Discussion about this post

User's avatar
Paul M Sotkiewicz's avatar

Brian, as you alluded to me a while back, you would get to this! Pun intended, it is a measure above! The history is great at showing the folly of measurement as being untethered to human choices and decisions. But the punchline you deliver at the end is the lesson we should take away. Some really import at things cannot be measured: motives, emotions, psychology, preferences (ordinal yes, cardinal no). Data is just a collection of measurement in which the measurement has been agreed upon by those using it. Big data is just a larger set of measurements. But data itself does not tell us anything about linkages or causality, though we can find correlations. The big mystery in the world is how do these data fit together? That is where data and empirical observation (if this, then that) come together to build models of the things we want to understand in the world. Economic phenomena, weather, physics, chemistry stem from observation and then we use data to build models to replicate what we observe. But in the end we need an agreed upon base by which comparisons (the measurement) can be made.

Still, data cannot explain human behavior, decision making, or personality traits like the dark triad. Those come from classification of observations. And even if data are available, we all interpret data differently based on our own experience.

We omit measuring things like “home production” as having value in GDP. Why? There is no money changing hands. If I pay somebody to do it, it counts in GDP. If I do it myself, no transaction occurs and it is not counted. In both cases the service gets done, but the measure is quite different. What is my value of happiness from reading your Substack? Certainly more than what I pay, but does that get measured? No. Why? Can’t measure something without a transaction and we do not transact intrinsic happiness.

Hope this finds all who read this happy in the most immeasurable ways so they do not become shackled to a spreadsheet!

Expand full comment
Susan Linehan's avatar

A really intriguing post. It reminds me of Wordsworth's "we murder to dissect" (though I don't agree with him about books! Nor that we should stop thinking and just "immerse."). I've noticed that things like unemployment figures or inflation have a gazillion [note the precise measurement figure] different definitions. No wonder we talk past each other so much.

Obviously those varying measure are created for different purposes. But those are not clearly defined in most reporting.

I've been reading several books on "predictive brain" theory which says (much over simplified) that our own reality is shaped by the predictions the brain makes to fill in the gaps in perception. Most of the neuroscience supporting it seems to deal with things like pain perception and the effect of predicted assumptions on the kind of arousal that leads to trigger-happy cops.--this is an unconscious process: the cops may truly believe they are not prejudiced. So it helps explain the underpinnings of systemic racism.

And it does seem to explain confirmation bias, amongst other things. The same can be said about the "tech bros" reliance on data as the be all and end all capable of explaining the world. We need a lot less hubris and closer attention to the way we parcel out "reality." Thinking about measurement and its attempts to corral reality is a good place to start.

Expand full comment
16 more comments...

No posts