Overall notes:
Remember the main point of a chart is to clearly show(off) your data.
Your chart design should include ways to highlight and emphasize that data and
should avoid anything that detracts from the data and your main message.
Title:
Most published figures do not have a title. The axes labels and plots
should clearly identify the point of the figure. Other relevant
information (including statistical values like R2 and p-values)
would be found in the figure text below the figure.
If you do use a title, the title should not be redundant with the axes
labels. (i.e. it shouldn’t be: “Temperature vs.
depth”, for instance. We can see that by looking at the
axes.) Rather, the title should give information that’s more
general: perhaps a lake name and date of the sample.
Axes:
labels:
- labels should include the parameter
that is being represented on that axis.
- they should include the units used
(metric).
- labels should not include
abbreviations other than those for units: e.g. don’t write
”D.O. (mg/L) ” but rather
spell it out: “dissolved oxygen (mg/L)”
- it is ok to use (standard)
abbreviations for units
scale:
- scale should be selected to highlight
and fully show the variability in your data.
- If your data range is from 1010
– 2900 units, then the scale should start at 1000, not zero.
Optimize your usage of space.
the text size of your scale should be large enough to be easily read, yet small
enough that it doesn’t take focus away from your plotted data.
Figure 1. A not-so-good example of the chosen range of values given
on each axes. The data are de-emphasized and there is a lot of
wasted white-space.
Figure 2. A much better example with the ranges of values chosen to
highlight the range in values observed in the data. The trend is much
more easily seen and values for each individual data point can easily be
estimated from the graph.
Color:
use color only when
absolutely necessary (or for oral presentations). Printing in
color is expensive. Lines & markers that are only distinguished by
color differences don’t reproduce well (photocopiers are still b&w).
Often, color is used in a distracting manner. Instead, use different
symbol shapes and line patterns (solid, dashed, dotted) when needed. Use
black, grey, and white markers.
Legend:
if your graph only has one plot (for example, only has temperature vs. depth),
you should not have a legend since there are no lines that need to be
distinguished from one another.
the legend should not be the same size as your plot (or even close to the same
size as your plot). It will draw attention from your data, which is
why you created the plot in the first place. The legend should be
placed in a corner of your plot where it does not interfere with your data, or
neatly off to the side/bottom.
Plots:
markers should be of an appropriate size: small enough that they are
distinguishable from each other, and large enough that they are distinguishable
from the line connecting them.
Lines without markers (like Secchi depth or photic depth) can be distinguished
by using different types of lines: dotted, dashed, solid, etc. rather
than color since color doesn’t photocopy well.
You should not use the grey background that Excel uses as a default. Use
a white background. It saves toner/ink and is also easier to read.
Gridlines:
gridlines generally detract attention from the data you are presenting and
should be avoided. Of course there are exceptions to this, but they
should only be used when absolutely necessary for getting a specific message
across in your chart.
Figure Text:
figure text should be below your graph. It should start with
“Figure 1.” or other appropriate number. It should then
briefly describe the graph, include any relevant statistics (p-value, R2,
etc. ).
Formatting:
the modern age of computers and software allows for correct usage and placement
of superscripts and subscripts. Take advantage of that.
“ oC ”can be made simply by using a superscript letter
o, placed before a capital C. If writing H2CO3,
use the appropriate subscripts. Use the help feature of your software
to discover how to do this if you don’t already know.
