The Perils of Data Categorization

4 thoughts on “The Perils of Data Categorization”

  1. I enjoyed this post – thank you.

    The more I try to comment (gathering my thoughts before I type), the deeper I go down the rabbit hole (or the tangle of tunnels that is what’s left of my old brain).

    Categorization is definitely a faster way to think, however the faster I go more chance of error I make. And the faster I go the less ability it seems I have to think more deeply. In business, speed is life – what’s the old saying “Run fast and break things.”?

    I’d guess speed is not the point with YouTube but volume. They want us to watch more content and pull more ad revenue.

    I also remember taking surveys with multiple choice answers and sometimes none of them fit my response. Or worse yet, having to ‘adjust’ my response to a personnel performance appraisal as to not trigger a ding on my Supervisor or Manager (because you know there’s always potential retribution). Whatever that category was, then it was falsely weighted. Maybe I’ve digressed into metrics here…

    I have to wonder how many people think about being categorized. Do they care? How do they think if affects them (or not)? When I click the thumbs up rating on an Amazon delivery, what happens behind the scenes? When I don’t follow a medication recommendation from my physician (and he gets dinged in his metrics), what category do I get placed in? Just because I disagree with a risk factor doesn’t mean I don’t value the physician, I just don’t want to support the system that rates his performance in a way I don’t agree with.

  2. … i think you’ve over-indexed on “intent” … because it really is impossible to know, with absolute certainty, what our intent might actually be. realistically, it’s a confluence of a number of different known and unknown factors / biases.

    categorization is still very useful, as you say, because it provides context. but what you’ve gotten wrong is where you put the blame; it’s poor technology models / implementations that have “metastasized” categories and made them less useful… better, more empathetic tooling solves this by removing things like algorithmically-controlled feeds — this has been my focus for the last 4+ years as i’ve also grown tired of being man-handled by algorithms… i just want a feed that is not controlled by corporate interests (i had to build it myself).

    categories are still useful but not all tooling stays useful.

  3. Love this data reflection. I cringe upon seeing Google implementing more autocomplete options in Google Docs, comments, emails, etc. I fear what it does to the brain to not have to fill in one’s own gaps along with removing the ability to pause before figuring out what you want to say. I wonder if it’ll lower our tolerance for starting from scratch and rob us of building more creative muscles. Plus, I’m ashamed to admit that trust in my ability to spell words correctly has greatly diminished in the last 4-5 years as autocorrect takes over!

  4. “Categorization is part of the human condition. Our brain uses categories to help us make sense of a lot of facts we experience. It is how we learn. As humans, we need categories to contextualize our world, and that includes each other. What is more important is the intent behind the categories.”

    Om – this resonates with my focus on narrative design for organizations and people over the last six years. Narrative lives at a higher meta-level vs. storytelling. Stories are directional – from me to you. Narratives are bi-directional and open dialogues that allow for “context” and “clarity” to arise between parties.

    The other counter-intuitive element to narrative is that it is a “pull strategy” versus the push of a story. To tie that back to your comments above, people will navigate and adopt the narrative that aligns with their belief systems. We are in an era of “slippery slopeism’s” with technology and social media amplifications ready at the touch to distort, twist and scale.

    Hence the perils you speak of.

Comments are closed.