31 May 2012

Finding photos


Walking home last night, I spotted this big beautiful moth near our library:


I took out my smart phone, got a picture, did some simple editing (cropping and rotating), and shared it online within a few minutes.

I mention this because it reminded me of a saying: “The best camera is one you always have with you.” Many of us usually have a camera at hand. And not a camera that takes low-resolution, grainy images, but something that gives you pretty decent result with a little effort.

Yet despite the ease of getting high-quality digital images, when people are looking for images to put on their poster or presentation, the first stop for many people is Google Images. That’s okay, as long as it’s not the only stop. For instance, I was helping on student with a presentation, and complained about the low-resolution image he had of a piece of equipment. He replied, “That’s the best picture I could find in Google Images.”

“Wait. Stop,” I said. “This is a piece of equipment that you yourself are using to do your own research, right? It’s sitting in your lab?”

“Yes.”

“Do you have a camera?”

“Yes.”

“Then take your own photo!” I said, miming a cuff upside his head.

The best photos will often be the ones you take yourself.

Failing that, there are also stock photo companies, which can be cheap and useful. Unfortunately, stock photos can also be cliched and banal. I’ve just resigned myself to seeing “science” represented by people wearing white lab coats and safety goggles, surrounded by flasks full of multi-coloured liquids, pipetting something. (For an antidote, check out This is What a Scientist Looks Like!)

I am a big fan of Flickr. First, they are more likely to have higher-resolution images. More importantly, if you go to their advanced search page, you can scroll down and select “.”

The bottom line: The fastest way to make your poster more attractive is to add photographs. Photographs are real, recognizable things that will almost always provide a more immediate entry point for your readers.

24 May 2012

Link roundup, May 2012

Colin Purrington has a template for conference posters. Regular readers will see a lot of things that I suggest here. The major difference is that he spreads the “fine print” all along the bottom, whereas I tend to favour the bottom right corner.

I normally don’t normally like animated GIFs, but in these case, I’ll make exceptions: “What I feel like when my poster is the best one in the row.” And “When my poster wins an award."

In praise of margins.

Don’t be afraid to do something that you’ve seen elsewhere:

So much poor design has been made in the pursuit of “originality,” “creativity,” and “grabbing the viewer” (don’t try this in person).

For instance, what if someone felt the need to make every stop sign unique?

Think you know serifs from sans serifs? Prove it! It’s harder than you think when you’re under the gun to use the gun... I Shot the Serif is a cute little online game (even though the pun is past its use by date).

Stripped Science looks at ancient poster sessions.

DrugMonkey asks what kind of poster do you want to see at a poster session? He prefers them a little rough around the edges.

An infographic generator. Use it wisely. In particular, see this critique from Flowing Data (my emphasis):

I'm having trouble getting on board with these tools. Easel.ly, for example, provides themes, such as the one on the right. There's a guy in the middle with graphs around him and pointers coming out of his body. You get to edit however you want.

So in this case, you start with a complete visual and then work your way backwards to the data, which I'm not sure how you can edit other than manually changing the size of the graphs. (Working with the interface takes some patience at this stage in the application's life.) It's rare that good graphics are produced when you go this direction.

There is a new typeface coming that is intended to make the area of a number (from 1 to 999) in proportion to its value. The planned name? FatFont. (May require registration / subscription to read).

A big ol’ Prezi on poster presentations.

If you’re an organizer of a small conference, you might want to have a look at this post on speakers behaving badly. Sigh. People suck.

Sibling blog to this one, NeuroDojo, makes a cameo appearance at Context and Variation, which has some great insight into poster designs.

I asked students to share what they liked and didn’t like. Why was one poster successful where another wasn’t? In a few cases disciplinary biases impacted the posters they liked, but most of the time their preference was driven by design features. The posters that were striking, confident and accessible were the clear favorites.

17 May 2012

Critique: Personality types

Today’s poster is shown with the permission of David Caldwell. You may click to see David’s handiwork in larger size.

I like the use of emphasis in the title. I like it a lot. I’ve not seen that done before, and I think it’s excellent way to make it easier for people who are scanning for posters while walking to find the poster.

The colours are well chosen. They are consistent and attractive. I do wonder if some of the greens (e.g., in the graph at the top of right column) are distinct enough. If printing was poor, or light was low, it might be hard to tell those apart.



I'm not sure I get what the central column is showing me. I get that they are emails, and that the orange circle is telling me something about the different groups of students. But I don’t see how these six differ. They might need a little bit of explanation in a caption somwhere.

The text hierarchies could be strengthened more. “Objectives” and “Methods” should be set in bigger text than the title of the “Summary characteristics...” table below them. Likewise, “Conclusions” is getting lost because it’s not noticeably bigger than anything else around it.

