Asking questions of the data vs. asking questions based on data
Of course I come across many good examples. Unfortunately, that is not always the case. Some organizations know that they can (or should) do something with data, because there is already value in it. When you ask them what decisions they want to make more substantiated, it surprisingly often remains silent.
There are quite a few organizations that never formulate questions based on scenarios to identify the best option with data. Instead, they formulate loose insights from data that they simply collect. In other words: not 'asking questions to the data' but asking questions based on available data.
That seems like a trivial semantic difference, but it isn’t. Because while there is such a thing as ‘data serendipity’ (discovering something unexpected and useful in the data while you were looking for something else), it is precisely that don’t-ask-the-data-questions mechanism that also makes working with data (profiles) dangerous.
Also read: The secret behind data-driven organizations
The problem is first of all: data and segmentation are south africa telegram data not neutral , as Miriam Rasch shows . Data and the categories that follow from them are always based on choices. They therefore reproduce a perspective on reality. This effect also increases as we make this form of reality measurable.
Secondly, people like to put other people in a box. That helps us to understand the world and the people we encounter. And that conflicts with the core of a data profile, which according to Rasch does not provide a clear picture of someone. Because data profiles are not consistent : different ordering principles can overlap and the categories are not mutually exclusive. You score on everything, but not to the same extent.
And our fallible thinking doesn't help either
It is precisely this multi-interpretability that makes a data profile so valuable. But also so dangerous. The question arises whether people are really able to distinguish, value and use these different perspectives on a person? Or does our natural tendency to classify people, to approach them one-dimensionally, ultimately prevail?
Especially when you know that our judgments are driven by lightning-fast, automatic processes, over which we have no control and which we do not see, it quickly becomes problematic. That judgment runs on what Sarah Gagestein calls the fast-considered system (a more catchy name for Daniel Kahnemann's 'system 1' ). This is a kind of automatic pilot, which treats subjects and people in an emotional way and from stereotypes.
