Slightly OT: Graphic of Unemployment in the United States
in [SAS]
|
From: "Vandenbroucke, David A" on 2 Dec 2009 13:28 It seems to me that the people who object to this display would want to refine it to the point where you'd need a couple pages of documentation to understand what you're looking at. The point of graphs is to present information quickly. It's certainly fun when you come across an intricate visual presentation that rewards careful study, but most people are going to spend no more than a few seconds watching this presentation. As others have said, the point is to show the geographic spread of unemployment rates. As an economist, I found it interesting to see which areas turned dark first and which never did. It's true that the map doesn't tell you a lot about the total number of people who are unemployed, but, as I see it, that isn't its purpose: it doesn't show you how many people like chocolate, either. It uses geographic building blocks that people are used to, in the size and shape that they recognize. If one squints a bit, one can watch what happens to specific cou! nties, in regions with which one is familiar. I wouldn't call it the best graphic ever, but I think it's clever and useful. I wish I'd thought of doing it. Dav Vandenbroucke Senior Economist U.S. Dept. HUD david.a.vandenbroucke(a)hud.gov 202-402-5890 I disclaim any disclaimers.
From: Dale McLerran on 2 Dec 2009 13:38 Paul, I was thinking the same thing and had just located a pdf file with a number of examples where geographical boundaries are distorted so that area size represents population or some other size measure. I believe that such projections are generally attributable to a cartographer named Walter Tobler (although I could be mistaken on this point). Here is the pdf with other similar examples: http://www.santafe.edu/~mgastner/publications/thesis_compr.pdf Note that sometimes the shape of these projected map is itself the salient feature of the graphic. Figure 2.15 on page 35 of the thesis which I have referenced shows two maps, one in which the size of each state is proportional to energy consumption and the other in which the size of each state is proportional to energy production. Shape differences draw your attention to states which are big producers and low consumers (and vice versa). Wyoming and West Virginia immediately stand out for high energy production and low energy consumption. I would note that these two maps are imperfect themselves. The maps are not scaled to have the same total area, although they should have been. If the plots were scaled to have the same total area, then one could immediately identify states which are energy exporters, importers, or energy neutral. For example, my eyeball inflation of the energy producers plot to place it on the same scale as the energy consumers plot would suggest that Pennsylvania might be energy neutral. It should be noted that these plots are more difficult to construct. We probably would not want to map counties using such projections because of the complexity of the algorithms for reshaping the plot boundaries. It would probably be preferable to reduce data to a state level. But then we would lose much of the information about NYC as it would be absorbed into a larger geographical region. However, the effect of unemployment in NYC would be felt in both the size of the state of NY as well as in the state unemployment rate. But just for the record, I do feel that the graphic which Art pointed to does convey quite well a huge increase in unemployment and the attendant recession. It is not meant to show precisely what the unemployment rates are across the United States. It is meant to show the increase in unemployment across the United States. In using an animated graphic, the plot conveys what is intended. Dale --------------------------------------- Dale McLerran Fred Hutchinson Cancer Research Center mailto: dmclerra(a)NO_SPAMfhcrc.org Ph: (206) 667-2926 Fax: (206) 667-5977 --------------------------------------- --- On Wed, 12/2/09, Choate, Paul(a)DDS <pchoate(a)DDS.CA.GOV> wrote: > From: Choate, Paul(a)DDS <pchoate(a)DDS.CA.GOV> > Subject: Re: Slightly OT: Graphic of Unemployment in the United States > To: SAS-L(a)LISTSERV.UGA.EDU > Date: Wednesday, December 2, 2009, 8:36 AM > Haven't read the whole thread, so > pardon if this was mentioned .... > > This is similar to the red vs. blue election maps - > population > cartograms are much better at proportionally representing > geographic > based information. > > Mark Newman of the Department of Physics and Center for the > Study of > Complex Systems at University of Michigan has a great web > page on this. > > http://www-personal.umich.edu/~mejn/election/2008/ > > > Paul Choate > DDS Data Extraction > (916) 654-2160 > -----Original Message----- > From: SAS(r) Discussion [mailto:SAS-L(a)LISTSERV.UGA.EDU] > On Behalf Of Ben > Powell > Sent: Wednesday, December 02, 2009 7:38 AM > To: SAS-L(a)LISTSERV.UGA.EDU > Subject: Re: Slightly OT: Graphic of Unemployment in the > United States > > I agree, brings to mind the nerve density maps of the human > body, where > the > finger tips are oversize, > > Rgds >
From: Jonathan Goldberg on 2 Dec 2009 14:02 I completely disagree with: "What the author identifies as "the best graph ever made" is one that is so confusing (to me) that I would never try to replicate the methods used as I'm sure it would totally confuse at least some readers." It helps to see it in good resolution and color. The graph is indeed amazing in packing a lot of information onto one page in clear, easy to decode form. Like Wainer and Tufte (AKA god) I think that in almost all cases more information is better if it can be done clearly. As a general rule, concerning graphics, Tufte is right. Think hard before disagreeing with him.
