Shadows are not edges

A question recently landed in my inbox about alignment. If some elements in a layout use shadows but others do not, should the shadows align with the edges?

Probably not.

This is a case where the psychology of visual perception is may be helpful. Human eyes are very good at detecting edges of things. Even implied edges. This is a phenomenon known as “illusory contours,” or occasionally “cognitive contours.”

Usually, a shadow implies the edge of an object, rather than being seen as the edge itself.

In the Better Posters book, Figure 19.3 and 19.4 on Page 212 make a similar point. The graphs in 19.3 at the top are technically aligned: they all take up the same space, same height, same, width. But the axes exert such a strong visual pull that it looks far better to ensure the axes line up.

Another example of this principle is hung punctuation: the practice of moving opening question marks outside of the boundary box created by text. The body of the text along the left forms a stronger edge than the small quotation mark, so the letters are aligned rather than the opening quotation mark.

Now, we could debate whether it is good practice to have some elements with shadows and others without, but that is another post for another day.

P.S.—Thanks to Samir for emailing a question! You too can email me at betterposters@gmail.com.

Related posts

Hung punctuation

External links

Samuel L. Jackson placeholder text generator 

Carl Sagan placeholder text generator

Placeholder text generators (not all links current)

Ten simple rules for conference posters

I am being a lazy blogger and turning a Twitter thread into a blog post. John Butler asked for, “Something like 10 steps for a good conference poster.”

So I made something up off the top of my head.

1. Read the instructions. Sounds easy, but printers say “wrong size” is the #1 problem they see.
2. Your title is most of your communication effort. It’s all most people will ever read. Spend a lot of time on your title! Don’t just use the first one that comes to mind. Simple declarative statements of the main finding work well.
3. You should be able say what your poster is about in one sentence. Too many people want to show everything they have done. Focus.
4. Make you one sentence in an ABT (and, but, therefore) format. What are a couple of facts? (“We know X and Y...”) What is the problem? (“But X doesn't hold in this case...”) What is the consequence of that? (“Therefore...”)
5. Make a grid. A three column grid is really hard to screw up. It’s not the only way. More or fewer columns can work. Rows can work. But three columns is a robust layout.
6. Leave space. Lots of people, because they did not focus enough (#3), make skinny little margins to try to fit more stuff on. It’s hard to read. Be generous with margins between columns, and with white space between text and graphics.
7. Be consistent. Consistent fonts. Consistent colours. Consistent column width. A common is that people make graphs before the poster, and don’t go back to make the graph fit with the rest of the poster.
8. Bigger is better. A common question is, “What’s the minimum font size for a poster?” (This is often coming from people trying to shove too much stuff on the page.) Accessibility guidelines usually recommend type be several times bigger than what many people use.
9. Practice what you are going to say. Do this before you print your poster. Sometimes you’ll find the order you laid stuff out in (because “It just fit there”) is not the natural order when you talk it through. The visuals and your explanation should follow the same order.
10. Do the “arm’s length” test. When you’re done layout, print your poster scaled down to a single letter sized piece of paper. Hold it at arm’s length. If your vision is reasonable, you should be able to read your shrunk down poster at arm's length. If you can’t, it’s too small.

The Twitter thread has marginally related GIFs.

External links

Ten simple rules for a good poster presentation

(PLOS Computational Biology)

Journals need to set a better example for posters

I recently ran across a figure in a journal (and a journal I respect) that made me look sideways at it. I am suspicious of composite, multi-panel figures at the best of times. But oh boy.

First, this was the layout of the elements in the figure.

Nine elements should be easy to lay out. Make a three by three grid heights and widths. Instead, we get this bizarro layout where only two elements are the same height or width (bottom right corner). The five and eight sided polygons in the middle are making me cringe. 

But based on the blocking of the elements, I thought I might be able to go through the figure in a sensible way. It looked to me like I should go across, then down.

But no. When I looked at the numbering and figure legend, I found I was supposed to read the figure in this order.

So the actual reading order for this figure is a drunkard’s walk that includes right to left transitions and backtracking across previously covered terrain.

At least the ordering starts in the upper left and ends in the lower right.

I can understand authors making a figure like this. Scientists are not graphic designers. But what I can’t understand is why a journal editor didn’t have something to say about this. Something to say as in, “Redo that figure.”

