 Are there any announcements anyone has? I don't have any. We're just going to then move on into the next talk. We're happy to have Professor Mike Schatz from Georgia Tech to talk to us about scientific writing. OK. Take it away, Mike. Thanks, Mark. Actually, I'll take this opportunity to make some announcements. So this afternoon and then for the next two afternoons, we'll be having our sessions on professional development. And toward that end, we make assignments of people in this room mapped to session leaders on particular topics associated with communication. So for today, these are the assignments. So you should look up here and see if you find your name. I have one that you don't see your name, so we'll fix that. But you should find your name. And if you don't see it here, see me. And I will make an assignment after my talk. But it's important because this is how in preparation, how each of you submitted a poster. We're going to start this afternoon in tutorial sessions where you will get in small groups with your peers and a session leader and work on your poster snapshot or writing. So everyone has submitted some material that is a poster in electronic form. And that's the basis on which all these elements will be discussed. So look for your name here. And then after the end of the talk, I'll go over and just point out whether you're, again, we're going to be some of you stay here. Some of us will go up the hill to the M Lab but focusing on these communication tutorials. Any questions about this? How many other people don't see their name up here? Do not see your name up here. OK, so we just have one person. OK, so I am not Harry Swinney. However, the slides I'm going to present today have been put together. It was a talk based on a talk delivered by Professor Harry Swinney, who is one of the founders, the founding directors of these hands-on schools and as well as been the advisor and mentor to many of the session leaders. And so he has, through the years, in addition to being an outstanding scientist, he is, in my view, and I'm sure a number of the session leaders who know him, one of the best communicators of science, in particular of written science of scientific communication and writing that we know of. So he has put together a series of tips. And I'm going to step through that and point out some of the things that I learned from Harry when I was his student. And it's really a good advice about writing. The first thing that I want to point out is that there's no one way of communicating writing papers. Writing any form of communication is as much art as science. And there are several paths to being effective at communicating. That said, there are some tried and true methods. Each of us, when we read a paper or listen to a scientific communication, we have a sense of whether that was effective, whether we've learned something. When we're on the other side in producing that product, writing the paper, it's sometimes harder for us to see, well, what is it about that communication let's say writing a scientific paper, which makes it effective. And we're going to go through some of those ideas based on tips that Harry has put together through the years. So first, let's talk about the issue of writing. And this won't advance here. And let's say you have published an article. You put it in a scientific journal that's pretty well known, one of the scientific journals of a professional society or one of the well-known commercial publishers, commercial publishers. What happens to that article? So this is information. This is information that has been gleaned based on when people are searching for papers online or they are starting to read a paper. So it's based on what's called clickstream, the kind of online interactions with various ways in which we access publications. And basically, the numbers, roughly speaking, fall as follows. That within five years of publishing an article, typically, through some search or some keyword search that there are ordered 10 to the 4,000 researchers who actually access that article. And either getting the keyword or have done a title search and have brought up, let's say, it's your paper published in a well-known journal. It's been out there for five years. A board of 10,000 hits along those lines. And that's a pretty large number. But that doesn't mean that they have actually read your paper. So of those 10,000, probably about 10% or so, actually make the next step. They look at this result of that search, and they say, OK, let me see if this is something I want to continue reading. So they access or read part of the abstract. Now, that doesn't mean they actually read the paper. There's the next level down of those number of people. Probably of order 100 or so will actually read some aspect of your paper, not necessarily cover to cover, but actually will go next level down. And from there, if you're fortunate, you're going to get something of order 10 citations for the paper. Now, what is it about this process? This is a lot of winnowing down that happens. And the basic fact of the matter is that people who are reading a paper will find plenty or who start down this trail will find plenty of reasons for not continuing to delve more deeply into that paper. And I would say the main point of this talk is to try to design your scientific paper in such a way that you don't give the person who is going down this path a reason to stop reading. That