Achieving Data Trustworthiness Through Humility

This is the second in a series of posts about moving beyond data quality to data trustworthiness. You can start from the first post here.

The most critical key to trustworthy data isn’t a technology, process, or procedure: it’s the virtue of humility.

Every person, no matter how gifted or experienced, is fallible. Every system, no matter how well designed or implemented, is built by people and is therefore also fallible. Mistakes will happen. Data issues and inaccuracies are inevitable. Try though we might (and we ought to try, with all our might), there will never be perfect data.

Of course there’s nothing profound about that. We’ve all found issues with datasets, data analyses, data pipelines, and data systems — even the best of them. It’s easy to see fallibility in others.

It’s just hard to believe it of ourselves.

This misperception is the greatest threat to data quality. When we think we know what we’re doing, we skip quality assurance checks we don’t think we need. We take shortcuts and unnecessary risks. We explain away concerns rather than investigating them. We get defensive when people share concerns or ask questions. When someone else’s data doesn’t match ours, we insist they are wrong.

And the more experience we have, the more likely we are to fall to this hubris. New data workers know they are going to make mistakes, they just don’t know the right practices to prevent or catch them. Experienced data workers know good quality assurance practices and are tempted to view them as training wheels — and what senior data scientist, analyst, engineer or developer needs training wheels?

Humility starts by acknowledging our fallibility. It recognizes that no matter how talented or experienced we are, we will make mistakes. It owns those mistakes, is diligent to prevent them, and is vigilant to catch them. It takes every concern seriously and dives deeply to resolve any questions. A humble data worker is predisposed to high data quality.

A humble data worker also earns more trust. They take other people’s issues and concerns seriously. The nine times out of ten they prove there’s no real issue are nine times over they’ve proven the trustworthiness of their data. They don’t get caught explaining away real issues, and they rarely make the same mistake twice. And since their humility has super-sized their concern for data trustworthiness, they have the fewest mistakes to begin with.

Humility is the foundation of data trustworthiness, weaving through every other key principle. In fact you can’t have our next principles without humility: safety and transparency.

Related posts:

Have you ever wondered why stars twinkle, or why the sky is blue in color, or how a rainbow forms? Or why water floats on lotus leaf like marbles? Science works in mysterious ways and has a reason…

Foster Learning Pakistan is a social enterprise; aims to develop the soft skills of Pakistani youth through its flagship training program. Flagship is a two months training program for youth and a…

This article will usher you to configure a plurality of VOD content as an input to the media live channel and play them in a sequential looping playback fashion. Before we start with detail on how to…