3 ms·
I have not read that book. I am, however, a data analyst with very strong opinions about dark patterns. My perception is that you're seeing an attempt to misd
by existencebox 8y ago
I have not read that book. I am, however, a data analyst with very strong opinions about dark patterns. My perception is that you're seeing an attempt to misdirect where there is only, at worst, lax handling of data, and writing off a worth-discussing core concept as a result. (there IS one aspect of the chart I'd complain about but it's not the 50-70 bounds) Here's why.
1. If I wanted to lie with statistics, a single 50-70 chart doesn't really "do the trick." You want to have a 50-70 chart alongside a 40-50 chart, ideally without clearly labeled Y axes, or something like that. Really distorting the trend comparison is where the money is at.
2. The aspect I would have called out is the fact that they're citing a trend inversion with less than a year's of data. this could be a season effect, natural volatility, any number of other things, and this single trendline is not a hugely confidence-building piece of evidence on its own.
That being said, the facts of the article stood up to my understanding, and is something I've long taken issue with in our recovery from 08 (Highly uneven distribution of gains and recovery); so I'd hope that discussion at least not be lost in the noise.
- dandare 8y agoAre you the author? I will give you the benefit of Hanlon's razor and read the article.
- existencebox 8y agoNo, not the author, sorry for any confusion. I can't say it's always clear when it's malice or sloppiness, but I can usually "smell" an agenda, and the chart doesn't even really seem to be the key lynchpin for arguing that agenda, just the most visible aspect. I'll grant that the author didn't use nearly as much data as I would have liked, but that's why I hope there can be a conversation. (To make this more substantive, a typical comparison is a side by side charting of labor productivity vs labor share of wealth, I just found the "stock ownership" angle to be another interesting datapoint)
- sharkmerry 8y ago>> The aspect I would have called out is the fact that they're citing a trend inversion with less than a year's of data. this could be a season effect, natural volatility, any number of other things, and this single trendline is not a hugely confidence-building piece of evidence on its own. I must be lost, what trend inversion are they citing with less than a years data?
- existencebox 8y agoNo, you're correct, and I misparsed the chart _entirely_, thanks for calling out; for some insane reason I thought it was a monthly cadence (since I think I glanced and saw "april" and elided the fact that it was year incremented.) I'd stand by the crux of my statement that I don't think the chart was intentionally misleading, however, with one fewer complaints about the nature of the data :)