is, you have to try to think of ways in which you can perhaps boost this number that you work very hard on your scientific research. You think it's great stuff. And you want other people to read it. That piece, that bridge between your work and you're trying to get that word about that work out there is a very crucial step, a bridge. And there are ways in which basically when the reader is going through this, there may be things in the paper that just cause them to stop reading. Some things you have no control over, but there are things that we want to make sure we go through and try to make sure that we are not the barrier. We don't put up obstacles to the reader such that they will stop at some point and don't proceed further and don't actually read our scientific research. So that's the key idea is we want to try to find ways to avoid giving the reader reasons to stop reading. Now, so let's start with looking at the paper title. So this is among the first things that a reader will look at. What are some key ideas? First of all, be descriptive. So you have a particular scientific topic. You want to make sure you carefully think about your title in a way which communicates accurately the science that you're trying to convey, but at the same time doesn't obscure it. What are the ways that you can obscure it? Well, first of all, one thing that's really important to do is think about who your audience is. So you try to write, and the general tip is try to write for as broad an audience as possible. Now, that doesn't mean you're trying to dumb down your research, but really the idea is you have a particular journal in which you are submitting your article, the journal of nature, or perhaps physical review letters, or a specific journal. There is a scientific audience. Among that audience, what you're trying to do is you want to try to place or describe in the title your work in a way which conveys it very clearly to the broadest possible audience in that venue in which you are presenting your paper. So what are some ways to do that? First and foremost, I'll skip down here, is really avoid jargon, avoid acronyms. We had a discussion this morning about different ways of doing particles and finding particles and tracking particles, a common acronym that was used, PIV, which Mark defined. But if your article describes that, and you have an audience that may know what PIV is, particle image velocity symmetry, but you have a broader audience, say, you're trying to write this article for nature, then probably, if it's key, then you should describe that and put that in the title, but it should be explained. So almost always avoid use of acronyms because they are basically the language of specialists. So you try to, to the extent possible, do that and avoid specialized terms because every time you put it in the term which is not known widely, you narrow the audience, you give the reader a reason to stop reading your paper. And the other thing is, in the title, be succinct. Sometimes you see these very long titles you read through and by the time you finish you say, I don't have any idea what this paper's about, I'm gonna stop reading it there, put people to sleep and just reading the title. So you wanna be succinct, try to be short and snappy with your title. Again, a key take home lesson in all of this is write and rewrite and rewrite. Take your paper, even elements of your paper and show it to your colleagues and circulate these as you write to get feedback. And the more feedback you get, the better you can hone all aspects of your paper. But being short and succinct is really an important step, particularly for the title. Now, imagine that the reader is going through your paper or imagine you are the reader and you're going through someone else's paper. Question for you, in what order do you read the paper? So the topics here, the structure of the paper is more or less, I would say, listed in sequential form here. There's the title, there's abstract, the introduction, the background, methods, results, discussion, conclusions, that's sort of the sequence. Figures are throughout, references at the end. But think about when you are going through and reading a paper, what are the first, let's say maybe you counter the title, what are the first few of this list? Let's say, what are the first one, two, or three things that you actually look at when you're in the process of looking at paper? So take a moment to think about that. Okay, you wanna go ahead and give a. Okay, so we have, tell me your name? Yes. So Junius does abstract, right? Title, abstract, then jumps to the conclusions. That's an interesting sequence. Notice, not in the sequence laid out here. Anybody else have some, perhaps the same or different? Abstract and then you go to the results, okay? Anybody else? Abstract, methods, conclusions. Look at the pictures, okay? Title, go to the figures. Abstract in what? Figures. Abstract in figures. I'm trying to. Okay, okay, what's the point here? Okay, now when we're writing the paper, we may have in our mind, right, so you are acting as readers, reading somebody else's paper. Think about what you have in mind when you're writing your paper. Perhaps you have in mind, and a lot of writers have this in mind, is that they think that the reader's going to start with a title and work their way all the way through, and well, let's say they go from the abstract to the conclusions. On the conclusions, you've