Communication skills


Chuck Doswell

This document was created to provide some thoughts that I hope will be helpful to those seeking to communicate scientific ideas. I consider it to be copyrighted material [© 1998 C. Doswell], so if you want to reproduce it in any way, electronic or otherwise, please contact me.

Many science and engineering majors have neglected their communication skills during their education. If this is true for you, it can be a real problem because a lot of what you will be doing will involve communicating what you have done to others. In fact, your ability to accomplish that communication effectively can be crucial to your success on the job; promotions, allocation of resources, bonuses, etc. all can be strongly dependent on how well you communicate. Don't let this slip by in your enthusiasm for your technical subject matter.


1. Writing

For scientists, their success is judged mostly by their publications. If you can't write so other people can know and understand what you have done, you will have a pretty rough time getting papers published in refereed journals. Writing your thesis and dissertation can be a good chance to learn what it takes to become an effective author, but developing the skills to write good scientific papers is not a matter of one or two tries. Be prepared to learn about writing skills for the rest of your career.

An excellent path to developing your writing skills is to read a lot of papers in your field. Often, contributing authors are asked to review papers within their domain of expertise. Take this job of refereeing seriously! As I've noted already, it is easier to see certain sorts of mistakes in someone else's paper than it is to see in yours, so if you give an author a hard time for doing something, be sure that you aren't guilty of the same sin. This is not the venue for a discussion of reviewing papers, but one benefit is that if you do it conscientiously, it almost certainly will make you a better author.

Clearly, writing takes practice, so you should plan on reporting on your work as often as appropriate. You can't learn a lot about writing just by reading about how to do it. You have to participate, and everything I've ever heard or read about good writing says that you will be doing a lot of revisions . Go ahead and plan for revisions, and don't try to write a perfect paper before you share it with someone else. This means you should pass the paper on to someone locally whose judgment you trust and who is willing to give it a careful, critical reading. Develop a thick skin when it comes to criticism of your work! A severe critic is your best friend in learning how to write well; that critic will force you to clarity of expression and will seek and find any weaknesses in your presentation. Do you really want to go to press with a weak presentation, or flaws in your reasoning, or misleading expressions of what you did, or poorly-executed figures? Better to find these and fix them in the review process than to have them pointed out to you after the paper has appeared for all in your field to see! Too many young writers (and even some experienced ones) shy away from criticism. Don't let this happen to you, and you will become an improved author every time.

Knowing when to stop the revision process before submitting your paper to review is going to take some experience. In general, at some point in the process, you will no longer be able to tell if the paper is really getting better with each revision. This may be so simply because you are thoroughly sick of the paper! This may be a signal that you need to submit your paper to some sort of review (perhaps among your colleagues or perhaps through the formal review process), and waiting for the reviews to come back is a good time to tackle another project, to get your mind off the paper. Then, when the reviews come back, you will be refreshed and ready to finish the last round of revisions in response to the comments you have received.

Good papers that you read in the journals (Yes, there are good papers out there!) are not the result of a single round of writing by the author(s). Instead, they have been subjected to many revisions after many people have done careful reviews of the content and presentation that paper contains. Very few papers appear after having sailed through the review process without changes, so get used to the idea. It helps if you can at least create a good starting point with the first pass, so learning what works and what doesn't work in your technical writing is probably worth your putting in some effort. You should get better at the process, with time and experience.

Avoid certain pitfalls! There are several mistakes made by new scientific authors (and some old veterans, too!). One is expositions demonstrating the obvious. If you did something that confirmed that some obvious answer did not work, that might deserve a short mention, but avoid belaboring such points. A scientific paper need not describe in detail everything you did; only some parts of the total work effort end up being pertinent. By the same token, be careful not to gloss over some point that needs to be explained carefully. There is a fine line to be walked between these two extremes and for most of us, experience is the way we learn how to find the proper balance between detail and brevity. It often comes down to a matter of focus: what are the main points you are trying to make? The following are some other common problems in writing.


a. Figure description

A well-conceived figure does not need to be described in the text. If you write sentences like "Figure 21 shows a graph of chinkadera frequency as a function of time." then you are simply repeating the content of the figure caption. This serves no obvious need, since the reader certainly can read the caption. This is just deadwood. Instead, you should cite the figure as you make the point: "Chinkaderas decreased monotonically with time (Fig. 21), showing that omphatel production had ceased."

