I am the sole statistician in my group, and I am scared of making mistakes

Question

I am fresh grad statistician and I am working in the agronomy industry (first time working with this kind of data), I am doing the best the best in my knowledge to produce analysis within the given time frames but I am scared that I may make a mistake or interpret a problem in an incorrect way.

I try to validate my results with my colleagues, but I can't validate statistical methodology or discus methods with them.

How I should deal with it?

Answer 1

I suggest you reach out to your manager with your concerns. Perhaps he or she has advice or knows someone elsewhere in the company with the expertise you can consult with. Aside from that, perhaps you can indeed discuss your methodology with your team. You might be underestimating them, they may have intuition to compare against your results, or barring that, the act of explaining can help you realize mistakes where they are present.

Answer 2

Try not to fool yourself - but you are already skeptical of your work, which is good!

One tip for checking results in uncharted territory or where you have no peer to oversee you: identify 2 or 3 different, ideally independent, ways of verifying these results. If all give the same outcome, you are in good shape. Basically, it's the same principle as with "checksums". This can hold for individual formulas as much as for statistical statements, where having several ways of looking at a set of data is good practice anyway to understand what it says.

Answer 3

You go and do your best. Everyone makes mistakes. Many of yours will not be found, or they will be found by you. Be open if your results are questioned, because you know you can make mistakes, but show confidence. If it happens, verify extra carefully.

PS. Sometimes it’s hard to find the correct result, but relatively easy to verify that a result is good or bad. If that happens then make sure you verify your results if possible.