From: Dale McLerran on 2 Dec 2009 14:58 I agree that Minard's plot of the Napolean Russian campaign is a highly informative plot. It deserves accolades. However, it is somewhat complex. Would it (or some similar creative graphic) be good for all audiences? Probably not. Often, it is beneficial to present graphics which are within the audiences common experience - graphics which can be readily understood with only a brief look. I think the take-home-point is "Know thine audience." Dale --------------------------------------- Dale McLerran Fred Hutchinson Cancer Research Center mailto: dmclerra(a)NO_SPAMfhcrc.org Ph: (206) 667-2926 Fax: (206) 667-5977 --------------------------------------- --- On Wed, 12/2/09, Jonathan Goldberg <jgoldberg(a)BIOMEDSYS.COM> wrote: > From: Jonathan Goldberg <jgoldberg(a)BIOMEDSYS.COM> > Subject: Re: Slightly OT: Graphic of Unemployment in the United States > To: SAS-L(a)LISTSERV.UGA.EDU > Date: Wednesday, December 2, 2009, 11:02 AM > I completely disagree with: > > "What the author identifies as "the best graph ever made" > is one that is so > confusing (to me) that I would never try to replicate the > methods used as > I'm sure it would totally confuse at least some readers." > > It helps to see it in good resolution and color. The > graph is indeed > amazing in packing a lot of information onto one page in > clear, easy to > decode form. > > Like Wainer and Tufte (AKA god) I think that in almost all > cases more > information is better if it can be done clearly. > > As a general rule, concerning graphics, Tufte is > right. Think hard before > disagreeing with him. >
From: Mary on 2 Dec 2009 15:26
Yes, and what the graphic of unemployment is showing is an important thing- that unemployment has spread to areas that aren't normally affected, though it would have been nice to have more categories towards the top of the unemployment numbers, such as showing the areas with higher than 10% unemployment more differentiated from those below 8.5% unemployment. As for the rural versus urban divide, yes, there is a population issue, but there are other explainations as well, such as: 1. Rural people, even poor people, are more likely to have cars than urban people and thus can drive to more areas to find a job, whereas inner city people may not even be able to get to the suburbs of the metropolitan area in which they live. 2. Rural people have always traveled long distances for other things like going to the doctor or taking children to school, thus are more amenable to traveling long distances if need be to get to a job. 3. People in extremely remote areas are more likely to be self-employed (farming, ranching), or can live on less income for a while from things like gardening. 4. People in rural areas view taking unemployment, welfare as a sign of personal failure and thus are more likely to try to find work of some kind than declare themselves unemployed. However, note this does not mean people in rural areas are richer than people who live in cities; in particular their transportation expenses can be a very large part of their budget. However, the graph did show that unemployment is reaching into many more areas than just the traditional inner city/suburb divide; that did make it interesting; that so many people are picking up and changing the states in which they live indicates a seriousness to the situation not seen since the depression. -Mary --- stringplayer_2(a)YAHOO.COM wrote: From: Dale McLerran <stringplayer_2(a)YAHOO.COM> To: SAS-L(a)LISTSERV.UGA.EDU Subject: Re: Slightly OT: Graphic of Unemployment in the United States Date: Wed, 2 Dec 2009 11:58:25 -0800 I agree that Minard's plot of the Napolean Russian campaign is a highly informative plot. It deserves accolades. However, it is somewhat complex. Would it (or some similar creative graphic) be good for all audiences? Probably not. Often, it is beneficial to present graphics which are within the audiences common experience - graphics which can be readily understood with only a brief look. I think the take-home-point is "Know thine audience." Dale --------------------------------------- Dale McLerran Fred Hutchinson Cancer Research Center mailto: dmclerra(a)NO_SPAMfhcrc.org Ph: (206) 667-2926 Fax: (206) 667-5977 --------------------------------------- --- On Wed, 12/2/09, Jonathan Goldberg <jgoldberg(a)BIOMEDSYS.COM> wrote: > From: Jonathan Goldberg <jgoldberg(a)BIOMEDSYS.COM> > Subject: Re: Slightly OT: Graphic of Unemployment in the United States > To: SAS-L(a)LISTSERV.UGA.EDU > Date: Wednesday, December 2, 2009, 11:02 AM > I completely disagree with: > > "What the author identifies as "the best graph ever made" > is one that is so > confusing (to me) that I would never try to replicate the > methods used as > I'm sure it would totally confuse at least some readers." > > It helps to see it in good resolution and color. The > graph is indeed > amazing in packing a lot of information onto one page in > clear, easy to > decode form. > > Like Wainer and Tufte (AKA god) I think that in almost all > cases more > information is better if it can be done clearly. > > As a general rule, concerning graphics, Tufte is > right. Think hard before > disagreeing with him. > |