Here's a statement of the obvious: The opinions expressed here are those of the participants, not those of the Mutual Fund Observer. We cannot vouch for the accuracy or appropriateness of any of it, though we do encourage civility and good humor.
Support MFO
Donate through PayPal
Why It’s So Freaking Hard To Make A Good COVID-19 Model
I think it is more a case of confounding factors for which we cannot control. Data is data; it really can't be 'good' or 'bad'. The collection design can be poor; or we may not be able to isolate a single variable, but that's about it as far as the data involvement goes. Where we tend to screw up is in the INTERPRETATION of the data; especially data which contains confounding factors. People then tend to see what they want to see or, indeed, see things which are entirely illusory. I think the article did a good job of showing the difficulties involved in collecting data for which evaluation would be straightforward.
Well, to quibble about "good" or "bad" data, what about the 737Max? Data from a faulty sensor was input to a computer which essentially caused two aircraft to crash. (This of course is a very simplified, but still factual description of the complex circumstances of the crashes.)
I could nitpick and say it's imaginary 'data', or not really data at all, but that's just splitting hairs. In truth, I wasn't thinking about completely bogus 'data' as being a possibility; I more or less presume that when I'm measuring something, it is really there! Good catch and spot on, Old Joe.
@racqueteer- Thanks. I'm perhaps a little more prone to that perspective than other people because I spent years working with data systems that controlled, measured, and reported on San Francisco's public safety communications systems. For example, the status of an emergency generator at a remote radio site, or for that matter, on the activity of the radios themselves.
That sort of experience teaches one to trust nothing absolutely.
Comments
Was that data good, or bad?
That sort of experience teaches one to trust nothing absolutely.