4

In a colloquium, an eminent senior professor showed a graph from a peer-reviewed journal (published by some other group). The graph contained three curves:

  1. black dots (blurred)
  2. black line (blurred)
  3. red line (thick and bright)

All these curves in the graph reports the same property (obtained with different methods).

When talking about the graph, the speaker mentioned, "Since the authors thought that this data (by pointing the red line) is most important, they have used red line"

My question is, is there such practice of using red-line to highlight good results in research articles?

Note: I ask this question because, in Origin Software, the first and second curve are, by default, black and red. If those colors are used, it may given an impression that the red line graph is the better result (compared to black line graph).

closed as off-topic by Herman Toothrot, user3209815, Coder, E.P., Buzz Oct 17 '17 at 15:03

This question appears to be off-topic. The users who voted to close gave this specific reason:

If this question can be reworded to fit the rules in the help center, please edit the question.

  • 2
    why is this related to academia? – Herman Toothrot Oct 17 '17 at 12:18
  • 4
    It's simply unrelated, doesn't matter that similar questions were asked before, why not ask in statsexchange? So if this plot is shown outside academia it would use a different color? – Herman Toothrot Oct 17 '17 at 12:36
  • 2
    The question is related to the research practices. Thus, I believe that this question is suitable only in Academia SE. – phenomenon Oct 17 '17 at 12:57
  • 2
    I think this was a joke. The black dots are data points. The black fuzzy line is a confidence interval, and the red line is the line of best fit. – Dawn Oct 17 '17 at 12:58
  • 2
    Well, it makes sense that the data that the authors would like to highlight is in brighter colour. If the 3 lines were just 3 colors (red blue green), then I would not assume that the red is more important, but if all data is in blurred black color but 1, then obviously that 1 is meant to be highlighted, right? – Ander Biguri Oct 17 '17 at 13:39
11

As far as I know, there is no well-established standard. The only universally recognized guideline is "use colors that are well visible, also when printed on black/white (if it's an article) or displayed on a beamer with bad colors (if it's slides)".

I think the most likely explanation is that the speaker was joking.

Personally, in my talks I like to use what I dubbed the "Star Wars light-saber color coding": the 'good guy' (new algorithm of which we are trying to show the merits) is in blue, the 'bad guy' (leading competing algorithm) is in red, and green and purple are used for additional variants (with green=good-ish and purple=bad-ish). But that is also a joke and not a standard.

  • 2
    To add to a Star Wars theme: there were also researchers who used TIE-fighter plots in their published papers: scieastereggs.tumblr.com/post/157262173243/tie-fighters – lukeg Oct 17 '17 at 11:50
  • If there is any kind of convention, wouldn't it be to avoid red because of the prevalence of red-green colourblindness? – curiousdannii Oct 17 '17 at 13:15
  • @curiousdannii First of all, the problem is only when you use red and green at the same time --- using only one of them is perfectly fine. In addition, the red and green tints used by default in Latex/Beamer/Tikz/Xcolor are distinct enough to be distinguishable also by a colorblind person. – Federico Poloni Oct 17 '17 at 13:39
  • 2
    @curiousdannii: Presumably the convention would be to avoid color schemes that require you to differentiate between red and green. It doesn't mean that you need to avoid both colors; you just have to pick one. See here for more details. – Michael Seifert Oct 17 '17 at 13:39
  • 1
    @FedericoPoloni I was more thinking of some colours being less distinguishable from greys. That can be a problem with some red inks even for the non-colourblind! – curiousdannii Oct 17 '17 at 13:48

Not the answer you're looking for? Browse other questions tagged or ask your own question.