When I'm preparing research findings for my clients, I find that a well-thought-out use of colour can be a great way to spice up a research report. Whether it's setting the tone with a carefully selected sequence of colours or relating to your audience by tastefully integrating elements of your client's own branded colour scheme, colour is an important element in the overall look of your report. However, making a good choice of colours is hard: proper use of colour communicates while errors in colour choice confuse.
Here's just one example, pulled from an actual report:
The choice of two red tones may say something to the reader because it suggests a grouping, perhaps a similarity between the brands. But the chart's designer did not intend for this to happen. In reality brands A and B are no more similar to each other than brands A and E are to each other.
To make things more complicated, colour-blindness is a lot more prevalent than most people think: one-in-ten men are colour-blind. Consider the following example, a popular practice is to replicate colours used in traffic lights by showing Green for "go" and Red for "stop". The average reader sees the following:
Unfortunately, those with Deuteranopia (the most common form of colour blindness) are presented with this unhelpful scale:
Thankfully, cartographers have been working on this problem for a very long time, and have developed come basic principles we can follow when selecting colour palettes for our research reporting. These are based on whether what you are doing is Counting (sequential scale), Contrasting (divergent scale), or Grouping(qualitative or categorical scale).
You can read about these principles, and how to apply them in the latest VCU Best Practice Briefing, "Best Practices for Using Colour in Research Reporting".
However, not even these principles work in all cases, particularly when considering the different types of colour-blindness that can occur. So my advice is to use colour in your research reports, but never rely on colour alone for encoding your data: always label your data clearly, and use the colour principles to ensure your choice of colour works in support of the data.