Kamis, 08 Januari 2015

Latest modern science | Critique: Plague - Si Bejo Science

Alison Atkin has an interesting and award-winning poster here. Click to enlarge!


The first thing that stands out about this poster is that it is hand drawn. Wow. I�ve only had, I think, one other completely hand-made poster on the blog before. That was done by someone with fine art training, but this is different. It�s lo fi, and personal.

What I love even more about this poster is how it invites you in to come and play...



For more examples of �interactive� (that is, pop-up) panels, make sure to read the full blog post.

It�s a little difficult to judge the poster in its entirety here, because Alison notes the image was is a composite. Assuming that this is reasonably true to the original, the only thing I would have liked to have seen would be stronger visual cues to read across in rows, not down in columns. This could be done by making the horizontal gaps a bit wider than the vertical ones, or by placing the test very consistently at the top. The critical first two panels put the text at the top, which set the pace for the rest of the poster.

I love this poster. Something like this would stand out at any conference for the amount of work it represents, its uniqueness, and its charm.

Hat tip to, er, Alison Atkin and Wellcome Trust for this Storify on accessible scientific writing.

Related posts

Combining art and science: Karmella Haynes interview

Jumat, 02 Januari 2015

Latest modern science | Interview with visual information specialist Karen Nelson - Si Bejo Science

Conference posters are typically made by amateurs (and boy oh boy, does that ever include me!). It�s rare to find someone who makes posters as part of their job, who is not an academic, and who has training in design.

This makes Karen Nelson rare as the proverbial hen�s teeth. Her email signature describes her asa �Visual Information Specialist� for the United States Forest Service. She graciously agreed to answer some questions and show a couple of her posters. (Click to enlarge the posters!)


Q: Tell us a bit about your background. How did you become a �visual information specialist�?

A: I graduated from the University of Wisconsin with a B.S. in Art (and an emphasis in typography and graphic design). After graduating, I was hired by the U.S. Forest Service, Forest Products Laboratory (FPL), as an entry level �Visual Information Specialist� (Graphic Designer). My skill sets improved over time, and I�m now a senior designer.

Q: Before you started designing posters, had you been to scientific conferences and seen the posters?

A: I did not attend scientific conferences before I started to design posters. However, I saw many research posters created by a former designer at the Forest Products Laboratory, who was my mentor before he retired.

Q: Do you go to conferences now? If so, what�s your general impression of the state of conference posters?

A: Due to budget constraints, I do not travel to conferences. However, I do review posters presented at conferences hosted on-site by my employer. My general impression of conference posters is that they are usually �journal articles on a large piece of paper.� Posters are overloaded with text, charts, tables, etc. Color choices are poor. Proofreading is overlooked. Primary messages are lost due to information overload.


Q: Describe the process of working with the researchers. How much of the text and graphs do they give you, and how much do you create?

A: The researchers are responsible for providing text and graphics. I ask researchers to provide (1) conference guidelines for the poster (2) text in Word, (3) charts (with data) in Excel (whenever possible), (4) tables in Word or Excel, and (5) original, unaltered, copyright-free photographs at the largest file size available. If it�s necessary to use a copyrighted image, the researcher must obtain permission for use. I specifically request that they do not embed photos, charts/graphics, and other elements in a Word or PowerPoint document. I work more efficiently when I receive individual files, and I think that quality is lost when I have to copy/paste an image from Word into Photoshop.

I begin the design process after I receive all content. I work directly with the researchers to eliminate unnecessary information. I�m a good proofreader and copy editor, but I consult with an on-site technical publications editor as needed. If the photographs are low resolution, I request different files (or I will look for them myself). I frequently re-create graphs and artwork (flow charts, diagrams, etc.), especially when I receive low-resolution image files that are not editable.

After a draft poster is done, I meet with the researcher(s) for review, corrections, etc. I usually output a small but readable print for markup. Sometimes, simply emailing a PDF will suffice for review.


Q: What is your poster design process like? Is it purely digital? What software do you use to put posters together?