Further, you may find yourself saying something like "As seen in Fig. 17a, a boundary passing through central Illinois separates the region of strong knutenary concentrations in northern Illinois from weak concentrations in southern Illinois." If your figure shows knutenary concentrations directly, then isn't this simply describing what the readers can see for themselves? Is there some point to be made by the figure? If so, make that point and cite the figure to illustrate the point, saying something like "The knutenary concentrations in Illinois (Fig. 17a) indicate that the gorphatz process is dominating the situation." You do not need to fill the text with descriptions of what the figures show; doing this amounts to creating another form of deadwood. This discussion also applies to tables as well as figures.


b. Too many figures

When you find your paper contains a lot of figure description and relatively little else, then this can be a symptom of having too many figures. This often springs from a desire to show everything you did. In general, it is not necessary to document everything; you need to show just enough to validate your interpretations. Opinions differ in this regard, but every figure ought to serve a purpose in the paper, not simply to document that you considered each item that each figure represents. Figures require captions, they take extra work to create, and (in my opinion) ought to be pared to the bare minimum to get the point of your paper across. Figures also increase the cost of your paper in several different ways, including page charges for publication, if page charges are indeed required. Should your formal referees and/or internal reviewers ask for more figures, fine. Go ahead and include what they asked for. If they ask for them, it is a good sign that you have been diligent in attempting to minimize the number of illustrations.


c. Too many or too few references

Some authors feel obligated to show how well-read they are, and so their papers read like a literature search. This attempt at being "scholarly" often goes too far, with so many references, there seems to be a bare minimum of original thought in the whole paper. On the other hand, it is quite possible for there to be too few references, thus giving the impression that the author has been working in a vacuum in the literature. This is just as inappropriate, if not more so, as having too many references. On the whole, I prefer too many to too few but it is best to try and strike some happy medium. As always, the editors and referees have the final say on this topic, of course.


d. No focus

Many authors seem to have trouble deciding on what is the main thrust of their research and sticking with it. If you have kept a research journal (or diary), you may be especially prone to this, because you will have a record of everything you did. The object in writing a good, terse paper is to find the main point(s) and stick to it (or them), without a lot of wandering around. If you use the method of outlining to organize your paper and can stick to the basic outline once it's finalized, it should be easy to stay with the point. If you start wandering away from your outline very far, you need to think through whether or not the wanderings are really relevant to your paper. Those off-the-point thoughts may belong somewhere else.


e. Redundancy

It's been said that you have to repeat something eight times before you can be reasonably sure your audience will remember it. Unfortunately, some people carry this too far, by repeating themselves in the Abstract, the Introduction, the text itself, and the Summary. If you are going to say over and over what it is you did and what you found, at least confine the lengthy expositions to the text proper and keep the summaries at the beginning and end very short! It gets tiresome to read the same things over and over again, so most readers will start skimming instead of reading with care at about the second or third repetition. This may cause them to miss important points buried in the midst of all that redundancy.


f. Acknowledgments

I offer a simple word to the wise here, paraphrased from my own graduate advisor's good advice. If you are uncertain about how much credit to give to someone who helped you out along the way, always err on the side of giving more credit than you are certain is due to that person. Think about it. If you helped someone and they never even acknowledged that help, how likely would it be that you would help them again? You can offend someone in this way and they might never say anything about it, but you might wonder why they are always too busy to help you out when you need it. It's possible to make an enemy for life in this way and not even know what you've done. Obviously, it's unwise to burn bridges this way. It doesn't cost you much to be grateful, and it can pay off in big ways later. Of course, you should want to be grateful to those who have helped you because it is the right thing to do, not because of the benefits you are going to reap. But if you only do it for the benefits, it still works.

Offering co-authorship is the highest form of acknowledgment, with inclusion in the "Acknowledgments" section as the next form down the scale. Clearly, referencing your sources is also a form of acknowledgment that is important. Where someone's contributions fit in this order is up to you, of course, but keep in mind the basic rule (above). If you offer co-authorship and the person chooses not to accept, they will not be offended by the question; if they feel their contribution deserves co-authorship and you choose to offer a lesser acknowledgment, how "collegial" are they likely to be in the future?


