Link round-up for December, 2018

It’s a small link round-up for this holiday season, but I have one I want to share, particularly given that last week’s post was musing about how we train students in graphic design.

This article talks about teaching data visualization to kids. Fourth grade students, in fact. That’s what makes it a perfect holiday post, because kids love holidays. Or something.

We might be taught how to read line, bar, and pie charts in elementary school because they have been around longer than others and are used the most. ... I’m not foolish enough to think I could teach 30 kids an array of new graphs in one afternoon, but I could at least help them understand that there’s more to the world than line, bar, and pie charts.

The post also talks about a Match-It game for data visualization that looks interesting.

And for a lighter touch, here’s a graphic artist’s breakdown of all the Marvel movie posters.

That wraps up the year for the blog. Next year will be big for this blog. I have some very cool things in the work in the coming months!

Should we train all students in graphic design?

This blog exists to help solve a problem: that academic conference posters are ugly. But I am under no illusions that this blog is going to fix the problem. So, what would move the dial the quality of conference posters?

When faced with this kind of question, I often see people say, “We should include this in our training for students!”

As an educator, I never want to be the person to say we shouldn’t train students. I’ve done it myself, often. I support this sentiment, but I’m wary of calls for “more training,” for two reasons.

First, suggestions for “more training” make me worry about mission creep. Over time, I’ve heard that students need more training in statistics. And in writing. And in ethics. And in grantsmanship. And in social media. And diversity issues. And in dealing with media. And in public outreach. And so on.

Don’t get me wrong, these are all worthy topics where I think training would be beneficial. But there is only so much we can realistically expect to make students competent in during their time in our programs. It’s hard enough to attain competence and eventual mastery in one discipline.

Second, graphic design is a professional skill that takes years of study and practice. “Training” research students in graphic design would probably end up being a few credit hours over a multi-year program, taught by a non-professional (e.g., a research scientist in the department who is smart enough not to use Comic Sans but never took a class in design) rather than a skilled graphic designer (e.g., someone from over in an art department who does this on a daily basis). (And I say that as someone who has been asked to do those kinds of classes and workshops. I mean, I’m that guy!) I worry that calling for training could trivialize the skills needed for excellent design and become a curricular box-checking exercise.

Instead of expecting academics to become one person bands, we should try to create more access and respect for experts in other fields and be willing to use them, credit them, and pay them.

Picture from randychiu on Flickr; used under a Creative Commons license.

Lessons from battle scenes: It’s all about the build-up

The Lord of the Rings: The Two Towers features a fantastic battle scene: the battle for Helm’s Deep. There’s a lot of reasons why it works. The scale and physicality of it is awesome. But there is one factor that is underrated.

In the DVD commentary, director Peter Jackson talked about watching famous sieges and battles of film, like Zulu. He said that what he learned from those was that the secret was all in the build-up.

The movie spends a lot of time making sure that the audience understands the situation. The fortress is literally set against a mountain, creating the impression of “backs against the wall.” The city is outnumbered, with hundreds of defenders against thousands of attackers. 

And you see that army of thousands marching in. Not doing anything at first. Just standing there. Then thumping their spears and yelling.

More waiting.

And then the army charges and all hell breaks loose and it is on. And when those orc charge and start throwing ladders up against the fortress walls, you are invested and ready to see what happens next.

The battle itself is only about ten minutes of film. It would have been easy to not show the orcs doing their “haka” and charge into the action, particularly in a movie that asks you to stay in your seat as long as the Lord of the Rings films do.

On a poster, the Results section is like the action part of the battle, when arrows are flying, axes are wielded, and explosions are going off. Too many poster makers want to charge into the action. Many poster presenters want to get to the stuff they find exciting as soon as possible.
 
But an audience member has to care. They need the set up. They need to know what the stakes are. They need to know the landscape the project is set in.

That means you should spend a lot of time thinking about the first part of your poster. This is important both in the design of your poster, and how you are going to talk about your poster. How are you going to bring someone who has never thought about your subject on board and interested in the outcome?

By telling them what the conflict is. By telling them the complication you are trying to solve. By setting up a scenario, then throwing in a “But...” moment. You have to be crystal clear in that introduction. If you can do that, people will follow you through the battle to see the resolution.

External links

15 years later, no one’s matched LOTR’s Battle at Helm’s Deep

Critique and makeover: Buffer it out

Today's poster is a contribution from William Elaban. This was not for a conference, but a class. Click to enlarge!

Now, I have to apologize to William here, because my first reaction to this poster is not a kind one. But sometimes, my first reaction to a poster is:

“Blow it up. Blow it all up. Blow it all up and start again.”

This poster has deep structural issues. There is too much text. The reading order is all over the place. When the problems are that big, you want to see a fresh page.

But first impressions can lie. Then I calm down and start tinkering. And by following some of the usual design principles, the poster slowly but steadily gets better.

The first thing I did was get rid of lines. Underlined text and boxes were immediately banished. Headline case was replaced with sentence case.

Next, I tackled the table. I gave it a more standard format, with just horizontal lines separating the top, header, and bottom. I cut the large number of decimal places down to a more reasonable three.

Getting rid of the long numbers in the table made it more compact. I started pushing the elements around so I could line up the left edge of the table with the text blocks above. I did the same with the figure on the right. Columns started to take shape.

All the headings were made bold.

The text was a mix of Calibri and Arial, so I made it all Arial. I continued to try to make the text the same point size whenever possible.

