen by comparing two bars.

This tornado graph is an example of what Edward Tufte calls a “duck.” Sure, it's different. But what is the point?

• • • • •

Daniel Gonzalez has a late entry to the “social distancing graphic” collection.

Sorry, but it’s not as good as others, since the quantity of animals described in numbers and the quantity shown do not match.

• • • • •

Since comics inspiration is a recurring theme in this blog, let this great blog post introduce you to graphics design legend Ira Schnapp. He carved the entire motto of the postal service in stone in a New York post office when he was 16.

This post is the first in a series on Schnapp. Part 2 gets to the comics work he is known for, including one of the most famous logos of all time, Action Comics #1, the first appearance of Superman. Which Schnapp probably created in a day. Part 3 talks about another famous Schnapp creation, the Superman logo.

Are posters right for online meetings? Thoughts from The Allied Genetic Conference virtual poster session #TAGC20

I have been watching many academic conferences move to online presentations this year with great interest. I want to see how they are going to do it and how well each format works.

In my Twitter feed today are discussions from The Allied Genetic Conference (#TAGC20). They have made their content free for anyone to view. Their poster session guidelines are here.

TAGC is handling their poster sessions by having people upload PDFs of their posters to Figshare, which has created a special collection for posters from this meeting. All the posters are available to view now, and no more can be uploaded. You can search through categories, or by keyword. (Zero hits for “crustacean” – d’oh!)

Here is an example of a poster where the presenter, Katja Kasimatis, recorded a ten minute walkthough to go with the PDF. Abigail Feresten did both a two minute and a ten minute version of her walkthrough (I think the short version was part of a “Poster preview” that is mentioned on the session guidelines but not completely explained.)

Finally, there will be Q&A sessions next week. I’m going to try reaching out to presenters to see how it goes. The FAQ notes there these Q&A sessions are not going to be archived.

Poster Q&A sessions will not be made available beyond the live discussion, so presenters may choose to discuss data in those sessions that they are not comfortable publishing on their poster.

The TAGC FAQ contains some interesting advice about format.

Although any dimensions can be used, remember that attendees will be viewing posters on their computer screens. ...

We recommend arranging information blocks like

1 > 2 > 3
4 > 5 > 6
7 > 8 > 9

instead of the print poster tradition of

1 4 7
2 5 8
3 6 9.

In other words, layout your poster in rows instead of columns. This makes sense, as the idea is that you are going to scroll down, and you don’t want people to have to scroll back to the top of the page all the time.

But if you’re going to do that... why not just a single, continuous column?

The TAGC20 advice on layout, probably without them realizing it, strikes at the heart of something I have been thinking about a lot with online poster sessions.

What, specifically, makes an online poster presentation a “poster”?

In this format, you are first uploading a document to Figshare. Why make that document mimic a single large piece of paper? Why not just upload a manuscript?

In this format, you have the option of uploading a video walkthrough. Why show section after section of a single large page? Why not just upload a slide talk?

The conference poster has always been a strange hybrid format that was born out of necessity more than anything else. If conferences go ever more online (an existing trend that the COVID-19  pandemic is only accelerating), we should not continue to have posters just because, “We’ve done it that way for decades.” Maybe there is no need for posters to continue in virtual conferences.

External links

TAGC Virtual poster session guidelines
TAGC Figshare collection

Critique: Anorexic mice

This poster was done as practice by a group of students, not for a presentation at a conference. Click to enlarge!

You could put this up in a conference and it would probably be one of the nicer looking posters there.

The layout is quite clean. There is lots of white space, and the text is readable.

The mouse graphic is nice, but would be better placed if the head was facing into the center of the poster. We look where faces look, and this makes you want to look off the page. Possible solution: flip mouse, left justify title, move logo to right side instead of bookending title.

There's no standard way of presenting a boxplot, so you need to specify somewhere what the components mean. (Is the center line mean or median? Are the whiskers quartiles or confidence intervals or something else?)

The boxplots are a little hard to read from a distance. Maybe thicken the lines a little or fill the box with a light colour?

In the middle column, the first line of first paragraph seems much closer to graph than anything else. If you have guidelines to make the margins between columns even, the particular layout with ragged right is defeating the purpose. Putting a soft return before "was" might solve it.

