Why user researchers should learn to code

I’ve never been very good at coding. However I’ve found that the bits I do know have helped me become a better user researcher. In this post I explain how this knowledge has helped me, and why I recommend that all user researchers should learn how to code.

Understanding and empathy

User researchers work on technology projects, whether it’s games, websites or applications. Although we often interact with designers and product owners, often the solution will actually be implemented by programmers, in code. Learning how code works can therefore help user researchers in a number of ways.

Understanding how code works can help improve researcher’s understanding of the product, and the feasibility of solutions. We’ve all had sessions where participants have suggested what they believe are ‘minor’ changes, which are actually a huge amount of work – such as adding a multiplayer mode to a single player game. Learning how to code can help researchers accurately assess the amount of work that solutions would be, and help prioritise what should (and can) be fixed.

This can have a knock on effect on communication with the client – by not making ridiculous suggestions, researchers are increasing their credibility and value to the team.

Statistical analysis

A significant aspect of a user researcher’s job is quantitative studies, and understanding statistics key to this. Certain types of clients often respond better to numbers and believe that they often “definitive proof” over whether something is an issue or not. One of the most common tools that researchers use for statistical research is R, on account of it being free, and its success has led to a wide range of custom made packages and pre-made analytical methods being available.

Statistical analysis with R uses many of the techniques from coding, such as for loops and if statements, which are basic techniques for coding software. An understanding of how to code is therefore directly applicable to being able to perform statistical analysis with R, and can greatly reduce the learning curve – to use R a programmer would only have to learn new syntax, and can rely on their pre-existing knowledge to understand how to do it. One of the main benefits of R for user researchers is that it allows for the creation of automated scripts, which can greatly speed up analysis, and allow us to present results to the client faster.

Creating custom tools

As part of my MSc, I created a custom tool in processing to allow researchers to codify social interaction observed during playtests. I did this using Processing, an “electronic sketchbook” which allows the quick creation of software using a Java-esque programming language. Processing is easy to learn and allowed me to quickly iterate designs and create a useful tool customised to the research requirements.

Many studies use methodologies that are highly bespoke and can benefit from customised research tools – some work upfront prior to a study can save valuable time during analysis, where the deadlines are often a lot tighter. By learning to program, user researchers can create these tools and improve their effectiveness at running and analysing sessions.

Convinced?

Most researchers I’ve spoken to have an understanding of how coding works, and have studied it in the past, even if it is not something they do regularly. Hopefully this post has encouraged you to develop and apply these skills to user research, and provide greater benefit to the teams you work with.

150 150 Steve Bromley

Leave a Reply