Joho the BlogOur national wave of vandalism (a story of metadata) - Joho the Blog

Our national wave of vandalism (a story of metadata)

At some point we noticed a crack in the windshield of our car. It’s a single line, about 8 inches long, not the spider web fracture typical of a pebble or assassination attempt. We don’t know exactly when it happened or how.

Our insurance covers it. But when I went to file the claim through SafeLight, the person on the phone insisted that I give a reason why the glass broke. The fact is that I don’t know. But that is not an acceptable answer. Rain? Hail? Pebble? Collision? Branch? Vandalism? Collision? I don’t know. I don’t know. I don’t know. I don’t know. I don’t know. I don’t know. So I said it was vandalism.

I wonder how many of our national statistics are skewed by a failure to provide an “I don’t know” or “Other” box… [Tags: ]

3 Responses to “Our national wave of vandalism (a story of metadata)”

  1. Hasn’t this issue played out in recent decades of death certificates and “causes of death” reports? Whereas citations of old age, heart failure, and natural causes might have satisfied those lesser folks before us, we now know that they didn’t really know. And, of course, we don’t yet know what we don’t really know yet. Meanwhile, I like to think that if standardized testing had included “I don’t know,” as one of the answer options, I might have scored much higher.

  2. Good point. Have I told the story about the motor insurance case file I tried to data-mine about a decade ago, that seemed to have a huge number of drivers aged 90 and thereabouts? Specifically, lots of people seemed to have been born on the first of January 1901 and the 9th of September 1909. If the insurer doesn’t actually require date of birth, and if the DOB field on screen is formatted and mandatory…

    But I wouldn’t worry too much about national stats. Commercially-derived figures are going to be particularly patchy on this; public-sector stats will generally be better & surveys better still. Designers of some survey instruments really go to town on ‘don’t know’ options – one drug-use survey I’ve seen asked if the respondent had smoked anything they didn’t know the name of, sniffed anything they didn’t know the name of, swallowed anything they didn’t know the name of, injected anything they didn’t know the name of… or taken any other drug they didn’t know the name of. That last question really should have had a free-text rider – “If YES, what on earth was it?”

  3. De la misma manera poli poquer caribbean poker web texas holdem combinaisons telecharger jeu pc poker juego al instante paginas internet…


Web Joho only

Comments (RSS).  RSS icon