In the comments I received for a paper, the reviewers would like some of my figures' panels to be "signposted". I have no idea what it means other than it is supposed to be a technique to make the figures easier to read. What should I do in this case? The reviewers are talking about figures with multiple (>4) panels.
Thanks1
In short, it sounds like they are saying they are just getting lost in your figures. As such, they request "signposts", or some kind of further instruction on how they are supposed to traverse your figures and writing to see the point you are trying to make. In my experience, especially with complex figures with multiple panels, readers tend to get lost and don't know how they are supposed to make sense of what you are presenting. Are the panels in some kind of important order? Is there some progressive meaning to each panel which requires comparing between panels, or are they independent?
Signposting can take the form of a more descriptive figure caption, better panel/figure titles, or further instructions in the text. For example, in a figure with 4 panels I thought the most important takeaway is that the points on the graphs got farther apart as you moved from left to right across the figure, showing increased dispersion between methods I was comparing - but readers had a problem figuring out what the figure showed at all, and why I thought it would be important. So I added some descriptive text, along the lines of "the panels are ordered by increasing effect size, and show an increased in dispersion between conditions" - and then I expanded upon that in the text, explicitly referencing the figures. Sometimes it is better to remake the figure entirely to make the key point more clear, and then use additional figures or an alternative design to show the less-central findings.
This is especially important with complex panels, as we who have built the figure have likely spent many hours staring at the results and see them as pregnant with meaning - but to a reader who is just encountering them for the first time, they don't see it the way we do. Even if there are supposed to be many important observations in the figure, at least 'signpost' - describe - a single, clear main idea that is most important. You can then reference it further in the text, returning to it for additional findings, but you have to make at least one key observation relatively obvious or most readers will give up on the figure entirely.
