3 ms·

"We decided against showing trends over time on ward pages because the overall trend in the number of tickets issued is too big and complex [...] it would have

"We decided against showing trends over time on ward pages because the overall trend in the number of tickets issued is too big and complex [...] it would have been outside the journalistic goals of highlighting systemic injustices."

mlthoughts2018 8y ago
It does feel like there is an element of cherry picking in the chosen data presentations. It invites believing “injustice” trends while obfuscating what the true multiple-comparisons adjustment for some type of significance or posterior credibility would be. The conclusion that tickets are enforced in a way that manifests injustice is probably broadly true, but the data presentation actually makes it quite hard to interpret what’s “real” and what might just be an artifact of cherry picking, not to mention what slices of the data could be exhibiting something like Simpson’s paradox, etc.
arcticfox 8y ago
I appreciate the goals of this project, but IMO they aren't being very rigorous about it, and really only arrive at the conclusion that "more blacks are poor, and tickets hit poor people very hard", which I think anyone would already agree with before this study. I don't really understand what the ward breakdown adds. "Which Ward Owes The Most?": right, poor people can't afford to pay tickets and build up debt. "Where People Are Ticketed Most": this seemed promising. I clicked and mostly majority-white neighborhoods popped to the top. This is explained away by saying they're more congested; OK, fair enough, but try to control for that by population density? Or something? How am I supposed to read "unfair to minorities" out of that? "How Ticket Type Varies By Ward": maybe there's something here, but this seems just as related to other factors to me as the previous point. From what I've seen in Chicago, the majority-black neighborhoods are vastly different than the majority-white neighborhoods (I feel like ProPublica would agree), so a difference in ticket type seems incredibly logical. Yet here it's not dismissed like they dismissed the previous category. "What Happens if You Don’t Pay": right, poor people can't afford to pay tickets and it ends in a property seizure or bankruptcy. "Who Appeals and Wins": sorting by "Percent contested and won" the minority neighborhoods fair very well, consisting of 5 of the top 8.
- IOT_Apprentice 8y ago
  Perhaps to call out the politicians who could do something about the preying on the poorer areas?

mlthoughts2018 8y ago

It does feel like there is an element of cherry picking in the chosen data presentations. It invites believing “injustice” trends while obfuscating what the true multiple-comparisons adjustment for some type of significance or posterior credibility would be. The conclusion that tickets are enforced in a way that manifests injustice is probably broadly true, but the data presentation actually makes it quite hard to interpret what’s “real” and what might just be an artifact of cherry picking, not to mention what slices of the data could be exhibiting something like Simpson’s paradox, etc.

arcticfox 8y ago

I appreciate the goals of this project, but IMO they aren't being very rigorous about it, and really only arrive at the conclusion that "more blacks are poor, and tickets hit poor people very hard", which I think anyone would already agree with before this study. I don't really understand what the ward breakdown adds. "Which Ward Owes The Most?": right, poor people can't afford to pay tickets and build up debt. "Where People Are Ticketed Most": this seemed promising. I clicked and mostly majority-white neighborhoods popped to the top. This is explained away by saying they're more congested; OK, fair enough, but try to control for that by population density? Or something? How am I supposed to read "unfair to minorities" out of that? "How Ticket Type Varies By Ward": maybe there's something here, but this seems just as related to other factors to me as the previous point. From what I've seen in Chicago, the majority-black neighborhoods are vastly different than the majority-white neighborhoods (I feel like ProPublica would agree), so a difference in ticket type seems incredibly logical. Yet here it's not dismissed like they dismissed the previous category. "What Happens if You Don’t Pay": right, poor people can't afford to pay tickets and it ends in a property seizure or bankruptcy. "Who Appeals and Wins": sorting by "Percent contested and won" the minority neighborhoods fair very well, consisting of 5 of the top 8.

IOT_Apprentice 8y ago
Perhaps to call out the politicians who could do something about the preying on the poorer areas?