Something about this looks familiar. Today�s poster comes from Igor Miklou�ic. Click to enlarge!
I love this idea. I�ve talked before about how it can be so helpful to base a poster off an existing design. Make a poster about Facebook look like Facebook. Brilliant. It immediately helps viewers recognize what they�re in for.
The poster runs into problems because it doesn�t follow the Facebook format closely enough! Facebook posts are usually short, and accompanied by a picture. Instead, we get some sizable blocks of text with no pictures, and they look gray and uninviting at a distance:
This is a limitation of copying another design. The design of a poster would benefit from changing the text size. But following the design of Facebook means you can�t, because then it won�t look like Facebook, which is, after all, the point.
This might be fixed by a substantial restructure of the middle of the poster to break the big posts into several small ones, perhaps with a few graphics. This would not be a simple change, but might be worthwhile.
Today�s poster comes from Arunas Radzvilavicius, and is shown with his kind permission. Click to enlarge!
The layout, the colour, the generous space, the use of graphic touches are all things to like on this poster. It�s very nice. But sometimes, a poster�s own worst critic is its designer. Arunas wrote:
The optimal amount of text on the poster is something I still can't seem to get right. I always seem to reduce the amount of text to the possible minimum, but that often leads to the poster becoming unintelligible to people not familiar with the details of my research.
How much to write on a poster is always a challenge, although most academics have the opposite problem of Arunas and leave in far, far too much.
The low amount of text is inviting to a reader from a distance, but perhaps confusing when you get up close. Here�s the start:
Isogamy: mitochondria inherited from only one (UPI) or both (BPI) mating types. Ancestral metazoan state. BPI if mutation rate was low.
This is so condensed, it�s close to shorthand. I struggle to revise this into full sentences, because some of the logical connections between words have been erased by the editing. I think this might be close to true:
In isogamy, mitochondria are inherited from one (uniparental isogamy, or UPI) or both (uniparental isogamy, or BPI) mating types. Isogamy is the ancestral metazoan state, with BPI favoured if the mutation rate was low.
Full sentences add more clarity than they take up space.
Seeing this poster shrunk down, it might benefit from the headings being a little more prominent. The poster is a little dark overall, and the reduced contrast dos not help the headings to �pop.� Likewise, using all capitals for the headings make them a little harder to read from a distance.
On to the poster that Vasco sent to me and let me share it with you. Click to enlarge!
My first reaction was that there�s a lot going on in this poster. It was a little overwhelming and intimidating.
The layout of the poster isn�t to blame for the feeling of busyness. The structure of the poster is actually reasonably clear and easy to follow.
A lot of the feeling of busyness has to do with the colours. Looking at it felt a like looking at a busy city�s business district:
There are five big blocks of colour on this poster: a red box, a green box, a yellow note, and orange note, and a light blue sidebar. And there is the data at the bottom, which also uses bright primary colours.
There may not be much that can be done about the data at the bottom, but the other five blocks might benefit from being more similar. Here is a quick and dirty example:
This redesign points out that the logos are also contributing to the business. Three of the five are dark blue, which isn�t in line with the rest of the poster. The dark blue blocks are also competing with the title for attention: the position says �the logos are important� (Cosmo principle), when the title should be most important.
Again, a quick revision that tries to bring the title out by repositioning and shrinking the logos (the title size is the same):
Now the emphasis is clearly on the title. Shrinking the logos helped emphasize the title by creating more white space to separate the title from everything else. The overall effect is a little calmer and more approachable.
Let�s revert back to the original colour scheme for a moment and have another look at that.
Over on the left hand side, the brightly coloured boxes again create a problem of emphasis. The highlighted colours and boxes, particularly from a distance, say, �I�m important, read me first!� The text supports this, too: �The problem� and �The solution� are in bold, and meant as key summaries.
If all the graphic and text cues say, �read me first,� why not put them first?
