Watch videos with subtitles in your language, upload your videos, create your own subtitles! Click here to learn more on "how to Dotsub"

BITC Publication Class 5: Figures

0 (0 Likes / 0 Dislikes)
OK, so we talked about figures as being one of the first things that you are going to develop in the process of writing your manuscript. And figures really can make or break a paper ... The old saying is "a picture is worth a thousand words." Well, a good scientific 'picture' expresses probably much more than a thousand words. So, we're going to take a bit of time to think about how to best develop (and note that I split an infinitive there!) how best to develop effective figures for your paper. So, first of all, it's a very good idea to use technology wisely ... when I began my career, we were drafting figures by hand, and you basically had to be a scientist AND an artist. Well, nowadays, you have many platforms on which to build very effective art to accompany your papers. Adobe Illustrator, Photoshop ... various programs that allow you to develop, design, edit, improve, and export your figures. So, one thing that I like to do is to save the original version that's IN the platform, so it might be Adobe Illustrator, or it might be ArcGIS, but save that original, so that if a reviewer asks you to change some part of a figure, you can very easily pull up exactly where you were when you finished the figure, and change that one detail, instead of having to redo the whole figure. And then you can export a jpeg or a png version to include with the manuscript for review. I find that to be a very, very time-efficient approach. So if a reviewer asks for modifications to figures, I go back exactly to where I left off, and I change that figure however the reviewer wanted it. Now, one very important consideration is color versus black and white. A BIG advantage of color is that the eye can take in more dimensions, essentially the eye can receive more information, and understand it. The downside, of course, is that color publishing generally costs a lot of money. Sometimes, it is 1000 dollars to publish a color figure with your paper. So, it has costs and benefits. I think that there are some things where color is supernumerary-- you don't need color to express what is in that figure, and then there are other situations where it is indispensable. And so use it wisely. One thing that you can do is use color illustrations in review, and then potentially--or not--substitute effective black-and-white figures in the publication stage. A solution that I like, and that I have been doing more and more, is to use color in the electronic publication version, and allow the hard-copy version to be published in black and white. More and more journals are allowing you to do this; that's an effective solution these days, especially given that most access to journals is via electronic versions now, so most people see the color version. However, if you're going to go back and forth between color and black and white, you really really need to print your color figure in black and white, and make sure that it is interpretable. Very frequently, when you can see something very clearly in color, you can't see it at all in black and white. I am going to give you an example of that in the next slide. So here's that example: this is a map of Luzon Island in the Philippines.. There's a lot that is nice about this figure: it has a very clearly visible legend, I can see the symbols very clearly, the labels are readable, I can tell the difference between islands and geographic features ... there's a lot that is good about this. Notice, however, that what I am supposed to be differentiating is this lowland-to-highland ramp (essentially it's elevation) from this green shading, which is the predicted distribution using the ecological niche model. Over here, in color, it's pretty clear. But look here When I print it in black and white, now, that area that was predicted by the niche model looks like highland area--OK, it's just very misleading to your eye. Another, more minor quibble that I would have is the one city that is labeled on this map has a symbol that looks a lot like an occurrence record of the species. That's a detail: for me, I would use a very different symbol to label that city. So one very effective thing that you can use to make these decisions about color is this website, ColorBrewer2.org. It's essentially an online tool for selecting color schemes for maps. And it can be very, very useful in avoiding problems like this: here, I've got a map with three types of shading, but then it has these two gaps. One of them is very very obvious, and this one is kind of hard to see. So, you can go to this website; you can give it the basic specifications about are you doing a presentation, are you doing a publication, is it in black and white, is it in color, and this website will help you, essentially, make the optimal choices for which colors to use in which situations. So even when we're talking about black and white figures, we can have effective presentations and ineffective presentations. I'll give you some examples here. Here, in this one that I've labeled 'bad,' notice that I have to do a lot of looking to understand this; for example, here, this is number of reef fish landings, but notice that I don't have a label down here. I have to look over here ... Worse yet, these three panels refer to something that is different, but I really have to go to the figure legend to be able to understand what they refer to ... I greatly prefer putting very brief and informative labels right on the panels. Now, this is a matter of taste, and notice that this is developed as kind of a matrix: These rows are "tons" and "tons" and "square kilometers," and these columns are "year" and "year" and "year," but each one is understandable. My basic priority in this is that you develop figures that are immediately understandable even to the somewhat tired reader. So, now, what I am going to do is to take you through some figures from recent manuscripts that colleagues and I have worked on and essentially I want to show you the sort of editing and fault-finding that I hope you can do, as part of your process of developing manuscripts. So this was a very early draft; I didn't like it at all, obviously-- this is basically just somebody putting up maps, and maybe a diagram of an insect, putting some text on there ... this is just a draft this is a mockup. So, here's another figure ... it has some things that I like: For example, this label refers to all of these panels, and this label refers to all of these panels, and then we have labels that refer to these [rows], but I am going to quibble with what's this axis and essentially can I understand this figure without a lot of reading. Another problem that I would point you to is all of this white space. I would wonder if there isn't a way to make the actual graphs a bit bigger, so that the eye does not have to work so hard to see the details. So here's a good example of a common mistake ... notice that we have five different lines, and certainly they are distinguishable, but not easily, and so I'm having trouble ... I am looking maybe for this 50 km line, I am having trouble deciding whether it is this one or this one. So you need to really step back from your figure ... you may have composed it with all of the care, and all of the detail in mind, but if it's not easily readable, even if somebody has a small computer screen or a bad printer, it has to be immediately and easily readable and interpretable. So, just to sum up, figures need to be as visual as possible; they need to really tell a story. They need to be presented so that they use the space on the page optimally. You may even have to redesign figures for journals that have different sizes of presentation. But, your figures are going to go at the end of the manuscript, and not [be] set into the text, there should be a single page that is figure legends ... so here's the legend for Figure 1, here's the legend for Figure 2 ... and then you present one figure per page. Figure 1, Figure 2, Figure 3, etc. And essentially that allows the reviewers and the editors to think about your figures very clearly without having the figures set into the text as if you were typesetting the paper. So save yourself all that extra work, that goes into putting those figures in the text. So that wraps up our consideration of figures; in the next piece, we'll talk about tables.

Video Details

Duration: 11 minutes and 6 seconds
Country: United States
Language: English
Producer: A. Townsend Peterson
Director: A. Townsend Peterson
Views: 34
Posted by: townpeterson on Dec 24, 2012

for subtitling and translation

Caption and Translate

    Sign In/Register for Dotsub to translate this video.