Hey folks,

Earlier this week, those of us in the US celebrated Memorial Day. For many, this marks the unofficial start of summer. I suppose the clock is now ticking until Labor Day, which is the unofficial end of summer. Let me be the jerk to tell you that you have 100 days left to accomplish all of your summer goals.

I suspect that for many of you writing papers and putting together conference posters and talks are on your list of goals. Generating attractive visualizations of your data is critical to those goals. For most people this is a challenge that fills them with existential dread.

But what if there was a way to get help? What if there was someone who would be willing to help you design attractive visuals, provide feedback, or give you a nudge in the right direction? If that sounds fantastic to you, I’m your guy.

I provide free 30-minute exploratory meetings to discuss how we might work together to design the figures for your next paper, talk, or poster. You can sign up by clicking the button below!

Although the next 100 days may seem like a race to accomplishing a “productive summer”, it’s really an opportunity to learn skills that will help you for the rest of your career. Let’s talk.

I remember giving the first draft of my first scientific paper to my PhD advisor. It was a mess. I’ve since learned that every first draft of every first scientific paper is a mess. My advisor patiently went through his edits with me. One of his questions was why I was using different words for the same concept. I thought it would be tedious to keep repeating the same word over and over. I also felt the need to use different adjectives to make the writing more attractive. No. Simplicity gives writing precision.

Imagine if I used “16S rRNA gene sequencing” and “amplicon sequencing” interchangeably throughout my text. I might think that by using different jargon for the same thing that they were going easy on the audience. But, it would get confusing to my readers because they might wonder if there’s a difference between the two techniques that they were missing. My advisor was correct, scientific writing requires precision. In this example, “16S rRNA gene sequencing” is a type of “amplicon sequencing”. I wouldn’t want to say something of amplicon sequencing that wasn’t true of 16S rRNA gene sequencing and vice versa.

Don’t get me started on the mixed uses of “16S rRNA gene sequencing”, “16S sequencing”, “16S rDNA sequencing” and so on. It’s “16S rRNA gene sequencing”. If you don’t know why, remember what my advisor told me. Those variants all imply different things that aren’t really true or are at best sloppy. There’s no DNA in a ribosome, “rDNA” is also used to mean “recombinant DNA”, “16S” describes the size of the small subunit ribosome and doesn’t mean anything apart from that context.

Now that I’m an advisor working with junior scientists on their own writing, I find that being comfortable with repetition and striving for precision is challenging. But what does this have to do with data visualization? Check out these two panels from a figure recently published in the journal Nature as part of a paper titled, “αKG-mediated carnitine synthesis drives DNA repair via histone acetylation”.

I’m not trying to pick on these authors. This is something that I see frequently. This example isn’t as egregious as other examples I could have picked. One panel might use a dynamite plot, another a box plot, and another a violin plot. All in the same figure. Why? I suspect it is the same as trying to change up the jargon. Perhaps because it seems boring to have 10 box plots.

Similar to writing, we need to get comfortable with some level of repetition. By changing the geometries we use to represent data, we risk confusing our audience who will wonder if different geometries were used to mean different things. Our audience may need to learn how to interpret the different types of visualizations. What do the error bars represent on a dynamite plot? What do they represent on a boxplot? What do they represent in panel a? What the heck is a violin plot? Answering these questions makes it more difficult for your audience to see what you want them to see and move through the story you are trying to tell. Keep it simple.

In this case, there are no more than 8 points for each condition. Just show the data and show the P-values or some indicator of significance. That strips away all the distractions. It lets the audience see the data as they are. Adding all the ornamentation of a bar and unnecessary confidence intervals and so forth is equivalent to trying to use your big SAT words to impress your audience. Keep it simple.

Just as we aren’t trying to be the next Steinbeck or Atwood when we write a scientific article, we also aren’t trying to be Rembrandt and certainly not Picasso or Pollock. Keep it simple. Strive for precision and clarity.

​

​

​

​