defined a bunch of terms, right? But that's okay because they started with the beginning and those terms were defined earlier in the paper, and you're going to assume that when they get to the conclusions, they've already understood the terms you've defined earlier in the paper. That's going to fail, right? It's going to fail because different people have different ways of reading the paper. The punchline is, other readers are like you. They skip around, right? So a paper that's meant to be read sequentially, start from the top, go through the end, won't make sense to a lot of people. And what is that going to do? It's going to cause people to stop reading, right? So this is something you want to avoid doing. Well, how do you avoid that, right? You can't control how people work through that. Well, one sort of very general tip is to try to make each major element as self-contained as possible. Now, it's not going to be possible that you're going to get the entire paper, let's say in the abstract, or the entire paper in the figures. But if you have in mind that other people are going to read the way you read, then if you try to the extent possible in each of these, particularly these main elements like the abstracts or the figures, we heard a number of people say, after title, I do abstract. After title, I do figures. If you think about trying to tell the story, and I'll say something about that, when you're writing your paper, it's all about storytelling, right? We as humans tell stories. We as scientists who are humans tell stories. It's all about trying to tell your story in as effective a manner as possible. So one key to doing that is try to build that story to the extent possible in these elements which people will access first, because if they read that part, they get to the abstract. You've told that in an effective way and got the main point across. They'll want to keep reading in general. That helps increase the odds that they will continue to delve more deeply into your paper. So that's a really key fundamental idea here. So let's talk about that in the context of the abstract. What should it contain? So this is one of the main things that the sessions that we're gonna go over this afternoon in communication, and that is in the writing session, the focus will be on writing in abstract. Now it's going to be the abstract for your poster, but the same ideas apply, right? The same sort of guidelines. So what is the length? You have to keep it short. Those typically in journals, they will tell you what your length limitations are, but really you want to keep it short doing a poster, but it should have very key elements, okay? So what is the problem you're studying, right? You should state that very clearly up front. What's the key idea, okay? Why should the reader care, right? Why should they care about that problem you spent five years working on, right? You want to communicate that. What did you do? What was your take? Experiment, simulations, what is the particular angle that you, angle of attack that you had on the particular problem, okay? Really importantly, okay, what's new? What did you bring to this scientific discipline, that problem, what did you find in that case? And why is it what you found? Why is it interesting? Once again, why is the problem interesting and why is what you accomplished or found or discovered interesting? So you make sure you want to communicate that. And then what does it all mean, okay? What does it mean going backwards in the past? How does it tie to what has happened before? How does it build on or perhaps correct what has come before, okay? And what does it mean going forward? This result implies these are directions to go, okay? It's a lot to do, but these are the kinds of ideas you should have in mind. That encapsulates the main story. This really will help keep people reading, right? If you try to embody in your abstract these really important points. Yes? Yes. Okay, so as I was saying before, you can't assume that people are going to, so you have in mind, I don't want to put in the abstract. I'll talk about it in the conclusions. They may never read your conclusions. You heard people read first, they're gonna read the title. Then perhaps the abstracts may be the figures. That's why you try to make these major elements, particularly the ones that come first in many people's minds, try to make them as self-contained as possible to communicate the main message. You might think, well, I'll just repeat myself in the conclusions. It won't hurt. Okay, so this is another, you spent five years on this problem, right? And you think if I say it more than once to someone else who has never heard it before, they will be bored. Quite the opposite. You need to say it many, many times to somebody who has never heard your story before so that they can begin to understand it, okay? So don't feel afraid about repeating yourself. In fact, you should, in a, again, artful way, communicate that main message in these other elements. Now again, each of these components of a paper can't communicate the whole message, but the idea is these first main points of these main elements play a huge role in determining whether readers are gonna continue delving more deeply. So you want to help them do that, not throw up barriers to that, okay? So it's a very good question. Okay, let's talk about figures. So I learned from Harry that even before you start writing a