Comments on: Can You Improve this Graph Showing Suicide Rates in Japan? http://flowingdata.com/2008/07/30/can-you-improve-this-graph-showing-suicide-rates-in-japan/ Strength in Numbers Wed, 23 May 2012 14:19:57 +0000 hourly 1 http://wordpress.org/?v=3.3.2 By: Annelie http://flowingdata.com/2008/07/30/can-you-improve-this-graph-showing-suicide-rates-in-japan/#comment-13271 Annelie Thu, 07 Aug 2008 02:33:49 +0000 http://flowingdata.com/?p=875#comment-13271 Also, even by normalizing the unemployment or the suicide rate, the sampling intervals are obviously different by a factor of five. Also, even by normalizing the unemployment or the suicide rate, the sampling intervals are obviously different by a factor of five.

]]> By: Annelie http://flowingdata.com/2008/07/30/can-you-improve-this-graph-showing-suicide-rates-in-japan/#comment-50910 Annelie Thu, 07 Aug 2008 02:33:00 +0000 http://flowingdata.com/?p=875#comment-50910 Also, even by normalizing the unemployment or the suicide rate, the sampling intervals are obviously different by a factor of five. Also, even by normalizing the unemployment or the suicide rate, the sampling intervals are obviously different by a factor of five.

]]> By: Robert Morton http://flowingdata.com/2008/07/30/can-you-improve-this-graph-showing-suicide-rates-in-japan/#comment-50909 Robert Morton Thu, 31 Jul 2008 18:28:00 +0000 http://flowingdata.com/?p=875#comment-50909 Howdy from Tableau Software! I really enjoy these challenges so I thought I'd participate this time. You can see my blog post here: http://www.tableausoftware.com/improve-this-graph-001 Or just a direct link to my visualization: http://www.tableausoftware.com/files/Japan_suicide_rates_0.png Howdy from Tableau Software! I really enjoy these challenges so I thought I’d participate this time. You can see my blog post here:

http://www.tableausoftware.com.....-graph-001

Or just a direct link to my visualization:

http://www.tableausoftware.com.....ates_0.png

]]> By: Darin http://flowingdata.com/2008/07/30/can-you-improve-this-graph-showing-suicide-rates-in-japan/#comment-50908 Darin Thu, 31 Jul 2008 05:34:00 +0000 http://flowingdata.com/?p=875#comment-50908 lol, just as I click submit, I went to Jake's graph which has % change. Jake's da man! =) lol, just as I click submit, I went to Jake’s graph which has % change. Jake’s da man! =)

]]> By: Darin http://flowingdata.com/2008/07/30/can-you-improve-this-graph-showing-suicide-rates-in-japan/#comment-50907 Darin Thu, 31 Jul 2008 05:33:00 +0000 http://flowingdata.com/?p=875#comment-50907 Epidemics occur when there is a significant increase (or decrease) of an event. One of the ways to present this graph would be the story that unemployment and suicide are highly correlated. The title says, "Suicide Epidemic in Japan"; this would suggest there is an 'outbreak'. If so, there should be some attempt within the graph to show that *each* increase in unemployment corresponds to an increase suicide and conversely i.e. *each* decrease in unemployment corresponds to an specific decrease. Since our data is yearly (not very fine considering how suicide and unemployment is measured, but hey, at least the OECD is gathering something), I think some sort of year-over-year % change would not only tell the story better, but would help to show or refute the correlation. Epidemics occur when there is a significant increase (or decrease) of an event. One of the ways to present this graph would be the story that unemployment and suicide are highly correlated. The title says, “Suicide Epidemic in Japan”; this would suggest there is an ‘outbreak’. If so, there should be some attempt within the graph to show that *each* increase in unemployment corresponds to an increase suicide and conversely i.e. *each* decrease in unemployment corresponds to an specific decrease.

Since our data is yearly (not very fine considering how suicide and unemployment is measured, but hey, at least the OECD is gathering something), I think some sort of year-over-year % change would not only tell the story better, but would help to show or refute the correlation.

]]> By: Jake http://flowingdata.com/2008/07/30/can-you-improve-this-graph-showing-suicide-rates-in-japan/#comment-50906 Jake Wed, 30 Jul 2008 22:09:00 +0000 http://flowingdata.com/?p=875#comment-50906 Revised.... http://bp3.blogger.com/_8rpY5fQK-UQ/SJDmQv17K0I/AAAAAAAAClg/5MBOFMkwoJw/s1600-h/SPIN2.png Revised….

http://bp3.blogger.com/_8rpY5f...../SPIN2.png

]]> By: Jon Peltier http://flowingdata.com/2008/07/30/can-you-improve-this-graph-showing-suicide-rates-in-japan/#comment-50905 Jon Peltier Wed, 30 Jul 2008 20:56:00 +0000 http://flowingdata.com/?p=875#comment-50905 Jake - (a) You should multiply not by population but by potentially employed population. At least in the US, only working age people who want to work can be considered unemployed. (b) The long term unemployment rate is not percentage of potential workers who are unemployed. It is the percentage of unemployed people who have been unemployed for a year or longer. This makes the math more complicated since you also need to factor in the "reguler" unemployment rate. Jake -
(a) You should multiply not by population but by potentially employed population. At least in the US, only working age people who want to work can be considered unemployed.
(b) The long term unemployment rate is not percentage of potential workers who are unemployed. It is the percentage of unemployed people who have been unemployed for a year or longer. This makes the math more complicated since you also need to factor in the “reguler” unemployment rate.

]]> By: Jake http://flowingdata.com/2008/07/30/can-you-improve-this-graph-showing-suicide-rates-in-japan/#comment-50904 Jake Wed, 30 Jul 2008 18:08:00 +0000 http://flowingdata.com/?p=875#comment-50904 2) Took population and multipled by suicides per 100,000 and divided that # by 100,000 to get # of suicides 2) Took population and multipled by suicides per 100,000 and divided that # by 100,000 to get # of suicides

]]> By: Jake http://flowingdata.com/2008/07/30/can-you-improve-this-graph-showing-suicide-rates-in-japan/#comment-50903 Jake Wed, 30 Jul 2008 18:06:00 +0000 http://flowingdata.com/?p=875#comment-50903 Hi Jon- 1) Took population and multiplied by the unemployment rate to get # of unemployed. 2) Took population and by suicides per 100,000 / 100,000 3) Divided # of suicides by # of unemployed Hi Jon-

1) Took population and multiplied by the unemployment rate to get # of unemployed.
2) Took population and by suicides per 100,000 / 100,000
3) Divided # of suicides by # of unemployed

]]> By: Nathan Yau http://flowingdata.com/2008/07/30/can-you-improve-this-graph-showing-suicide-rates-in-japan/#comment-50902 Nathan Yau Wed, 30 Jul 2008 15:25:00 +0000 http://flowingdata.com/?p=875#comment-50902 Jake used the same numbers, didn't he? He just didn't use the unemployment rates for the years there wasn't data for suicide rate. Jake used the same numbers, didn’t he? He just didn’t use the unemployment rates for the years there wasn’t data for suicide rate.

]]>