2. Oral presentations

As with writing skills, it is likely that you are going to have a continuing real need for making effective oral presentations. If you have time in your course schedule and you have not ever taken a speech course, this might be a good time to find a spot for a speech course in your schedule. The main idea is to get some practice at speaking in "public." If you can get this by joining a local Toastmasters group (or some other way), then this will work as well. The main idea is to learn by doing, hopefully with some feedback from people who are knowledgeable and critical (and, yes, here's another place where a thick skin is helpful). I make no pretense of being a professional, but there are many pitfalls to avoid and tips I can offer to help you learn how to make oral presentations.

a. Learn to omit "filler" from your speaking

Many of us put "filler" in our speech unconsciously. When we are pausing to gather our thoughts for the next burst of speaking, we often use such fillers as "y'know," "like,", "uhhh," "well," and so forth. In order to hear what you do, have some of your discussion within a small group recorded, preferably when you are unaware of it. Even when you are aware of it, the playback of your speaking still can surprise you. A simple device for avoiding this filler is simply to pause and concentrate on saying nothing during that pause. Perhaps we do this in our conversations to prevent others from leaping into our pauses and saying what they want to say. This habit is inappropriate in a presentation where we are not competing with others for talking time, and ends up being a distraction. Not every moment of your presentation needs to be filled with your voice!

b. Maintaining eye contact with your audience

As you speak, it is useful to let your gaze roam about the room and make eye contact with individuals in the audience. This gives them the feeling that you are making the speech directly to them. If the group is small enough, you can do this with everyone in the audience at one time or another. In a large audience, this may not be possible and people in the back of a large room may not ever be able to see your eyes, but you still should look at individuals in the audience, if for no other reason than to keep yourself aware of talking to people, not the air. Don't talk only to the overhead projector or screen, please.

c. Limit the amount of material covered

Although there are no hard and fast rules, you cannot make many points in a short time. I use as a good "rule of thumb" that I can only make about one point of substance for each five minutes of presentation time. Therefore, in a 15 min. talk, I try to make no more than three major points. Plan your talk accordingly and you should have no problem finishing your presentation in the allotted time. By the way, I have never heard anyone complain when I finish early, on occasion!


d. Limit the number of visual aids you use

Employing a visual aid, such as a figure, or a table, or a photograph takes time. Each such visual aid usually involves at least a brief explanation and usually requires that you take some action (unless you have someone switching viewgraphs or slides or whatever for you at just the right times with no prompting on your part). I use the rule of thumb that I should have no more (and preferably, fewer) visual aids than there are minutes in my presentation. In a 15 min. talk, there should be no more than 15 visual aids. And if the visuals are multi-media (e.g., both slides and viewgraphs) then each change from one medium to another is allotted the same time as a visual aid. Thus, if I have 10 visual aids, and I make 2 transitions back and forth from slides to viewgraphs and back (so the talk goes from slides to viewgraphs to slides to viewgraphs to slides), that makes four transitions and I am just about at my limit for a 15 min. talk. If you have more visual aids than this, you will find yourself rushing through them so rapidly that most of the audience will get very little from any of them. A well-conceived visual aid still must be visible for about a minute to be certain that the audience has grasped its contents

e. Do not read your visual aids aloud

If you find yourself reading the contents of your visual aids to your audience, they will be able to read them silently much faster than you can read them aloud, so you will end up falling behind your audience. This is also a monumentally boring way to make a presentation. If I can read your viewgraphs, why do I need you to do that for me? A better strategy is to use your visual aids to prompt you to discuss items in a certain order and to avoid inadvertently leaving something out.

f. Make sure the content matches the presentation

Most of us have heard talks that seemed wonderful because the presenter was exciting, or dramatic, or funny, or whatever, and then when we had time to think about what was said, we discovered that the actual content of the talk was pretty shallow. And most of us have heard a speaker who had a lot of real substance to say but who was simply boring and clueless about how to make an effective talk. Your presentation should not overshadow what you are trying to say; rather, it should be focused on getting your points across. Being "cute" or making forced attempts at humor that are out of context (e.g., starting out with "A funny thing happened to me on the way to the Symposium ... " story) are good ways to alienate your audience. Technical presentations do not have to be boring, but neither should they be viewed as mere entertainment.

