Currently I am on my way to Montreal to attend one of the largest geochemistry conferences in the world, Goldschmidt, where I will be giving a presentation on some of the work I have done in my PhD. Overall this is by far a good thing, I get to travel overseas (for the first time – Tasmania doesn’t count), meet other researchers from around the world and experience the culture of other cities and geologists. However part of being able to attend the conference is producing a presentation and that includes creating diagrams.
Scientists love diagrams as they allow us to clearly show data in a visual sense, rather than by simply displaying numbers, making the meaning much clearer and instantly understandable. The difficulty comes in making that diagram as informative as it needs to be without including too much information for the audience to absorb; which varies depending upon the purpose of the diagram. Diagrams for oral presentations need to be easily understandable from a distance and in a short time frame as they are generally only displayed for about a minute, meanwhile diagrams diagram in written work, such as journals, can be much more detailed as the reader has more time and the caption can be used to help explain the diagram, diagrams used on posters are somewhere in between in complexity as they can make use of captions but generally are only viewed for a short time. All of this means that at the very least diagrams need to be altered if not created newly for each purpose; for example diagrams for a journal are generally in greyscale whilst posters and presentations allow full colour, lines, text and data points are much larger for presentations and colour contrast is extremely important.
The diagram shown above is one that I prepared recently for my presentation which evolved gradually over about a week, it shows the variation of oxygen fugacity (fO2) relative to pressure (which is equivalent of depth) for some mantle samples from South Africa. Key considerations included making sure that all the major points could be seen easily from a distance and also separate data sets distinguishable using both colours and the symbols. Other inclusions include literature data as this immediately gives the audience an indication of how my data correlates with previously published work as well as reference points, lines or fields that give a context as first glance. I also added text to quickly explain the diagram (and to make sure that I don’t mix up terminology).
On top of all the issues regarding would to include the choice of program to use, as with many scientists I tend to use Excel to store and manipulate my data, leading to the temptation to use it to create the diagrams (especially graphs). However the default outputs aren’t that great visually or content wise so lots of tinkering is required to get the diagram you want. Of course there are other options available and many people choose them, however as with most people I tend to only be creating serious diagrams when a deadline approaches. As such the learning curve of new software, particularly graphics programs is generally too steep when push comes to shove. Therefore I end up creating most my diagrams using excel, telling myself that I will learn how to use Kaleidagraph or Illustrator next time, but when next time comes around I am generally in a rush so the process repeats inself.
In conclusion, diagrams can be very useful but always take much more effort than expected to get them just right.