I Never Thought I Would Utter Those Words . . .
It is heretical.
But even the least experienced data scientist is aware of “Data Bias”.
We have all seen it.
The data “analysis” that proves any point you want it to make, the two departmental groups arguing about the conflicting interpretations of “their own” respective data, and the correlation that is misinterpreted as causation.
BUT, data is, like cash, a critical resource for business, whether in executive, operational management or “getting the job done better every day” positions.
Anna Shraga raised a challenge this week with the question “when making a decision, do you tend to trust your instincts or the data?”
It is an interesting question as it touches on inherent biases.
Data bias – I appreciate and am very invested in data but am very aware of data bias.
Cognitive bias – I trust my instinct up to a point – but am also painfully aware we all suffer from cognitive biases (halo effect, Dunning Kruger etc) which means we have to challenge our own instincts as much as otters . .
So, it turns out that my answer is be careful and use as many different perspectives and data points as possible, especially if it is an important decision!
The psychologist Daniel Kahneman refers to the problem as WYSIATI – What You See Is All There Is. Reliance on data may blind you to the obvious more often than it alerts you to the unknown.
As Jeff Bezos once commented, “When the anecdotes and the data disagree, the anecdotes are usually right.”
Renowned behavioral science expert and Ogilvy vice chairman Rory Sutherland, argues businesses that rely too heavily on data look clever, but often act dumb.
So, how do we resolve this dichotomy?
We need data, we need multiple perspectives and ALWAYS with context. Context, in business terms, is the process view.
Data without sufficient context or of too narrow a focus is dangerous. That’s why a focus on end-to-end process data is so key. It provides much needed perspective.
“Beware data, it can blind you to the obvious”
Rory’s article in Management Today on this “Data Dichotomy”, is worth a read, here . . . You may need to register but the article is free of charge to read.
If you have an opinion on this week’s “Something to Consider”, whether you agree or disagree, I would love to hear from you.
Getting as many different perspectives and data points as possible turns out to be more useful than I thought!
Thanks for reading . . .