The callout box with the summary is good, but I had to dig down to the middle of sentence before I understand why the authors made all the measurements. Maybe start with, "Anorexia is a human eating disorder that we studied in mice by measuring mitochondrial activity, neuropeptide levels and hypothalamic glucose uptake" and continue from there.

“But the title says anorexia! Why are you complaining that you don't know what the summary is about until you're halfway through?” Because the callout has about as much, if not more, visual weight than the title. The callout is in the upper left where people look first, and it’s highlighted it with attention getting red. The title is downplayed by the choice of such a thin font.

The right edge of the callout box doesn’t align with the edges of the flowchart or text below it.

The position of the mouse brain drawing is not at the point of need. The hypothalamus is mentioned in the first normal paragraph. When I read that first sentence, that's when I probably want to know, “Where is the hypothalamus?”

Probably a lot of the paragraph text “Differences in hormonal activity” could be changed into a statistic summary, inserted into the corner of the box plots.

Critique: The world belongs to squirrels now

I’m always fascinated to see show presentations of a project change. Today we have two posters from Sarah Westrick. Click to enlarge! This is the first one, chronologically, from the 2019 Animal Behavior Society meeting.

This one was from the 2020 Society for Integrative and Comparative Biology meeting.

Sarah wrote:

Lately, I’ve been trying poster designs that are different from my usual grid-like structure. For the Animal Behavior Society, I was able to focus specifically on one study which made it a lot easier to fit everything in that I wanted to. For SICB, I was advised to include more of the work we’ve done, so it was a challenge to fit the various studies together and still be concise.

The lack of grid structure generally works on both posters, and it works best when it’s clear that it’s a deliberate choice. The bottom text box in the Animal Behavior Society (“Offspring from...”) is a good example, because the text box is solidly far past the edge of the underlying box for the graph.

The text box above it (“Highly attentive moms...”) is less successful because the overlap is so much smaller. It’s enough to convince you it was deliberately instead of incidentally placed there, but just barely. I might have expanded the white box the graphs are sitting in to the left, so that the text box has more overlap, imitating the box below.

The “hero” shot of the squirrel by Erin Siracusa in both posters adds visual appeal. In fact, it’s because of these that I first saw Sarah’s work.

The drawback with edge to edge photographs is that it can be difficult to keep the text readable.

In the first poster for Animal Behavior Society, the white text on a dark background used for the title and the two main boxes fares much better than the dark text. In one way, that is good, because the two main conclusions “pop” compared to the rest. It highlights the things you want people to remember!

The second poster for Society for Integrative and Comparative Biology, the results on the right are not quite as successful as standing out as in the previous poster. It’s partly a matter of text size. The result summaries are physically smaller on the page.

But the lack of text “pop” is not helped by the complexity of the box, either.

The text boxes used for the main results have four overlapping layers:

1. The text in foreground.
2. The semi-transparent text box under the text.
3. The large number “1” under the text box.
4. The photo in the background.

Particularly around the word “predicted” in the excerpt above, your eye has to make sense of all four of those layers right next to each other. While I like the “wayfinding” provided by the big numerals, there might be some other ways of accomplishing that. Even moving the numbers out from the text boxes might help.

I’m not sure that centering all the text has any particular advantage. I would suggest left aligning most of it.

Pandemic publishing plans: Better Posters book update

When I started working the Better Posters book over two years ago, the original plan from the publishers was to release it in the first quarter of 2020. Which would have been the end of March, 2020.

Er. Yes. About that.

I never mentioned the “first quarter of 2020” date publicly, but because we are now past that date, this is a good time to talk about the status of the book.

The writing is done. The copy editing and revision is done. The cover design has gone through a few versions, and is probably close to being finalized. Planning is going on for production and promotion. But we’re not going to be close to that original planned publication date.

Part of the reason we slid past the planned publication date was my own fault. Copy editing took longer than I anticipated. I always knew it would be complex, but things that I didn’t think of cropped up, like getting permissions for quotes. And I was still teaching full course loads when the copy editing was going on. And it was partly because the copy editing wasn’t just text, either. I ended up revising a lot of figures, and even creating a few new ones.

But the other big reason it will still be some time before you can hold the Better Posters book in your hand is not my fault. There are more important things going on than my little book.

So you may have noticed we’re in the middle of a global pandemic.