I justified the text blocks to emphasize that things are aligned on the page now instead of scattered higgildy piggildy.

The deep problems remain – to get rid of those you really do have to blow it up and start again. But I’ll be darned if the poster doesn’t look noticeably better. And there isn’t anything complicated about what I did here. It mostly boiled down to:

1. Get rid of lines and boxes.
2. Line things up.
3. Put space between things.
4. Make the text consistent.

Sometimes, I’m kind of amazed by how much the appearance of a poster can improve with simple fixes. It’s not crazy complex stuff. It’s like how getting a good haircut and a little makeup can take years off someone’s apparent age.

Link roundup for November, 2018

This month’s contender for “Best conference poster” was spotted by Greg Fell:

Clever!

• • • • • 

If you have a fabric poster, Crystal Lantz can show you how to turn that ol’ science communication into a lovely tote bag!

She’s got detailed instructions, but you’re on your own for the sewing machine. Hat tip to Crystal Lantz and Caitlin verder Weele.

• • • • • 

I missed these tweets from Suzy Styles about poster club back in August, when there was discussion about harassment in poster sessions:

The First Rule of Poster Club is...
🤛🏻you 🤛🏻do 🤛🏻not 🤛🏻 talk🤛🏻about...
🤛🏻the presenter’s appearance
🤛🏻the presenter’s phone number
🤛🏻who you think ‘actually’ wrote the code
🤛🏻basically anything other than the poster and relevant scientific context 🤷🏻‍♀️

The Second Rule of Poster Club is...
🤛🏻you 🤛🏻do🤛🏻not
🤛🏻try to look down the presenter’s top
🤛🏻stand unnecessarily close
🤛🏻touch the presenter 🙅🏻
🤛🏻block the presenter with your body
🤛🏻talk about anything other than the poster and relevant scientific context

The Third Rule of Poster Club is...
🤛🏻be aware that if you are much taller, standing close can be intimidating
🤛🏻be aware that if you are more senior, standing close can be intimidating
🤛🏻check - am I being a jerk atm?
🤛🏻do not talk about anything other than the poster++

• • • • • 

Phylopic is a great resource for biology presenters. It provides silhouettes of different animals. I tried, “Crab.”

Weirdly, I tried clicking Callinectes sapidus, which has an icon next to its name, and found there was no image of that species! But with a little more clicking, it provided this outline of Liocarcinus vernalis.

The “illustrated lineage” feature is also pretty nice. Hat tip to John Vanek.

• • • • • 

More conferences should use magnetic name badges, says Jennifer Rohn.

OMG - a magnetic name badge so you don’t have to pierce your expensive clothes! Where have you been all my life?

• • • • • 

This blog post mainly about bootstraps (in the mathematical sense) but also contains warnings against the problems of bar graphs and error bars.

(T)here is no substitute for a scatterplot, at least for relatively small sample sizes. Also, using the mean +/- SD, +/- SEM, with a classic confidence interval (using t formula) or with a percentile bootstrap confidence interval can provide very different impressions about the spread in the data (although it is not their primary objective).

• • • • • 

John Burn-Murdoch has a nice analysis of this wide-distributed graphic of US election results.

With that constraint gone, you can label them all directly! Immediately more readable.

• • • • • 

Kathleen Morris gives us a list of free graphics resources.

Hat tip to Lisa Lundgren and Emily Rollinson

Critique: Stimulating brain to lower pain

Today’s poster comes from kindly contributor Emma Taylor, who was generous enough to share her honours project. Click to enlarge!

Quick thoughts.

• Good structure that lets me know I read in rows.
• Visual hierarchy is generally good.
• Looks crowded. Would benefit from more space around everything, but particularly between text and edged of blue boxes they are in.
• Underlining text in discussion would probably be better as bold or italic text.
• Another editing run to cut more text might help. (I know, that bit is hard.)

Critique: Digitized manuscripts

This blog mostly uses sciences as examples, so I am always positively delighted when I get contributions from the humanities. Today’s contribution is from Cornelius van Lit. Click to enlarge!

One of the things like about getting other people’s posters is they try stuff I would never do. I’d never put my title in the middle of the poster. And yet, it works here.

The poster is a great example use of using size to indicate reading priority. That large text in the middle makes it very clear where you are supposed to start reading. Nothing competes with that title.

The downside of having the title in the middle is that there is some potential confusion about how you are supposed to read the remaining text. But it’s okay here. After reading the middle introduction, people will jump up to the upper left corner (which starts “Scholars use digitized manuscripts...”) because that’s just where you look first when you read English.

After reading that section, I think most people will read across to the top right (which starts, “In one chapter...”), because of the proximity of the text. Having that big title in the middle stops you from looking down and trying to read in columns. If the title and introduction were at the top, people would get lost. (But with only four sections, they wouldn’t get lost long.)

I tried making two changes, both subtle, in the revision below.

First, I moved the author information and the QR code from the top of the poster to the bottom. I really didn’t like how the QR code was sitting “corner to corner” in the first version, so I lined it up with the map below. Besides, both bits of material looked like “fine print,” and fine print is more logically placed at the bottom. It might also be easier for shorter people to scan the QR code if it’s lower rather than higher on the poster.

Second, I added a very subtle neatline around the map in the lower left corner. (You may have to enlarge to see it.) Three sides of the map have segments with clear straight edges, but the left side doesn’t, making the map a strange, irregular shape. By using a thin, light gray line, the shape of the map becomes more consistent with the shapes in the other three corners.

External links

Among digitized manuscripts