There’s a mixture of alignments in the right column; some is left justified, some is centered.

The Y axis of the bar graph is unlabelled, and the graph has no error bars. Also, I’m not a fan of putting numbers on a bar graph; error bars would convey much more useful information. If the exact numbers are that important (which, on a poster, they almost never are), they should be in a table, rather than as redundant information in a bar graph.

I am not sold on putting the labels on alternating sides in the “Survey results,” either. This is a table, so one should follow the conventions for a table. Alternately, this could be a graph, or set of graphs, with means and error bars. It might be easier and faster to read.

Based on some of these comments, David did this revision:


The original might have a slight edge in being inviting to someone walking past, by virtue of that big central column with pictures. This version will reward anyone who stops and reads it with a clearer story.

15 May 2012

Tip jar

If you find this website useful, please visit my #SciFund project and consider making a contribution! I'm halfway through the fundraising campaign, and given the goal, your support makes a big difference.

And if you can't contribute (though I would love it if you did!), please help me out by spreading the word about the project! Hit the Facebook "Like" button! Tweet it! Share the video!

There are no ads on this blog, and I plan to keep it that way. But I'm not too proud to beg once in a while. ;)

The sand crabs I mention in the video have made little cameo appearances on the blog here and here.

Photo by Dave Dugdale (Learning DSLR Video) on Flickr; used under a Creative Commons license.

10 May 2012

For emotional impact, go big

A new paper on type size supports what headline writes have apparently known for ages.


Bigger text gets bigger reactions.

ResearchBlogging.orgThe work by Bayer and colleagues used electroencephalograph (EEG) recordings to study the response of people as they showed them words that tended to be associated with good emotions, bad ones, or that were emotionally neutral. They found that the bigger the text, the faster someone responded emotionally to the word, and the longer they responded.

At least, that’s the authors’ interpretation. One thing about this study is that, as far as I can see, they didn’t ask their subjects directly, “So, how are you feeling?”

If you have any words on your poster that you want to emphasize, and for people to respond to: make them big.

Reference

Bayer M, Sommer W, Schacht A. 2012. Font size matters—emotion and attention in cortical responses to written words. PLoS ONE 7(5): e36042. DOI: 10.1371/journal.pone.0036042

Photo from here.

03 May 2012

Critique: Bird songs and brains

Today’s poster comes from Justin Kiggins (a.k.a. Neuromusic), and is used here with his permission. You want to see it larger and more clearly? Then click, I say! Click!


This are many good design decisions here. There are clear columns. The colours have been chosen and deployed with deliberation.

But this poster is struggling to handle an enormous amount of information. The “Too long, didn’t read” section on the right is almost the authors’ way of admitting with a sigh, “Yeah. We know. It’s a lot of stuff.”


Given this, the first solution is to edit if possible. It can be hard, but sometimes, not every data point needs to be shown on the poster.

If, however, you want to leave all of that information on there, what else could be done to try to cure the intimidation factor?

First, I would try to create more white space. The margins between the three main columns are not much bigger than the margins between the two or three sections within them. This means that the columns tend to blur together, and the whole thing looks less structured and organized as a result.

I stuck this into my image editor, shrunk each of the three main columns by 5%, and spaced them out more. You can more readily see the three columns, particularly in the small version (but you can click to enlarge).


You can get away with this without losing visibility because this poster is huge: 88 inches across. The text and images are likewise big, which is great.

While the main columns only needed a slight tweaking of the margins, the arrangement within each column is much less disciplined. This is important, because the columns are wide, and are meant to be read in rows within the main columns. This could be okay, because the reading pattern is consistent.


In the left column, each row has three sections... but none of them are the same width or height. The middle column is a little cleaner, except perhaps at the top. The left column starts lower than the other two, and has an asymmetric subdivision that doesn't line up.

Again, more attention paid to alignment (that the second line of the title doesn’t line up with the institutional affiliation bugs me) and more clearly defined margins between each section would help. Here’s a version where I’ve laid down a partial grid. Ideally, each red line should not touch a graph or text underneath it. The middle column comes out almost unscathed, but the ones on either side don’t fare as well.


Finally, there’s the issues of the colours. As far as I can tell, they were chosen deliberately. I think the orange is representing physiological data; purple consistent is linked to “context,” and blue to “target,” etc. But there’s a lot of colours there. Even though they are well chosen and fairly subdued, they end up fighting with each other a bit. This might not be such a problem if the material were given more generous margins. Plus, the link between the colour and the concept is not intuitive; it would have to be explained. Rather than five separate colours, different shades of one or two colours might have accomplished the same thing.

The more information is on a poster, the more critical the space between that information becomes.