Some of the things I like about this poster? This poster has uneven sections, but there are visual signals that make it easy to follow. The lines between the columns is better done than on many posters, providing a clear guide that isn�t overwhelming. The use of subtle �A,� �B,� �C� icons help make the order clear and add a nice graphic touch. The sidebar clearly signals stuff which is nonessential to the main presentation of the poster. The spot for stickies is also a nice invitation for interaction.
The collection has three articles so far, each of which examines a different specific graph and how it could be improved. The one above takes on bar charts versus box plots and rescaling the Y axis.
This title for the upcoming Society for Neuroscience meeting is sure to ignite debate about whether posters should have funny titles:
12 things you didn�t know about high responder/low responder rats, stress coping, and the dorsal raphe. Number 5 will blow your mind!
Icons can be useful things for all sorts of graphics. There is a big icon library call the Noun Project that might be useful for in poster design. Its splash page boasts that it has 150,000 icons. I gave it a whirl by searching for �crayfish,� because those have been on my mind:
Not bad. You have a choice of downloading PNG or vector based SVG images, and it�s all available under a Creative Commons license. I will forgive them that two of their icons are definitely crabs and not crayfish. Hat tip to Paige Jarreau.
Astronaut Chris Hadfield has an album of songs from space. Fellow musician Jud Haynes (of Wintersleep) talks about the process of designing this cover.
Jud is an academic at heart:
I set out on the first phase of every good design project, �research�.
Hat tip to none other than the man himself, Chris Hadfield.
In my visual language, science is one of the easiest things to illustrate. There are so many nouns involved. The great thing with science, even in something as abstract as arithmetic, is there�s always some sort of image involved in it, and lots of stuff�whether it�s robots or plant material�that�s exciting to draw. It�s funny, sometimes when I do a science piece, I don�t have to draw things accurately, but I do want to value the science and the research.
Journal covers can have some similarities with posters: a big focus on key images and findings. Cell Press discusses how they pick their cover images: for the journal Neuron:
In particular:
(T)he editors have varying opinions within the team about what they prefer in a cover. Some like somewhat abstract images that require the viewer to stop and think about the connection between the visual and the experiment or idea it represents, and others prefer a beautiful scientific image over a metaphorical work of art. ... Images that look like simple reproductions of figures will most likely not be selected. In other words, no scale bars.
Rebranding a university is always tricky. I�m going through this process now. Penn State is doing this and has a new logo:. Compare the left (old) and right (new):
It hasn�t gone down well. I like the new Penn State logo in its overall design. The only problem is the eyes of the lion looks zombie-like, or, as one person said, �hypnotized.� But there isn�t much you can do about that when that is what the statue looks like:
I�d be a little creeped out having that on my campus.
I can attest that meeting location usually has little to do with the quality of amenities. Milwaukee, which had the lowest attendance in the last decade, is on a beautiful waterfront with lovely art deco architecture and great breweries. ... Portland and Albuquerque... were far from amenities and it was challenging to get to and from hotels and local restaurants. I hope more folks will realize this, and check out the places that aren�t as glamorous.
I�m also quoted in this article about the future of research conferences. The overarching theme seems to be that people want more interaction at conferences.
Opening up reader submissions for this blog is interesting. Sometimes, I make an audible sound when I first see the poster. Sort of a sharp intake of breath. Not quite a gasp. The sot of noise you make in the passenger seat and you see a car coming towards you and you�re not sure if the driver has seen it and you can�t hit the brakes or steer?
Maybe not quite that bad, but... it�s not a good sound.
Then there are times when you open up the file, and think, �Well, dang, am I going to have anything to write about that?�
Today�s contribution is more in the latter category than the first. It comes from Sourav Chakraborty, who gave me the okay to show this to you. Click to enlarge!
Sourav was inspired by a poster by Josefine K�hberger on this very blog, in fact. The result is a nice, clean, attractive poster. There is not a huge amount of text. The layout is clear. The base colours are subdued neutral shades (which I think is one of the main influences from Josefine�s poster), with brighter colours used to good effect for emphasis and highlighting, particularly in the code.
