Have You Been Gaslit By Data Science?


Machine learning excels at automating routine decisions.

But the decisions that matter most are far from routine.

The decisions that truly make or break an organisation are the high-stakes, one-off decisions where data is scarce, machine learning falls apart, and executive stakeholders are left relying on their gut.

These are also the situations where data scientists have the potential to add the greatest value - if they know how.

As a data scientist with a background in actuarial science and statistics, when machine learning has failed me, I’ve found myself falling back on actuarial and statistical techniques better suited to sparse data - while all the time worrying I was somehow doing data science “wrong”.

If I wasn’t creating a machine learning solution, could I even call myself a data scientist?

So when I came across How to Measure Anything by Douglas Hubbard, it was a revelation to me. Not because it taught me something entirely new, but because it confirmed I was doing data science right all along.

It was as if data science had suddenly stopped gaslighting me.

In the latest episode of Value Driven Data Science, Doug joins me to share how combining techniques from statistics, economics and decision theory can help data scientists tackle the problems that matter most.

You’ll discover:

  1. What Applied Information Economics is and how it works in practice [03:17]
  2. Why organisations are systematically measuring the wrong things [09:23]
  3. How the Lens Model can make expert judgement more reliable than the expert themselves [13:44]
  4. How AI can turbocharge the Applied Information Economics approach [21:10]

The decisions that make or break organisations rarely come with large datasets. This episode is about what to do when they don’t.

Listen now on Apple Podcasts or Spotify, or click the link below:

Episode 109: How to Measure Anything and Make Better Decisions

Talk again soon,

Dr Genevieve Hayes

Data Science Impact Algorithm

Twice weekly, I share proven strategies to help data scientists get noticed, promoted, and valued. No theory — just practical steps to transform your technical expertise into business impact and the freedom to call your own shots.

Read more from Data Science Impact Algorithm

The number one piece of advice I give to data professionals is to have more conversations. Conversations lead to good things. Things like job offers and stronger relationships. But more importantly, the opportunity to deepen your knowledge through learning from the wisdom of others. I owe many of the best opportunities I’ve encountered in my career to proactively seeking them out wherever I can. My podcast is a weekly example of the power of conversation. But if there’s one thing all these...

While working as a data scientist in a large organisation, I was once called on to attend a meeting with representatives of a big tech company and multiple senior leaders from my business area. The tech reps were there to sell us a new product. The data scientists had been invited because we were the intended users. Although there were a dozen or more people in the room, I was the only person who came prepared with any questions - 2 1/2 pages of them, to be precise. I let everyone else have...

Like many readers of this newsletter, I identify as a data expert. It took me years of study and hard work to build this expertise, and it’s something that I’m not willing to lose. This is not just because of everything it took to get here, but because my expertise forms an integral part of my identity. And the idea of losing it feels, in many ways, akin to losing a part of who I am. So, when I recently started hearing from data professionals who spoke of slowly losing skills as they...