Everyone wants to be data-driven: "We want to do things empirically correctly, so we just rely on the data to provide a value-neutral assessment of the best course forward."

Except for the most part, data-driven tends to be data-informed, at best. Because you can tell any story from a given set of data. We'll talk what data can actually tell you vs. what it can't, looking at specific case studies and examples.

We'll discover how to find the right data to help drive your decisions, and how you might actually be able to collect it. And at least once I will get so upset about someone using data incorrectly I'll start sputtering, which is pretty amusing for everyone.

Comments

Please login to leave a comment

Ian Littman at 14:34 on 26 Oct 2024

The only reason this is four rather than five stars is that the talk ran quite short (~25m in a 50m slot); in a shorter track format the presentation is excellent as-is.

This talk mixes explanations, suggestions, and anecdotes skillfully. Maybe I'm biased because these items resonate with my personal experience, but hopefully this shows up at future conferences elsewhere. Could even work in a single-track conf.

Joseph Lavin at 14:35 on 26 Oct 2024

Lot's of great and useful info! I appreciate the blog post with the same info.

Lane Staples at 12:16 on 27 Oct 2024

This talk was outstanding!

It just also ran short, which I suspect was due to some technical difficulties resulting in the last minute loss of speaker notes based on accidental eavesdropping from the front row. So, less than perfect rating due to wanting more outstanding content from the time, but almost perfect rating because Kait did such an amazing job adapting and delivering their content from memory.

I really loved the framing of this talk; the distinction between narratives and fairytales and relating the stories we live our lives through to data and our emotional attachment to all of the above was such a succinct explanation of magical thinking and its pervasive tragic impact. That framing is immediately part of my internal vocabulary. I'll try to remember to cite my source ;)