paper, you have a story to tell. Where do you start telling that story? You start with the figures. And so basically, you decide what those figures are going to be. And in fact, many of us who have worked with him, Matthias is nodding his head. You go over and over and over until those figures and captions tell the story. Because that's another key piece, okay? So again, the totality of all figures, you try to tell the whole story of the paper. Within each figure, there's a chapter of the story. You try to make it as self-contained as possible. Do not require the reader to jump to your text to then understand what's going on in the figure, right? Because that's another way to give them a reason to stop reading, okay? So you try to make this a self-contained story. If they're difficult to understand, they will move on to somebody else's paper, okay? Now, here's the key thing. Figures ideally communicate a message even without you needing a caption. Do we need a caption for understanding what's going on in this figure? You know, our brains work as Mark was saying. Our brains are tremendous at doing image processing and actually interpreting meaning from images. They've had millions upon millions of years of practicing that or survival, depending on looking at images and making meaning from those images, right? So we should take advantage of that. Figures build on that. And so you really try to design your figures so that the figures themselves can convey a message because there's a lot there that the brain can process if the figures are crafted in a way to communicate that effectively, okay? So that's a key point to remember. Now, there's a famous book which is titled The Visual Display of Quantitative Information by a man named Edward Tufte. This guy makes his living going around and telling everyone how to create effective figures and I have to say, he's very good at it. This book which has been around for decades is really excellent, right? And there are some, this doesn't encapsulate all the main ideas and we're gonna go through some, he has a number of case studies which are really useful to go through but here's some principles you can garner from that. When you create a figure, make sure you maximize the ratio of the data ink to all the other ink, right? And so we'll look at some examples and see what that means but I think of this really more in the context of this doesn't apply just to papers. Let's say they also applies to PowerPoint presentations. How many times do you see people take and have all sorts of fancy decorations and such on the slides and then all sorts of font colors and different colors? That's gratuitous use of graphics, ink, digital ink which does not contribute to communicating the information. So you notice these slides from Harry Swinney, his philosophy, very simple background and here is the ink you need to communicate the message. So this general principle maximized ratio of data ink to other ink is a universal. Plies to posters, if you're in my poster session I'm going to be hard on gratuitous use of color because I find it and I think others may find it just distracting. You want the ink that is there to be focused on communicating the information you want to get across. Again, there is an artistic component differ, people differ on this philosophy but this is a principle that you'd be well advised to have in mind for all sorts of visual communication. Now when you have graphs you try to make those as simple as possible. That's sort of as part of this but in some sense one example would be don't overload your graphs with information. Don't put 12 curves on with dot dash and dash and blue and green and red. You should think very carefully about the amount of information you need so that it will convey the main point as effectively in as simple a manner as possible. And here's one that he's big on and that is don't use legends. Don't force your reader to look at the small box in your figure with the 12 symbols and then go back to see which that symbol is and back and forth. Instead, if you have different curves that mean something differently, label the curves themselves, okay? So you don't need to force the reader to jump back and forth. Again, all these things are in service of trying to keep the reader from having an excuse to stop. So let's look at a couple of examples. All right, from a paper of Linus Pauling, double Nobel Prize winner, right? Lots of problems with this graph, okay? Question is, how can this graph be improved? So what I want you to do is think about it for just a few seconds. Dosh, before we're gonna start, just a minute. So we're practicing something which is called active learning and it'll be peer instruction and we'll say more about that in science education next week. So I want you to individually think about in your own mind what are ways in which you would improve this. So let's take like 30 seconds or so. How many different ways can you improve this graph? Okay, now what I want you to do is turn to your neighbor, talk to your neighbor and share and see if you have the same ideas or maybe come up with new ideas. How to improve this? There are at least six or seven or eight ways this could be improved. Okay, let's, okay, you have a chance to discuss. Okay, I will mention incidentally when you use this kind of method in your lectures to pose questions to your students, to your lecturing to, this is very effective. You see the activity