A: Yes, the design process is fully digital. If a poster contains tables, I use Adobe InDesign for layout (InDesign is great for importing and editing tables). If a poster contains charts and vector graphics that need to be redrawn, I use Adobe Illustrator (Illustrator�s graphing tool is also very helpful when data is provided). I edit and color correct raster images in Adobe Photoshop. We have an in-house large-format printer.

Q: Do you have any advice to help a scientist making a poster? Putting it another way, what are the pitfalls that people not trained in design fall into over and over again?

A: KISS! In a room full of 100 or 200 or 300 posters, let yours stand out and attract attention. Portray the main message and important results � not all of the journal article details. Make the design process easy � use large, pertinent photographs, succinct graphics, and a minimal amount of text.

Don�t use dark or brightly colored backgrounds. Instead, keep the background white or use a light, neutral color so that your graphics and photographs can pop.

Don�t use boxes! Instead, leave plenty of white space between columns and sections of information. If you �need� boxes, you have too much information.

Please avoid 3D charts and gradient fill patterns! Remove all �chart junk.� Read Edward Tufte�s The Visual Display of Quantitative Information to learn more. Choose a limited color palette and a limited number of fonts and font sizes.

Consider a handout or a business card with links to sources for more information.


Q: What other kinds of visual information are you charged with making in your work? How do the design considerations differ?

A: My work is divided between technical/research-related items for the scientists and other design products for semi-technical and public audiences. I have designed technical manuals, including all graphics contained therein, figures for journal articles, technical brochures, research-related PowerPoint presentations, and some web graphics. On the other hand, I create items for public consumption such as semi-technical handouts, fact sheets, web graphics, PowerPoint files about Forest Products Laboratory, and posters/displays that again are general in nature. The design considerations (principles), to me, are the same for both audiences, yet designs for public consumption allow for more creativity.

Q: Straight graphic design geekery now: Do you have a favorite typeface?

A: No! It depends on the subject and audience. For research posters, I like Myriad Pro Semibold for titles and heads. It�s an easy-to-read sans serif face. I also like Cronos Pro (sans serif) and two serif font families (Minion Pro and Adobe Text Pro).

Thanks to Karen for taking the time to answer some questions!

Related posts

Combining art and science: Karmella Haynes interview
Critique: geese and swans

Cronos Pro sample from here.

Kamis, 25 Desember 2014

Latest modern science | Link roundup for December 2014 - Si Bejo Science

I�m always on the lookout for re-use of posters, and Gary McDowell takes advantage of the new fabric ones:



This poster scarf actually predates a similar one seen at Neuroscience by a few days:



Choosing the right title for your poster is critically important. This New Yorker article shows that the headline changes the way people remember the content of the story you tell them next:

In the case of the factual articles, a misleading headline hurt a reader�s ability to recall the article�s details. ... In the case of opinion articles, however, a misleading headline... impaired a reader�s ability to make accurate inferences.

Dr. Attai took this picture at American Society of Clinical Oncology 2014 meeting. Um. Aren�t people proud of their work?


NatC has a conference networking tip:

New conference networking strategy: share cab to airport with strangers. Get career advice.

I constantly harp on people to make a grid. Here is a useful slide deck showing how grids are used to design a complex website (hat tip to Duarte and Garr Reynolds):



Kirsten Sanford nominated this as her favourite poster from the American Geological Union meeting. It�s colourful, I�ll give it that.



Business cards are an integral part of conference networking. Erik Peterson turned his business cards into mini-posters:


Slideshare has a video that claims to tell you four tips for making data visualization memorable (hat tip to Ethos 3):


The cheat sheet summary is below; the original paper is here:

  1. They look like something natural.
  2. Are pictoral.
  3. Use colours.
  4. Have high visual density.

Typeset in the Future is an obsessive single serve blog looking at typography in science fiction films. Try this post on Alien for starters. Hat tip to Adam Savage. (The mythbuster also throws in his favourite typefaces: Futura and Caslon.)

Before and After talks about using colour to make connections between objects. Very useful to remember in designing posters, and displaying data.

Here�s a fun article about secrets hidden in plain sight in logos. I knew a couple, like the FedEx arrow, but there were lots that I didn�t know.