Posters are deeply influenced by journal articles. They almost always have the same elements as a journal article, even though there is rarely a need to imitate the format. So slipshod graphics in journals have a ripple effect. They sets a poor example that readers who later make posters might imitate. People think it’s published so it must be okay.

It could be so much better.

I get the impression that a few journals have dedicated graphics staff who often work with authors on their illustrations. I wish that professional graphics designers were employed more widely by journals, and that the journals advertised that.

Lessons from Pac-Man

Shiz Aoki of BioRender sometimes gives this tip when discussing posters:

Imagine Pac-Man being able to travel between the sections of your poster without getting stuck.

This is good advice, but I want to run with it a little more. I think some people will read this and think that “blocking” is the problem. Poor Ms. Pac-Man has nowhere to go here.

But there is more to this it than ensuring there is enough space.

Here is an example where Ms. Pac-Man has all the space she needs to get anywhere on the screen... but it is still a frustrating screen to view and would be impossible to play.

The problem here is not the amount of space, but the alignment of it. There is always “enough” space in the sense that Ms. Pac-Man and the ghosts can get around the screen, but the width of the maze is always changing and none of the edges of the walls line up any more.

If you look at mazes, the corridor width is often kept absolutely consistent.

This can be true regardless of the overall shape of the maze.

Or what the maze is made of.

Mazes like these (including Pac-Man) are fundamentally grid systems. It is not enough to just have “at least” a certain amount of space between elements. It should be the same amount of space whenever possible.

Garden maze from here.

Critique: On spec(trographic)

Today’s contribution come from Michael Young. This will be shown at this year’s Astronomical Data Analysis Software and Systems conference, but he’s given me permission to show you as preview. Click to enlarge!

Because the poster is intended to show off the website, it features near the middle of the poster, which is appropriate placement.

Michael’s own assessment of the poster is that despite much editing, it is still too wordy. I agree, but the good news is that the typography is clean and understandable, apart from a tenuous touch of the institution name with the title bar. The descender on the “y” in “University” is killing me.

I also the colour combination of blues and tans. It’s consistent throughout. The blues provides a good contrast to the logo, appropriately situated out of the way, unobtrusive and quiet.

The underlying structure of this portrait style poster is a solid two row grid:

I might have preferred a little more space between the rows to separate them, but the space between each row is not the biggest problem here. You might not notice the simple two row layout because of the placement of everything in those rows. Let’s highlight the pictures and sidebars:

Just to make it a little more obvious, here’s the position of all those elements without the distraction of the poster contents:

Now it looks just a bit like a Mondrian painting.

The point, though, is to highlight that there’s no underlying plan for those pictures. Almost no two are the same size. They’re only aligned if they happen to be along the poster’s edge.

Consequently, the reading order of the poster, while clear (thanks to the underlying two row grid), is circuitous. You have to wind and wend your way around all those pictures.

What might have helped this poster is a stronger secondary grid. How is the row going to be divided? Could it be quartered, or otherwise sectioned into smaller pieces?

Related posts

Avoid the tenuous touch

Common problems

I teach a technical writing class in which I ask students to create posters. I give the students some instruction and and good ideas for poster layout. Generally, their work is as good or better than conference posters I see.

The top three problems are:

1. Clutter: It’s hard to generalize the sources of clutter. It ranges from unnecessary borders to gradient fills to too many elements on the page.
2. Proximity: People underestimate how much white space they need around objects. In particular, they don’t put big enough margins around pictures.
3. Text greyness: Even though my students generally do a good job of making the text large enough to be readable, they often end up with too much of it.

If there is a common element to these three problems, it’s the urge to put too much stuff on the page. Editing and cutting is a tough skill that takes a lot of practice to master.

Bad poster bingo

Play it at your next conference! Click to enlarge.

At most conferences, the game is not whether you can get a bingo, it’s whether you can do it on a single poster...

External links

Bad presentation bingo

If iPads were made by academics

This is the default screen of a new iPad. Calm, cool, focused.

This is what this screen would look like if it made the way academics make their conference posters.

Everything squeezed, nothing aligned, lots distorted.

Or maybe this.

Boxes and Comics Sans; now that’s the stuff.

Seriously, people. It feels like 90% of conference posters look like those bottom two.

To repeat the lessons:

• Big margins.
• Make everything line up.
• Lock the aspect ratio when enlarging or shrinking.