because then you start talking and getting involved. This is an example of active learning. So this is a really important thing. We'll talk more about that next week in science education. Okay, let's have some, someone volunteer, some number of people volunteer. How can you improve this graph? Let's each, let's say someone raised their hand and name away this could be improved. No improvements, okay. Yes. We discussed that the crosses are unnecessary. Okay, what the heck are these crosses doing there? I mean, that is gratuitous ink, right? That is not data ink. That's other ink. Get rid of them. Okay, that's a good point. Another, another suggestion. Okay, it is true that perhaps color would work here. Maybe not in 1947, but, but perhaps one could use color to make this more effective. Other suggestions. Yes. Sorry? Missing a legend. Missing a legend. Why does it need a legend? But I think what you're driving at is, what the heck are these curves? You know, why are there curves like that? What is, what's the meaning? There's something that you don't understand about, there's this line up broke and then there's another one. What's going on here? So it's true. There's something about those curves which, you know, you're trying to like, why are they drawn that way? Right? We'll talk about maybe some of us can figure out what this is and you'll see it very clearly, but that needs to be explicated. There is information that your eye is drawn to and you have no clue what it means and you're going to be frustrated by that. I'll just choose somebody else. Another suggestion. Yes? That's right, that's right. So to do more explanation about that, someone else, so we have only, we had a lot of discussion. Session leaders can chime in too. Yes? It's not at the middle of Y axis. Sorry? The label of Y axis is not at the middle of the. Okay, so the label, okay. So there's one thing, it could be elevated. I'll tell you a problem I have with this. Why not rotate that? Atomic volume, right? Why not rotate it? Now again, sometimes you can't do it. Space limitations. There's space to do that. Rotate it so you can read it. Bruce. Rotate the labels for sure. Rotate the labels, the same thing on the vertical axis. Right? What else? Yes? Go ahead, and then I'll take you. Yes? There's no units. Where are the units? What does this mean? Atomic volume, in what units? If you're a chemist, you might know, right? We know this has something to do with like the periodic table or something, but you know, what's atomic volume? Maybe I had it in chemistry, freshman chemistry 30 years ago. No idea what the units are. Yes? Sorry? Title, perhaps, yes? It doesn't have a caption. Okay, well, you can see there are lots of things. Let's look at what Tufti kind of suggestions Tufti made. Okay, so a lot of them what you said was actually have been incorporated. Identifying what these curves are. So what you see here actually, these are the alkaline metals, right? And then now with that in there, you start to say, okay, atomic number, ah, maybe what this curve means is that you are in a particular, okay, well, I forgot. This is group, or is this group? Remind me, it's this particular, the horizontal part of the periodic table. Chemists back here. It's the rows. The rows, okay, the rows. This is an official chemist called the rows. These are the rows on the periodic table, right? And then you add, when you add a shell, right? You have this jump in the volume. Okay, so that, now, even without having labeled that with the different, the row number, shell number, basically you can see that from these alkaline metals. So that gives you more information. Couple of other things. So this is the issue of rotating the label. Notice these labels not just only rotated, but fewer of them. You don't need all these labels for the axis, right? We can determine between 20 and 40 is 30, right? Reduce that amount of ink. Another thing that Tufti's big on is eliminating these boxes. He doesn't like this kind of thing where you're doing the top and the side. That's sort of a matter of taste, I think. No units here, Tufti missed that one, right? Should have units. Okay, so these are things, again, thinking about when you're communicating, making your plots tell a story and do that in a more effective way. Let's look at another case study. This is, in Tufti's view, like the worst graph ever, right? Let's take a look at that thing. What the heck is it about? Just look at that for a second. That is quite a monstrosity. Okay, even the title. Age structure of college enrollment. I mean, how obscure jargon-y can you be? Okay, so you know what? There are five data points here. They're plotting five data points. Okay, let's look at what Tufti has to say about what's bad about this graph. There are five data points. We'll show what those are. So what this is showing is basically percent of college students over a certain, let's say, 25 and over for a few years. For five consecutive years. That's what it's showing. So why do you need the 3D business, right? Why do you need the color again? This is like, way, doesn't do anything. This is the same information as that. 100 minus this produces that, you know? This percentage. Can't read the fonts. What the heck are these curves? All these curves connecting the points. Why is it curved? What's the reason for the break here, right? This is just like really a bizarre thing. So let's look at how can you communicate