The knowledge about the bias variance trade off is very important for you as a statistician as it comes across your work - especially exploratory work - all the time. Taking care about it will help you avoid problems of over- or underfitting your models.

Jenny has encountered several highly competent people throughout her career who have admitted that they entered one path or another based on certain assumptions that were not entirely correct. Further, she has talked with interns about their future career interests and she sees them teetering on the edges of some same fallacies. It’s all fun and science until you have to get a job. As you will make this decision at some point, Jenny wants to give you a few key points to consider when evaluating potential statistics / data science careers in drug development and healthcare.

The COVID-19 pandemic has put pharmaceutical research firmly into the spotlight, with the urgent need to develop new therapies, tests, and vaccines. For those outside the industry, hearing announcements in the mainstream media, it is not always easy to understand the intensity of the work and collaboration that fuels game-changing new developments.

However, we know that there are thousands of ‘unsung heroes’ around the world, working in different capacities on different clinical trials to make the breakthrough. It has been a herculean collective effort.

Should we oversimplify statistics, or should managers learn more about statistics?

If you've been working as a statistician for a while, you know that you have to keep it simple if you want managers to understand you. Of course, I speak about those managers not being statisticians themselves.

There are a lot of ways to explain statistical concepts to them, but this adds time to the presentation. Sometimes there won't even be a presentation, and all you have is sending the report via email.

From reading, from my experience, and from talking to other experienced statisticians, I have distilled the knowledge about this challenge into 5 steps.

Outside of statistics departments, we are often seen as tactical and operational implementers. This branding makes it hard for statisticians to be included in strategic discussions. This limits our ability to influence the organizations and projects. It can even lead to entire departments being outsourced or not appropriately funded because of this terrible reputation. I'm long on a journey to fight against this branding. This journey also requires us statisticians to learn about thinking and acting strategically.

Our job as statisticians focuses primarily on the analysis and communication of uncertainty in its many aspects. For this episode, I think of uncertainty as the variability in the population we are studying, the uncertainty around estimates, or the model uncertainty, which we might capture via looking into various models.

What is change management? Is it only something for big corporate initiatives or does it also apply to smaller changes?
Why does it require so much consideration?
What's hard about change?

Change management is something I never thought about being interesting for statisticians, but recently, I needed to dive in to this frequently. I learned that this helps me when creating new designs or processes as it answers lots of questions about getting from an idea to an implementation.

Knowledge sharing is close to my heart and Nelson, who is a very knowledgeable statistician, has shared a lot of articles in LinkedIn and on his website.

Nelson has always been keen to learn from others and share that knowledge broadly, so we'll be exploring similarities and differences in how he's seen that be achieved in different circumstances.

Do you want to review a data visualization effectively?
Do you want to provide your organization with a simple tool to improve your data visualizations?
Do you need a checklist for yourself to make sure you covered everything for your data visualization?
This episode is for you!