This poster uses bulleted lists, which I generally don�t like. Let�s have a closer look:
This list might be improved by creating a stronger and more distinct hierarchy between the different levels. The main bullets are black squares, and the secondary bullets are black circles.
It�s good that the two levels have different shapes and sizes, but the differences are not that big. I might try reducing the point size Particularly from a distance (or when reduced in size on the screen), the squares and circles look pretty similar. If you�re going to use different levels of lists, you want to make it clear that they are different.
Here�s a quick change to make them more distinct: a hollow circle instead of a filled one.
The difference alone is not enough: you also want to make sure that the differences work in the right direction following expectations of hierarchy. Here�s an example, where I create the same difference (hollowing a symbol), but the other way around:
Lightening the squares works against viewer�s expectations. You�ve made something important lower contrast, making is less noticeable, signalling that it is less important, not more. But the position says it�s more important, not less.
Here�s one more revision where I shrunk the secondary bullets to about 80% of the original, again to create a bigger difference between the different levels of text hierarchy.
Now it�s clearer which are the main points, and which are the secondary points.
Egalitarianism is great socially, but it�s not so great in text design.
A couple of months back, I was one of the instructors in the #SciFund poster making class. We had decided to require everyone make their posters in Adobe Illustrator, which I have never used before. This freaked me out a little bit, and I knew that if I was going to be useful to students, I would have to figure out Illustrator myself.
I decided that I had to make a poster at the same time the students were. I just had one problem: I wasn�t going to a conference this summer, so I had no actual need to make a poster. I decided to tackle the data on a paper that was going through the editorial process at the time, and was finally released today (Faulkes 2015).
I wasn�t extraordinarily diligent in documenting my process, but I did try. This first one is fairly early in the process (click to enlarge):
What surprises me in retrospect is that from a distance, this first one is very similar to what I ended up with. The basic layout decisions � five columns, three pictures in the middle � served me pretty well. But you could not hang in a conference. Obviously, pictures are missing, and if you click to enlarge, you will see a lot of silly placeholder text (from a variety of sources).
Despite that I normally tell people they don�t need logos, I included one mainly because it looked like I would had space left over. This was a simple way to fill it, and the colours matched the picture.
A few steps later, and the poster already looks very close to done. But as we�ll see, looks can be deceiving.
First, I ditched the standard �IMRAD� headings. My idea was to try to make the poster quickly readable by making every heading a key question or finding. That way, you only had to read a few sentences to get the gist of the poster.
Second, I pulled in colour. It just happened that the pictures I found tended to have green and orange in them, which, coincidentally enough, was the colour scheme for the new University of Texas Rio Grande Valley mascot. I used the eyedropper to duplicate colours from the mascot and photos to the headings, the box around the pictures, the title, and so on.
Third, I put in the data. I considered making graphs, but I kept thinking that these were simple, easy to understand numbers, and there were not very many of them. The central graphic is, in essence, the thing I tell people to never put on a poster: a table! But it�s a table with photos, lots of space, and no �data prison.�
Fast forward a few more steps:
The obvious change when you see the thumbnail is that I�ve moved the mascot. I placed the mascot in the lower right corner following the Cosmo principle: that�s where the least important stuff goes. The problem was that the Vaquero was facing outwards, leading your eyes off the poster. I moved it one column over, just because I didn�t want to move it very far.
But that wasn�t far enough!
Now the mascot is clearly facing into the poster, leading your eye into the next section of text. Much better.
You can�t see at a glance are all the changes to the text I�m making as I go, too. But trust me, there is a lot of editing and rewriting going on.
This is the final version:
I know it doesn�t look all that much different from the second image above, but there are so many chances that you can�t see in the thumbnail. They are the little things like increasing the text size, changes in wording, and the space between the lines. They are almost subliminal differences, but they all add up to a much nicer appearance, as I wrote about here.