this? You could put the five points in a table or you could plot them, right? Okay, so you get the message, right? Figures are really important. You want to make it, your eye really is very effective at grabbing and getting and obtaining a lot of information all at once. Communicating a message, but you have to design the visual so it does that and doesn't get in the way of the readers getting the message, okay? So figures, tell the story. Okay, what else? Let's talk about the writing aspect. So, you know, brains have been developed for many, many years for image processing, but sort of interpreting symbols in the form of writing has not been around that long, right? A few thousand years. And in fact, most of the world didn't know how to interpret symbols in terms of writing until, you know, relatively recently, 100 to 200 years ago, right? 10% of the world's population was literate at around 1850. So I've certainly, that my family back, less than six generations, was illiterate. They came from, they were peasants and the Russian step and I should say in the Ukrainian step, they were farmers, right? I'm sure they were not literate back in the late 1800s. So this is very recent. What does that mean? That basically it takes more for your brain to process, to interpret the writing. So you have to really be careful about crafting the words to communicate clearly. Now, what are some really key things to get across? So we talked about, so those first major elements, abstract figures. Some things that should really make sure you put in the paper to make it scientific. The key thing, I would say, punchline, is that you make sure you put enough information in there so that someone else wants to replicate the results they can do it. So if they're going to do the experiment, okay? In order to do that, they're gonna have to dig more deeply but make sure that if someone chooses to do that, that information is there, okay? Parameter values for your experiments and ranges. Units, right? If there's sample preparation, try to describe that in as much detail that enables replication. Simulations, boundary conditions, initial conditions, right? Theoretical analysis. So these are the key things. If it's not enough information to do replication, it's not really, so you can write short letters. That's difficult to do but at least it should, a short paper should point to a larger paper ultimately than which this information can be found. Now, references. Okay, it's our civic duty to give, do where it is due, to acknowledge prior work and when you're building on the work of others. If you may not, you may have a personal, something personal against someone else who is working your field. They scooped you or they treated you badly perhaps and you sort of have it in for them. That is not a reason for omitting them from your reference list, right? So make sure it's an ethical, you're ethically bound as a scientist to really cite the work that you built on when you communicated or you were doing your science. So it's really important. There's a nice policy statement about this in detail at the American Physical Society that's worth reading. Okay, so we're gonna wrap up here. What's the most important thing to have in mind? You rewrite and rewrite and rewrite, okay? So I mentioned that starting with that idea of the process of let's say a title and figures and you go over those over and over and over again, right? Get others to look at those early stages, you know? Give feedback. You may not like what they say, but actually it hurts when people are saying that this looks really horrible, but it's really valuable. You should value that kind of feedback because it helps you make it get better, right? So if someone who spends the time to give you a critique that is sharp but is on point and you may not feel good, but it will help your science just get better. Read it out loud to others, to another person to see if it sounds good and cut out unessential material. Yes, Ken. I say, this sounds like crap and it has to be regretted and just keep doing it until you read it and it actually, you're not saying that it can work, it actually reads pretty well. Yeah. Yeah, that's good advice. All right. Yes, and then edit, try to make it succinct, cut out unnecessary material. Again, the mantra that I have always in mind is I don't wanna give the reader an excuse to stop reading. The person who's reading it will have plenty of excuses to stop reading, right? Don't give them extra excuses. So that's, I think, a good punchline to have in mind. So it is a really important part of doing science, right? Communicating your science. And in fact, the very process of writing your work up actually is, you might think, well, I'm communicating, not doing science. Actually, when you are writing, it is an excellent opportunity to sort of rethink very deeply about the work you've done and you'll find out many times you will make changes that are, oh, this doesn't sound right, I've missed something that impacts your scientific research. So it's, in fact, an opportunity to really think deeply about the science you've done and actually learn some new insights many times, often is not, okay? So those are those sort of tips. Again, there are tips that put together by a master, Harry Swinney. I am just a messenger, so I communicate those on to you and with that, I'll stop and thank you for your attention. I need questions or comments or anything.