Kamis, 28 Januari 2016

Latest modern science | Link roundup for January 2016 - Si Bejo Science

We have a new contender for �worst graph ever�: the pie cloud.


What... I mean... Why... I... I give up. Shudder. Hat tip to Andrew Gelman.

Pieter Torrez has an article on how to create a beautiful scientific poster.There�s good advice on use of colours, text, software tools. The only thing I�m not sure I agree with is adding a picture of yourself.

Eve Heaton decided to use the trick that every conference vendor learned long ago to attract passers-by:


Hat tip to Colin Purrington.

Because PowerPoint is so often used to make posters as well as presentations, I have to link to this long, thorough analysis of PowerPoint�s history and use. The history is impeccable, although the analysis of PowerPoint�s importance is variable and sometimes told in fancy academese instead of plain English. Here�s an excerpt I like (that applies to poster presentations, too):

Rich Gold, manager of the Research in Experimental Documents group at Xerox PARC and self-proclaimed PowerPoint maestro, characterized presentations as jazz. Slides are merely the starting point, the �bass rhythm, and chord changes over which the melody is improvised.� ... Reading from notes or slides violates the expectation that a speaker can lay it down fresh every time, connecting with the group around a commonly held artifact.




Check out the list of 2015�s most popular fonts. Plenty of gorgeous fonts, though quite a few would only be good in very small doses on an academic poster.

Kamis, 21 Januari 2016

Latest modern science | Critique: Thale cress RNA - Si Bejo Science

Today�s poster is from Andrzej Zielezinski. It was shown at the twentieth annual meeting of the RNA Society last year. Click to enlarge!


This poster feels very contemporary and in tune with the times. The style is very close to �flat design� seen a lot on the web: clean, primary colours, sans serif type, very little shading. (Indeed, the website mentioned on the poster has a similar aesthetic.) I love how the core of the poster (the intro, methods, and results) looks.

The abstract is problematic. At a distance or shrunk down, that big rectangle in the upper left just dominates the poster�s visuals. It draws you in, and give you... blocks of text as a reward.

I would have tried to lighten up that block so it isn�t so visually dominant. In this quick and dirty redo, I�ve made the text that nice green, for emphasis, but put the box into a lighter, more neutral grey.



It�s not quite right, but I think the balance is a little better. The better solution would be to remove it entirely!

The grey stripes in the background are subtle enough that they are not overwhelming. Like the abstract, however, they might be lightened up around the edges f the poster a bit. The stripes are running at three different angles, too: the set running across the bottom is not lining up with the upper left. And if the stripes are going to radiate out from the center of the poster, maybe they should do that in all the corners.

The title bar is unusual: very few people right align their titles, because that�s not where we are trained to read. In this case, because you have that big abstract block in the upper left, having the title on the left too would have been far too much. Having space around the abstract block helps the overall look of the poster.

The title text feels a little light. Because it is set in a low contrast light green in a thin font, with a few grey stripes behind it, it might not be easily readable or noticeable from a distance.

I�m a little puzzled that a website link shows up in two places: under the authors� affiliations, and down in the bottom green bar. I would be tempted to have it in one place alone. My instinct would be to cut the top one, so I could make the title and author�s section a bit roomier, or maybe larger.

Similarly, I can�t quite figure out why two logos are sensibly corralled in the bottom, while one is taking a primo spot in the title bar. I�m guessing the one in the title is the institution and the others are funding agencies?

The genus and species names (Arabidopis thaliana) are not in italics anywhere. My reaction:



Jumat, 15 Januari 2016

Latest modern science | Critique: Sea turtles eating trash - Si Bejo Science

This week�s poster is like those �Spot the difference� cartoons that used to appear in the classified ad section of newspapers (dating myself there). Qamar Schuyler sent me a work in progress, so there are two versions. You can click either to enlarge!