Neuroscience 2011: advice in advance

This week, I shall be at the annual Neuroscience meeting, which is one of the biggest, if not the biggest, scientific conferences, in the world.




A huge number of posters will be put together and printed this week. For those who do not want to have to dig through the last two and a half years of posts, here is a super quick guide to making your poster better.

The big tips

Shut up.

People want to read as little as possible. And at Neuroscience, who can blame them? There is more than any person can see, and big blocks of text scream, “This is going to take a long time!”

Show images whenever possible, especially pictures. Pictures of actual physical objects that people can recognize are always more inviting than graphs, which are generic and abstract.

Do your audience a favour: respect their time. Leave the fiddly bits for your dissertation or published paper.

Line up.

For the love of humanity, make a grid. Divide your poster up into evenly spaced pieces. Draw lines. Make every piece of text, every chart, every thing line up along those lines. Don’t just stick things here and there, or make them “almost” align. It looks sloppy.

Size up.

Make everything bigger. The text. The space between each line of text. The graphs. The labels on the graphs. Large size is a key factor in how long people will look at something. How do you know if it’s big enough? Does your poster pass the arm’s length test?

Free up.

Drawing boxes around everything is like putting your poster in prison. Solitary confinement is cruel. When your poster is well organized, all you need is white space to separate the sections.

Those are my main requests that can help make most posters better. Even I, with a lot of posters under my belt, forget these basics sometimes.

The next tips are things that can be ignored with a little practice and thought, but are good places to start.

You can’t go too far wrong with...

Black text on a white background

Black text on a white background works. You can do cool things with backgrounds if know what you are doing, but it’s easy to turn something looks horrible on a printed page. Most people should put away the gradient fills and the photo blow-ups, and stick with the tried and true colour scheme that you see in 99% of all books.

A three column layout

Photographers learned long ago that dividing things into three is pleasing. For big, wide posters, like Neuroscience, three columns may be too wide. If that’s the case, either create an odd number of columns (five or seven), or subdivide your grid into smaller sections.

Sans serif typeface for the main text

Some suggestions for typefaces that are fairly easily available include Helvetica, Arial, Verdana, Tahoma, Gill Sans, and Calibri. You may think these are boring because everyone uses them, but there is a reason people use them so much: they get the job done.

Personally, I like Gill Sans on posters. Its readability holds up at a distance.

Calibri – while a fine typeface – does convey a subtle message: “I can’t be bothered to change the default settings in Microsoft Office.”

And finally:

No frickin’ Comic Sans

Comic Sans says, “amateur.” You might be going for “humourous” or “approachable,” but you achieve that at the complete loss of the “professional” trait.

I’m not against fonts with a hand-drawn look. I am a big comic book fan. Many people could make their posters better by learning the lessons of comic books. But there are better comic book typefaces out there.

Do you have a poster at Neuroscience that you are proud of? Want to show it off? Get feedback? Or just chat? Email me! I will be at Neuroscience from Saturday, 12 November, to Monday, 14 November (and maybe part of Tuesday morning).

Unabashed plug! If this helped you, please consider supporting my scientific research.

Using grids in PowerPoint

I can usually recognize a poster that has been made in PowerPoint because nothing lines up. This is part PowerPoint’s fault, but not entirely. If you must use PowerPoint to make your poster (even though you shouldn’t), here’s how to make it look like it was done in a real graphics package.

Click to enlarge any of these images.

First, turn on the grid. Under the View tab, select the “Gridlines” button (just to the left of the magnifying glass).

The default grid size is a strange value: one twelfth of an inch. (Page layout is about the only part of my life where I routinely use Imperial measurements.) You can change the grid size by bringing up the Grid Dialogue box (the little pop-out button is just underneath the button you clicked to get gridlines).

If you drop down the “Spacing,” it will show your options for grid size in fractions rather than decimals.

I usually find it much easier to work with a coarser grid. Grid spacing at one quarter inch makes it easier and faster to align things.

PowerPoint has two “guides” that cross the vertical and horizontal center of the page. Because they cannot be changed, they are useless for most purposes.

As a work around, I draw in my own guidelines for columns and such using the line tool. This takes a bit of calculation, since you have to work out the exact position of each line in advance, and can’t tell PowerPoint you want a three column grid, like you can with Publisher.

