Aaron McDaid

Thinking recently about the relative importance of various skills required by scientific researchers, I wondered if expert programming skills would become more important in future, rather than less so. The same question goes for expertise with statistics.

It's easy to understand how some people might think that modern technology means that such expertise is less important for the 'scientist'. They would think that modern easy-to-use software means that scientists can concentrate on their domain knowledge (genetics, cancer, or whatever); they would think that all that 'hard work', processing and analysing the data, will be done by the software.

But I suspect the opposite is the case. I believe that serious software programming expertise will become even more important in future, just as scientists will continue to need to truly understand the statistical techniques that are applied to their data. Modern technology is obviously vastly improving the quality and quantity of scientific reseach. But it gives most assistance to those that are truly competent in software development. In order to stay relevant, all researchers will have to fully embrace computer programming and become good at it.

And they can't just hire a bunch of techy PhD students and delegate it to them. A competent scientist-programmer will have many simple hypotheses to try on the data. They will be able to run many simple tests in a short time in suitable software such as R. They will also be able to do more complicated analyses as soon as they have the idea. In contrast, imagine an old incompetent scientist had to write an email to a colleague/PhD/intern for each request; can you imagine the confusion and delays while trying to explain how exactly to process the data?

Aaron www.aaronmcdaid.com