5 ms·

Letting the y-axis start at 0 loses too much detail.

by cygx 8y ago

rayiner 8y ago
If you’re interested in the yearly swings, convert it to a annual % change graph (and start that at 0). What you really means is that using an honest scale makes the trend look less dramatic.
- cygx 8y ago
  I just googled the data[1], removed the duplicate entry for 2013 from the Excel spreadsheet when the methodology changed (shoddy on my part, I know) and made a graph from it that was readable in the five minutes before I posted my comment. Feel free to post a better presentation of the data. [1] https://www.census.gov/data/tables/2018/demo/income-poverty/p60-263.html https://www.census.gov/data/tables/2018/demo/income-poverty/...
  - chmln 8y ago
    The parent post has a point. Five minutes or not, it is ultimately deceptive. What really is a ~22% drop since 1970s looks like a 100% drop on first sight.
    - cygx 8y ago
      By all means, make a better graph, so I can tell you what you really meant by your choice of presentation after the fact :p
      - rojobuffalo 8y ago
        For real. > terrible choice of axis > honest scale > deceptive Be kind ya'll.
        klipt 8y ago
        I didn't know OP made the graph. If I had I would have been less upfront, e.g. "I would have chosen different axes".
- camgunz 8y ago
  > If you’re interested in the yearly swings Why would they be? They certainly didn't say they are. > What you really means is that using an honest scale makes the trend look less dramatic. I think "honest" is a loaded term here. It's a well known fact that income inequality has been growing in the US for decades now, and the trend is dramatic and undeniable. Are there ways to minimize it visually in a graph? Sure, you could blow out the axes, or you could do yearly deltas which wouldn't have the line moving past a few percent ever. Both of these suggestions would flatten this phenomena, and coincidentally both have been suggested here. But the phenomena itself is not flat, and therefore it is the flat representation that, by definition, would be misleading.

Letting the y-axis start at 0 loses too much detail.

by cygx 8y ago

Letting the y-axis start at 0 loses too much detail.

If you’re interested in the yearly swings, convert it to a annual % change graph (and start that at 0). What you really means is that using an honest scale makes the trend look less dramatic.

cygx 8y ago
I just googled the data[1], removed the duplicate entry for 2013 from the Excel spreadsheet when the methodology changed (shoddy on my part, I know) and made a graph from it that was readable in the five minutes before I posted my comment. Feel free to post a better presentation of the data. [1] https://www.census.gov/data/tables/2018/demo/income-poverty/p60-263.html https://www.census.gov/data/tables/2018/demo/income-poverty/...
- chmln 8y ago
  The parent post has a point. Five minutes or not, it is ultimately deceptive. What really is a ~22% drop since 1970s looks like a 100% drop on first sight.
  - cygx 8y ago
    By all means, make a better graph, so I can tell you what you really meant by your choice of presentation after the fact :p
    - rojobuffalo 8y ago
      For real. > terrible choice of axis > honest scale > deceptive Be kind ya'll.
      - klipt 8y ago
        I didn't know OP made the graph. If I had I would have been less upfront, e.g. "I would have chosen different axes".
camgunz 8y ago
> If you’re interested in the yearly swings Why would they be? They certainly didn't say they are. > What you really means is that using an honest scale makes the trend look less dramatic. I think "honest" is a loaded term here. It's a well known fact that income inequality has been growing in the US for decades now, and the trend is dramatic and undeniable. Are there ways to minimize it visually in a graph? Sure, you could blow out the axes, or you could do yearly deltas which wouldn't have the line moving past a few percent ever. Both of these suggestions would flatten this phenomena, and coincidentally both have been suggested here. But the phenomena itself is not flat, and therefore it is the flat representation that, by definition, would be misleading.