After you create your guidelines, you can start dropping in your text and pictures, safe in the knowledge it makes it much easier to get everything with a consistent width.

But after you’ve laid everything out, you have to remember to go back and delete all the guidelines you had for the margins and between the columns.

If you are very careful and consistent, you can make a poster in PowerPoint that looks like it was done in a proper graphics package.

For examples of what you can do with PowerPoint, check out this post at the I was lost but now I live here blog and the featured artwork. Or some of the artwork at PowerPoint Heaven. I particularly like this portrait of Jack Sparrow from the Pirates of the Carribean films (ZIP file) and this anime girl (ZIP file).

It’s a long, long way from yellow bullet points on a blue background. I think you have to work to hard for these results, but I am nevertheless impressed by the perseverance of the artists!

Related posts

The importance of alignment

Boxism

Boxes, more boxes, boxes within boxes, and boxes everywhere!

People making research posters seem to be obsessed with boxes. Walk through a poster session and you’ll find poster after poster where each section, sometimes each figure, is set apart by big, thick, heavy boxes.

Then there are those who draw a box around the entire poster, a box around the text column of the poster, a box around each graph of the poster, and the effect is much like staring at an image of a video camera pointed at its own monitor, a tunnel into some unknown dimension...

Don’t believe me? Here are just a few examples, picked because they were near the front of ePosters and the Pimp My Poster Flickr group. These are not particularly extreme examples; I just grabbed them to indicate how pervasive the drawing of boxes is.

Let’s start with this award-winning poster on ePosters (click to enlarge).

The boxes around the graphs are more subtle, because there is not a line around them, but the white against the coloured background still creates a box around the graph.

Here’s one from the Pimp My Poster group (click to enlarge).

These posters both have other issues besides the relentless boxism, but those are other issues for other times.

And just for the heck of it, one more from Some Beans, which doesn’t need my callouts to point out the box obsession:

Imagine how this blog would look if every paragraph was contained in its own little box. It wouldn’t be much fun to read.

Many posters would be substantially improved just by taking away the boxes, and using white space to separate everything instead of lines.

Boxes are usually a last-gasp attempt to enforce some sort of order upon a chaotic layout. Excessive use of boxes betrays the designer’s lack of a plan or a lack of confidence. To avoid descending into a Russian doll of boxes:

1. Start with a simple grid. Try three equal columns if flummoxed.
2. Make the margins between the columns wide. Clearly defined margins will help guide people in which way to read.
3. Make your material fit the grid; don’t change the grid to fit the material. You may have to go back and redo graphs to different proportions.

Boxes can be used to good effect. Boxes around entire columns, or callouts of highlights, can be attractive. But every single element does not have to be contained in its own box.

Type crimes: Do as I say, not as I do

From Failblog.

The eye loves the circle

Kimberly Elam wrote:

The human eye loves the circle and embraces it.

While I am a strong advocate of laying out posters on a grid, you can end up with a poster that is relentlessly rectangular. A circle can be a strong antidote to a poster filled with right angles.

Circles can be used to draw attention. It is no accident that circles are used in those ubiquitous bullet lists.

Circles can create tension. Like a ball, they suggest something that is mobile and not static.

Circles can be used to create white space. As everyone knows, a round peg will not fit into a square hole without leaving spaces.

Because they do tend to break, rather than reinforce, grids, circles are probably best used in small doses on a poster. A poster without a circle will not be noticeably missing anything. But it’s a useful exercise to consider how you might work a circle into a poster grid.

Reference

Elam K. 2004. Grid Systems. Princeton Architectural Press: New York. Amazon

Photo by user Oranguthingy on Flickr, used under a Creative Commons license.

No more slidesters, part 2: Three Publisher tips

I have used Microsoft Publisher a lot for posters, as I mentioned previously. I’m going to show three easy things that Publisher does well that are useful when making a conference poster.

Grids are a cinch

Grids are at the heart of successful poster design. Under “Arrange,” pick “Layout guides,” then pick the “Grid guides” tab.


From there, all you have to do is pick the number of columns you want to have, and the space between them. And you’ve now got an evenly spaced grid!


Setting up grid guidelines in other programs is rarely this painless. You usually have to do a lot of calculating to set up each individual guideline.

You can also set up rows in the same way. This is not as useful as often in posters, which are usually dominated by vertical divisions, rather than horizontal (apart from the title).