Keeping the knowledge level of your audience in mind is a common courtesy that few speakers seem to be able to do. If you are addressing a general audience, you shouldn't be solving nonlinear partial differential equations during your talk. If you are addressing a group of faculty members as part of a presentation of your research findings, you shouldn't include slides of your most recent family outing. This ought to be common sense, but it is disappointing how often speakers fail to take into account the type of audience.

g. Put some variety into your speaking

Most of us have heard speakers who drone on in a monotone, with little or no sense of feeling in their presentation. If it is true that you care about your work in some way, it is quite acceptable and even desirable to have some of that excitement show through in your talk. Vary the level of your speaking volume; at times you might be close to shouting, and then revert suddenly to a near-whisper. If your speaking is always at the same volume level, this encourages people's attention to drift. The purpose of this is not simply add drama to your talk ... that would be equivalent to putting entertainment ahead of content. The point is to get your audience's attention and keep it. If these "techniques" enable you to accomplish that, isn't it worth it?

h. Listen to other speakers

As with writing, it is a good practice to attend as many talks as possible. This allows you to evaluate what speakers do well and what speakers might do that you definitely want to avoid. Learn how to evaluate the presentation separately from the content of the presentation. Your impression of a talk is often influenced by the presentation quality more than the actual content. Hence, this skill is something that will obviously be helpful, and you need to learn how to distinguish these two independent aspects of a talk.

i. Dealing with questions

Most technical presentations include a question and answer (Q&A) session or at least include opportunities for questions from the audience. These can be intimidating for new presenters. There are some important things to keep in mind:


3. Visual aids

I've already begun to mention some of the basic issues related to visual aids in the preceding section on oral presentations and even in writing. Figures and illustrations can be extremely valuable tools in getting your point across, but if you use them poorly (e.g., figure description in a paper, or using poorly -executed figures in a presentation), then they can be a hindrance to getting your point across successfully. The term "visual aid" is thereby a misnomer in such cases. What are the characteristics of a good figure?

a. A good figure makes one point, clearly

The object of a figure is to show something; this might include some theoretical results, characteristics of observed data, or some structure in calculations based on either theory or data. If a figure makes too many points, it usually becomes confusing to the point of not getting any of your points across. If numerical quantities are involved, the figure makes clear such aspects of the numbers as the order of magnitude, the scale, and the quantity (or quantities) being represented. Simple figures generally are more effective than complex figures. If you find yourself writing a long caption to describe the content of a figure, then you should consider simplifying that figure. It isn't always true that long captions imply a bad figure, but as a rule of thumb, if the caption is more than two or three lines of text, the figure may need simplification.

b. A good figure is readily legible

Figures in papers usually are reduced from their original size. Figures used in a presentation may be viewed from considerable distances at the back of the presentation room. If the illustration is not legible at such distances, then the point will be lost on at least some of those in attendance. Journals likely will reject illegible figures but if you use them in a presentation, the audience's attention will be lost quickly. Visual aids in talks are notoriously bad at scientific conferences, where tables of small numbers or tiny equations are crammed onto a single slide or viewgraph. An illegible or confusing figure is a waste of everyone's time as it cannot possibly succeed in its mission to make some point. Speakers often use up too much of their allotted time explaining badly-crafted figures. If a figure takes more than a moment's explanation, it is not a well-executed visual aid.

c. A good figure has visual impact without going overboard

Although they are most common, illustrations or visual aids with black lines or text on a white background often have minimal visual impact. If possible, white text or lines on a dark background usually work better in oral presentations. The written versions may be acceptable to journals as black on white, but you should always be aware of the value of illustrations that are visually attractive. Avoid being garish by using wild embellishments like fancy textures or psychedelic colors, as these tend to distract from the point being made. As was the case with your manner of speaking, the object is not mere entertainment and the point being made should not be overwhelmed by the presentation.

d. A good figure's appearance is professional-looking

There is nothing inherently wrong with hand-done figures, but most of us lack the skills needed to make such figures look good. Right or wrong, the degree of confidence in the reader's or listener's eye is associated with the figure's professional appearance. Therefore, it generally is beneficial to have your illustrations done on a computer or by a draftsperson. There's no guarantee that good, well-executed figures will result from being done on a computer, though. It still takes care and skill to get a computer to give a professional-looking, effective visual aid. In my opinion, the skills of the professional draftsperson often are underrated, with the result that many professionals have bad figures that detract from their intentions.