One of the last changes was which numbers I used in the central graphic. I rounded the percentages up to got rid of the decimals. They just weren�t necessary. I also changed which numbers I showed in the second row, which much more clearly indicated the popularity of one species (almost half of all sales!).
The decision about which numbers to show on this poster, in fact, led to me asking the editors to make some last minute changes in the published paper. Because I was forced to grapple how to show things clearly and visually on a poster that I realized there were some nice improvements I could make to the paper.
I�ve given just a few examples of the stages in making this poster in this post, but you can watch the development with more steps in this video:
Clearly I had not yet taken on board the lesson of editing. This was a problem in a lot of my old posters I made before I started this blog; see here and here in particular.
Yes, there�s even an abstract. The one thing I will say in my defense is that the instructions specifically said to include the abstract, and I was still a few years from realizing there are no poster police, and becoming an abstract anarchist as far as posters were concerned.
At the time, I was happy enough with this poster to have my picture taken with it. I can�t recall who I loaned my camera to, but I�m so grateful to him, because this is a favourite picture of me to this day. I felt like this picture showed my in my natural element.
My reaction to the poster now?
Not so much a critique as a cry of anguish.
Too much stuff and not enough space. I cringed when I looked at the guideline settings and saw the columns were only separated by half an inch. Nowhere near enough of a margin.
There were also a few blatant errors in the text that I never caught until now. No, I�m not going to tell you what they are. I shall leave that as an exercise for the reader, as they say.
I am happy that this poster is laid out in columns, with at least a major grid structuring the poster. I also learned something very important from doing this poster: rehearse the poster out loud. This is the poster that inspired this story:
For one poster I did, I had a figure that ended up in about column four, quite far to the right of the poster. (Black and white image at top of column four - ZF, 2015) I thought it made sense to put it there given the poster space. It felt fine when you read through the poster.
But when I gave people �tours� through the poster at the meeting, I kept referring to that picture very early on, when people were mostly examining stuff on the left side of the poster. People had to look way over to a different section of the poster, and it disrupted the flow of the presentation. (In that case, it was exacerbated by the poster being over two meters wide. People had to look a long way over to see the picture.)
Because this is one of my own posters, I was able to open up the original Publisher file and start editing. I didn�t give myself anything that I wouldn�t have had at the time, like new images. Here�s the revised version:
I made all the margins two inches. I hacked away a lot of the text, and replaced the stupid abstract with a picture of the study species, which people can more readily relate to and understand. That one key figure that threw off my narrative because it was too far over to the right got moved up to the introduction, too.
It�s better, but honestly, I can see this version is still struggling with the baggage from the first effort. I�m not sure those three tables are helping my cause. And there is still too much text. But I am not going to redo the poster from scratch because I have better things to do than completely remake a poster from a conference more than seven year ago. (But apparently I don�t have better things to do than write a blog post about it.)
If I were to design the poster again from scratch today, it might be a lot more like these graphics that I made to promote the paper on Twitter. None of these graphics could be a poster as is, but they give an idea of the approach I took in making a compact version of the paper.
The one above has the picture of the shrimp, which is nice, but it needs more detail for the results. Remember, the point of this is not to be complete, but as an enticement to get people to click a link to a longer article.
This next one below is probably closest to a working poster:
Nice, simple, straight head to head comparison between to species. Put in a title, a picture of the animals, and this is close to something you could hang up on the conference poster board.
This last one has a clear title and some more detail:
I worry that it has a little too much detail, but that central panel really drives home the difference between what was expected (two separate cell bodies on the side) and what I saw (massive, hard to tell apart cell bodies in the middle).
As much as it hurts to go back into your old work, it is nice to go back and see how far you�ve come.
References
Faulkes Z. 2007. Motor neurons involved in escape responses in white shrimp, Litopenaeus setiferus. Integrative and Comparative Biology47(Supplement 1): e178. http://dx.doi.org/10.1093/icb/icm105