Note, though, that the article gets the story behind the BMW logo wrong (it�s not a propellor). But perhaps it can be forgiven, as BMW�s own histories have sometimes mucked up the truth!


Merry Christmas!

Kamis, 18 Desember 2014

Latest modern science | Critique: The golden blogosphere - Si Bejo Science

This poster comes from Joel Topf. Click to enlarge!


The idea of the layout is good. Having the text all concentrated in one short summary that looks like it can be read quickly may help viewers who want to skim. But some of this advantages are defeated because the poster is still extremely dense. For instance:

The address breaks the grid by curling around to the right of the �Introduction� heading. Those are too close. Each one is a separate element, and deserves its own defined space. Instead, the two sections are overlapping in space, and they will look better if separated.

A similar problem occurs with the bottom graph: its space is invaded by the graphs above. This is particularly noticeable where the �Total number of posts� graph (gray box) comes close to touching the blue data line in the graph below.

The text looks pretty brief, but wonder if it could be edited down even more. For example, I took this from 41 words:

Nephrology bloggers are rarely compensated and their writing is not usually considered part of academic production in regards to advancement. Without obvious advantages for the blogger, I thought that bloggers must thrive on internal enthusiasm and it may wane over time.

To 30:

Academic blogging is not usually rewarded in career advancement (e.g., tenure and promotion ). This suggests bloggers are intrinsically motivated, but this may wane if there are no extrinsic rewards.

The more you can edit, the more space you can open up.

I wanted a bit more guidance for all the data on the right side of the poster, so that I know what is being shown here. Some one sentence summaries next to the three main sections would be welcome.

The colours in the table are not explained anywhere. I am guessing �green�means statistically significant, and �orange� means... a decline in posts over time? Maybe that could be mentioned in the main text at the left.

The table is big and dense. Again, I wonder if it could be simplified, either graphically (first step: remove the vertical gridline!) or even removed. If I�m reading it right, some of the information in the table is repeated in the graphs to the right of the table.

The last line of the table - �Totals� - appears to be incorrect. It looks like most of those entries are means, not totals.

Also, the text mentions 30 blogs, but only 22 are plotted.

Where the QR code goes is a mystery. It�s a helps to tell people what they�ll get by scanning a code. Further, the bit.ly short link goes to the same site as the QR code. I suggest picking just one. I lean towards keeping just the QR code, because I have yet to see anyone type in that complicated alphanumeric short URL. But if both were left on the poster, I�d try to make the bit.ly URL the same width as the QR code.

Additional: Joel provides a well mannered response to my critique.

External links

Nephrology Blogosphere poster

Kamis, 04 Desember 2014

Latest modern science | Critique: Sex models - Si Bejo Science

Yes, I totally went for the gutter headline in introducing this poster from Amanda Whitlock. This poster is used with her permission. You can click to enlarge!


Amanda included this note in her email:

After reading your blog, I switched from using Powerpoint to Scribus and have become a huge convert and evangelist on its behalf.

I�m so pleased there is one less person in the world using PowerPoint for posters!


Amanda�s poster has a clean design. It starts with my favourite �hard to mess it up� layout: three columns, equal size.

I recently read an article that argued that anytime you overlay text on a picture, it should always be white text on top of the image. That message might be a good one for all poster makers. Especially when viewing this poster at a reduced size, I�m worried that the title (90% of your communication effort!) is barely visible. I tried a quick and dirty replacement of the black with white:



The edges of the text are badly pixelated because of the way I inverted the colours, but the title is more visible. Let�s try the same to the headings:


The difference is harder to see, but might be more obvious if made in the original document. The headings would also benefit from a bit of additional work to ensure that they are all evenly spaced. �Conclusions� looks closer to the bottom of its bar than �Results,� for example.

The same goes for the vertical alignment. A line that misses the letters in �Conclusion� hits letters in �References,� for instance.


While the poster has plenty of images and white space, it is a shame that the critical upper left corner is the least visually appealing part of the whole poster, with only text.

If you can�t see the title, and there is only words and data, it�s unlikely to gather any new readers who just happened to be walking by. While this poster will not make anyone cringe when they walk past it, they might just... walk past it.