3 ms·

This is the problem with quoting studies and not understanding what the stats actually mean with their relation to each other. It happens a lot with the studie

This is the problem with quoting studies and not understanding what the stats actually mean with their relation to each other. It happens a lot with the studies themselves, also, but is pretty prevalent in pundit style presentations. I'll give an interesting example from my twitter feed: "State records show that [Mr. X](1) claimed that his combined income in 1978 and 1979 was negative $3.8 million, allowing him to pay no taxes. A few years earlier, he had told the NYT he was worth more than $200 million." This is combining two factual statements with the intent to link them when the linkage is actually not there. People pay taxes on income not net worth(2). You spotted a stat that is closer, but actually still not directly comparable without further information and fixing the wording to indicate the same action in the same way. This doesn't even add the cases where studies have different basic facts about the observed and people trying to tie them together in problematic ways. Their starting wall example is also pretty foolish given the actual destruction of vibrant black neighborhoods in urban renewal projects starting with 1949 and the Housing Act. Why build a wall when you can take the land? 1) yeah yeah, we all know who the quote is about, but I have no wish to argue with BOTH sides again. 1800 here we come. 2) well, ok, there are some specific stuff, but its not really relevant to the setup. Also, yes, the graphics are horrible to the point of having a minecraft / 8-bit fetish is getting in the way of actually indicating what the information is.

srtjstjsj 10y ago
Please with draw your incorrect aspersions. Your example is not the same as the OP article, that faithfully reproduces the source data at: http://www.pewtrusts.org/~/media/legacy/uploadedfiles/wwwpewtrustsorg/reports/economic_mobility/pewsharkeyv12pdf.pdf http://www.pewtrusts.org/~/media/legacy/uploadedfiles/wwwpew...
- protomyth 10y ago
  PoorBloke123 gave a fine example that I expanded on with an explanation of what typically happens. I then gave a recent example of the same thing one step farther. I don't really have time or any want to read the source other than to comment that I have always enjoyed the fact they know how to use their graphic art department properly. Since, its an urban study, my only real concern is people using it to prove things about rural areas which is the 3rd in line after the first two I expanded upon. [edit - from the I really should be doing something else department] If you based all the data from figure 3 you miss quite a few quotes about "Overall, because of the uncertainty in the estimates, the patterns do not provide a precise answer as to whether neighborhood poverty plays a substantial role in explaining black-white gaps in upward and downward mobility, nor do they attempt to account for other factors that could be driving the results. The next set of analyses address these issues.". If I were writing a summary I would have started with the ending and worked backwards because the earlier stuff is about building their case. Given that the following figures have a lot of comments about not enough study and lack of statistical significance quoting the earlier stuff without coaching it the same language is pretty poor reporting and lessens the conclusions and build-up that Pew was very methodological about. Pew did a really fine job of answering holes with further points on the way to their conclusions. It will be interesting to see who rebuts (there is always one) and with what reasoning. In fact, taking the study and showing someone to compare it to the article would give an actually fine comparison between a well thought out and sober, easy to understand report and a 8-bit wonder with a shallow an uneven understanding of the source material.

srtjstjsj 10y ago

Please with draw your incorrect aspersions. Your example is not the same as the OP article, that faithfully reproduces the source data at: http://www.pewtrusts.org/~/media/legacy/uploadedfiles/wwwpewtrustsorg/reports/economic_mobility/pewsharkeyv12pdf.pdf http://www.pewtrusts.org/~/media/legacy/uploadedfiles/wwwpew...

protomyth 10y ago
PoorBloke123 gave a fine example that I expanded on with an explanation of what typically happens. I then gave a recent example of the same thing one step farther. I don't really have time or any want to read the source other than to comment that I have always enjoyed the fact they know how to use their graphic art department properly. Since, its an urban study, my only real concern is people using it to prove things about rural areas which is the 3rd in line after the first two I expanded upon. [edit - from the I really should be doing something else department] If you based all the data from figure 3 you miss quite a few quotes about "Overall, because of the uncertainty in the estimates, the patterns do not provide a precise answer as to whether neighborhood poverty plays a substantial role in explaining black-white gaps in upward and downward mobility, nor do they attempt to account for other factors that could be driving the results. The next set of analyses address these issues.". If I were writing a summary I would have started with the ending and worked backwards because the earlier stuff is about building their case. Given that the following figures have a lot of comments about not enough study and lack of statistical significance quoting the earlier stuff without coaching it the same language is pretty poor reporting and lessens the conclusions and build-up that Pew was very methodological about. Pew did a really fine job of answering holes with further points on the way to their conclusions. It will be interesting to see who rebuts (there is always one) and with what reasoning. In fact, taking the study and showing someone to compare it to the article would give an actually fine comparison between a well thought out and sober, easy to understand report and a 8-bit wonder with a shallow an uneven understanding of the source material.