There�s a lot to like. The sea turtle provides a clear cue as to what this poster is about. I wonder if a picture of a turtle ingesting debris might be an even better indicator of the poster�s topic. The trade-off could be that a poster of a turtle in trouble might be disheartening and a turn-off to a potential reader. Maybe the healthy, charismatic turtle used here is the right choice.

The main data, the maps, are up front and center. The big coloured map is placed just where it should be: right in the upper middle. The caption for it, though, is a little problematic, because it�s been severed from the image it describes.


In general, you want to place descriptive text as close to the image it�s linked to as possible.

A similar problem occurs with the smaller maps. While they don�t have to be read in any particular order, they do wind around, snake-like, between the colour map and the captions.


Part of the problem here is that five maps are the same size, and one � for Kemp�s Ridley sea turtle � is narrower. I would still try to put these in a more consistent two by three grid, and just suck up that the last one isn�t a perfect fit. Perhaps the figure caption could slot into the extra space, maybe like this:


Or this:


Of course, I�ve cheated in the sketches above because I haven�t relocated any of the text. Repositioning the figures would require a massive revision of the right side of the poster, perhaps moving the �Results� section into the upper right corner.

Here�s Qamar�s tweaked version. Spot the differences!


Some of the differences I caught (not intended to be an exhaustive list):

  • The box around the conclusions has been given a red border to �pop� the take home message. I like it.
  • A graph has been added to results. I like this, too. Visuals are better than words.
  • A poster number has been added. I�m very mildly against this, because I��m not sure it does much besides take up space. On the other hand, it is unobtrusive and might help someone.
  • The proportions have changed a little.
  • The �Contact me� box in the lower right has been tweaked a bit, and is better aligned with the box above it. I would like it more if it was the same width as the box above, though.

You can see this poster with Qamar at the Ocean Sciences meeting in New Orleans in February. If you can�t make it to the Big Easy, you can read the pre-print of the article here.

Reference

Schuyler QA, Wilcox C, Townsend KA, Wedemeyer-Strombe KR, Balazs G, van Sebille E, Hardesty, BD. 2015. Risk analysis reveals global hotspots for marine debris ingestion by sea turtles. Global Change Biology: in press. http://dx.doi.org10.1111/gcb.13078

Kamis, 07 Januari 2016

Latest modern science | The view from the floor of SICB 2016 - Si Bejo Science

 
It�s been a while since I�ve been to a conference, but this week I was at the Society for Integrative and Comparative Biology in Portland. These were things that I noticed while looking at the posters.

Fabric posters are still a minority, but I think you can always count on seeing a few. I finally saw a fabric poster made by Spoonflower. I�ve blogged about this service, but hadn�t seen one �in the wild,� so to speak. The presenter was generally happy with how it looked, although was putting in quite a bit of effort to make it hang right. It is a very stretchy fabric, almost like spandex, so tends to sag. If you are going to have a fabric poster, remember to iron it before bringing it to the session.

I ran across multiple posters that tried to say something about differences that were not statistically significant. I read text like, �The experimental group was slightly higher than the control (p = 0.07).� No! If the difference is not significant, saying anything more about the relative values of the averages is meaningless. Because if the difference is not statistically significant, you are saying that difference is due to chance, which mean that the difference you are describing could just have easily been in the opposite direction.

I referred multiple people to this blog post, �Still not significant.�

Too many titles were hard to read from a distance. The poster sessions are busy, with a lot of browsers, so your title should be visible from the moon.

I bugged many presenters about their error bars. Most posters I saw had at least one bar graph with error bars, and about 80-90% of those had no indication anywhere on the poster of whether the bars were standard deviation, standard error, or something else. This matters a lot for interpretation.

Update, 8 January 2016: My efforts to make a graphic for this post backfired. I�m leaving the image here, but several people busted me on an insufficiently nuanced quote about p-values. I�ll pick this blog post from Scientist Sees Squirrel for further discussion.

While the image here could be better, I think the larger point still makes sense: if your model says your results are probably due to chance (however you set that model up), describing experimental conditions as larger or smaller doesn�t make sense.