Once you’ve got columns, draw a single text box in each one and paste in your text. Because posters usually have many graphics, you’ll want to insert pictures, too. This is where another feature of Publisher comes in handy.

Text wraps around pictures

Many other programs either cannot wrap text automatically (like PowerPoint) or make you jump through hoops to get there. When you insert a picture into Publisher, the text will normally flow around it. If the picture is the same width as the column, the text will break and continue below it.


The default spacing is not particularly graceful, so it is worth fiddling with the settings after the picture is in so that the text isn’t too close to the picture.

Joining text across columns

Many people, particularly those working with PowerPoint, create their posters as a zillions small text boxes that almost invariably go out of alignment because there are so many of them, and they are all different widths. Publisher allows you to link text boxes into one continuous piece of text.

Click on a text box. If there is more text than will fit in the space, it will show an “overflow” symbol at the bottom. On the menu bar, there is a “Create Text Box Link” button. Click it, then bring the mouse down to the next column. The mouse icon will change to a bucket. Click it within the box to “pour” the overflowing text into the next column.


Continue this for each column. Once this is done, the text will automatically rescale through all the text boxes. For example, if you add a picture to text box one that displaces some of the text, the text will shift down automatically, through to the last linked text box. The same thing will also happen if you enlarge or reduce the text.

Columns: How many?

That most posters are wider than tall (landscape layout) means that the text and graphics will be arranged in a series of columns. This number and width of columns is probably the single most important factor in designing a grid for a poster.

How many columns should a poster have?

Some designers find even numbers of divisions too symmetrical and staid, like folded paper. Thus, you may want to try for an odd number of columns, like three of five. Alternately, you might work with a grid that is some multiple of three.

There’s another reason to consider trisecting your poster space. Photographers and graphic artists often talk about the “rule of thirds.” The idea is that the most visually interesting images, pages or photographs often have their focal points off center, at one of the four points formed by intersections of imaginary lines that divide the space into thirds both vertically and horizontally.

This is easier to show than describe.


The red crosses show the preferred points for locating the items of most interest. In photography, these might be the eyes of the face, say. Locating things on those points in a technical poster is a bit more complex, but that’s a post for another time.

If there is any text that is long enough to form a paragraph, the size of the type has a major impact on column width, and therefore column number.

A narrow
skinny
column of
text can
be highly
annoying,
because you
are forced
to keep
sending your
eyes back to
the start of
the line.

Reading extremely wide paragraphs of text makes it easy to lose which line you’re reading as you continue to scan across. I’ve had to make the text very small to show an example of this phenomenon, because I can’t easily adjust the column width for this blog, so I hope you can see the point. Sorry for the eyestrain.

According to Jury (2002), a good guideline for paragraph width, and therefore column width, is about ten to twelve words per line of text. You will find plenty of situations where this guideline is not followed; some magazines have narrower columns, maybe six to eight words wide. Nevertheless, those principles alone probably give you a very solid starting point for a grid.

References

Jury D. 2002. About Face: Reviving the Rules for Typography. Rotavision SA: Mies.

The importance of alignment

If anyone learns just one thing from this blog, I hope it is this:

Make a grid.

Let’s look at this award-winning poster, found at Creating Effective Poster Presentations. (Click to enlarge.)


There is much to like about this poster, but the layout could be better. Let’s look at it with a few gridlines superimposed. For best effect, click on the image to open it at a larger size. I’ve added these gridlines by hand to try to align to what is actually on the poster, and not put on a mathematically regular, even grid. A few arrowheads call attention to some of the errors that pop once the gridlines are added.


The Introduction and the Objective paragraphs don’t align. The histological pictures don’t align with their column’s text. The Conclusion and Acknowledgments paragraphs don’t align. The tops of the three columns don’t align. The top graph in the middle column is wider than the one below. The right column is wider than the left.

This poster shows every indication of being created in PowerPoint. I’ve seen it many times before. This sort of poor alignment is a typical outcome of using PowerPoint to make posters, because it does not provide a simple way to create new grids apart from those provided for projected slides.

Lining things up is at the core of typesetting and layout. It is the heart and soul of the craft. Failing to do so makes a poster scream, “Amateur hour.”

Related link

Designing with Grid-Based Approach, from Smashing Magazine