Normally, if the book was coming out in the next few months, it would be leading into, or in the middle of, summer. Summer is peak conference season, but every scientific conference that I know of has been cancelled for 2020. It is the worst time ever to release a book about conference presentations.

Of course, that problem even assumes that the book could be physically printed over the next couple of months. Much of book production is digital now and can be done remotely, but there’s enough stuff involving moving atoms around instead of bits that printing and shipping a book would be somewhere between “slow and difficult” to “Shut up and get out, and take your impossible demands with you.”

We’d be doing nobody any favours trying to get this book out in the next few weeks. There’s no point. (Thanks, COVID-19!)

I am still hoping that you will be able to read this book before the end of the year, but only time will tell. It always does.

Update, 4 April 2020: Yesterday, Beth Meachum, an executive editor at Tor Books, took to Facebook to outline why publishing is in a tough spot now. Some of this is specific to the US, and the Better Posters publisher, Pelagic Publishing, is in the UK. Still.

I want to talk for a minute about why publishing is in so much trouble right now. It’s way more complicated than most people seem to think.

First, you need to know that the vast majority of our business remains in hardcover and paperback books. Hard copies, physical objects. The second strongest sector has been audio books. Ebooks are a distant third.

Selling books is a very long and complicated supply chain. Ignore editorial – writers and editors can work at a distance and electronically. It really starts with the paper. Storing paper for the big presses takes an enormous amount of warehouse space, which costs money. Printers don't store a lot – they rely on a “just in time” supply chain so that when a book is scheduled to go to press, the paper is delivered to the printer. Most of that paper is manufactured in China. Guess what isn’t coming from China? Anything, for the last three months. Some of it comes from Canada. Guess what the Trump administration put a big tariff on at the beginning of the year?

So, we don’t have adequate paper supplies. Then consider, big printing plants are not “essential businesses.” There are only a couple printers in the US that can handle the book manufacturing business. One of them shut down last week. COVID-19. We started rescheduling books like mad to deal with that.

But supposing we had paper, and a printer and bindery, the books have to be shipped to the warehouse. Again, non-essential movement. The freight drivers moving books? Staying home, as they should. Not all of them. I hope they remain healthy, because dying to get the latest bestseller to the warehouse doesn’t seem quite right to me.

Now then, our warehouse. We have a gigantic facility in Virginia. Lots of people are working there, bless them, but it’s putting them at risk. There they are, filling orders, packing boxes, running invoices. Giving those boxes to the freight drivers who take the books to the bookstores and distributors. Again, truck drivers risking their lives to bring books to the bookstores.

But think again. The bookstores are closed. The distributors are closed. No place open to deliver the books to. Some bookstores are doing mail order business, bless them, but they aren't ordering very many books from our warehouse. Amazon isn't ordering very many, either – because they have (correctly) stopped shipping books and are using their reduced staff to ship medical supplies and food.

So the books that distributors and sellers ordered months ago are not being printed or shipped or sold. And because of that, they aren't making any money. And because of that, they are not ordering any books for months from now. Plus they aren't paying for the books they got from us last month and the month before. Cash flow has ground to a halt.

Now, audio books... turns out that people mostly, almost 100%, listen to audio books while they commute to work. Sales of audio books collapsed about three weeks ago. Fortunately, there isn’t a physical supply chain there, so theoretically that business can restart immediately upon resumption of commuting.

So given all the above, it’s not a good time in the publishing industry. The damage is going to last for a long time, the effects will be felt for at least a year to come, even if we do go back to business as usual in May. Or June. Or July.....

Oh let’s be real. We won’t go back to business as usual until there is a real vaccine for this coronavirus.

The insight is welcome, the honesty is bracing, and it’s just a little scary for a first time author to read. But I’m lucky. Very, very lucky. My livelihood doesn’t rest on this book.

P.S.—The artwork at the top is a snippet of a cover concept! Just a teaser to show production is happening, and to give you a little more than authorial angst over publication delays.

P.P.S—Because the book is still in production and should still be released, I thought this was a good time to give the blog header a new look. You may notice it is a little more in line with the cover concept sample above than the old blog header.

This was my first project in the newest version of CorelDraw. I was excited to play with some of the new features, like variable fonts. Looking for references for the “Soon to be a major motion picture” sticker was also fun.