To all my Data Scientists fellows: Stop using so many jargons!

Yukio
2 min readOct 13, 2022

--

Don’t be a data scientist/analyst that only uses jargon and/or terms that are incomprehensible to most people. You don’t appear to be smarter because of that. In fact, you are pushing people away, looking arrogant, , and probably making it difficult to implement your own objects (since you know what you want with that).

Instead of saying “the model got a f1-score of 0.75” or “the estimators are all significant”:

1. Prefer tangible examples: “As you can see, the estimators are all significant, which means there is really a 10% increase in the probability of default for customers with 60 days delay in payment.” Another example of speech from the same perspective would be something like: “out of every 10 customers who default, our model can predict/anticipate, on average, 8”.

2. Show the impacts your model may have on business results: “With this model, we can improve the relationship with non-defaulting customers and provide more effective installments for potential nonpayers. We estimate a revenue gain up to $10 million/year with more assertive installments “.

Both your peers and your leaders will appreciate this approach. You can rest assured, no one will think you are less intelligent for not using such intelligent terms and details. It’s just that you’re likely to gain much more morale and empathy from those involved.

Ps.: Since we are working with probability, be careful with the words you choose during your presentation. Remember that nothing is completely guaranteed when talking about statistics, it’s all about chance. False promises will get in the way of building a data-driven culture.

--

--

Yukio

Mathematician with a master degree in Economics. Working as a Data Scientist for the last 10 years.