Jump to content

 
Photo

Choropleth maps - a basic question

- - - - -

  • Please log in to reply
8 replies to this topic

#1
Pyzdra

Pyzdra

    Contributor

  • Validated Member
  • PipPip
  • 20 posts
  • Canada
Dear All,
Knowing only the basics about thematic map types and the different variable types that are best depicted by these maps one should be aware of the fact that it is quite advisable to standardize raw counts by enumeration unit area or by other raw variable before depicting them on a choropleth map, full stop. The reason why this is done is probably too obvious to be describe here, so I'll skip it. Having this always in the back of my mind I keep running into an issue that many maps produced by reputable institutions do show raw counts on choropleth maps and nobody seems to have any problem with it. An example is a population map by US Census <http://www.census.go...ensr01-103.pdf>, but other examples are widely available anywhere you turn. Am I being too sticky about this criterion of attribute standardization, or perhaps due to their prevalent nature some chorpleth maps depicting raw counts are accepted by the communities of map creators and readers?
Pyzdra

#2
Martin Gamache

Martin Gamache

    Ultimate Contributor

  • Associate Admin
  • PipPipPipPipPipPip
  • 980 posts
  • Gender:Male
  • Location:Washington DC
  • Interests:History of Topographic Cartography
    Topographic Mapping
    History of Relief Depiction
    Thematic Cartography
    Demographic Cartography
    Cartographic techniques, methods, and tools
    Orienteering
    Panoramic &amp; Kite Photography
  • United States

Dear All,
Having this always in the back of my mind I keep running into an issue that many maps produced by reputable institutions do show raw counts on choropleth maps and nobody seems to have any problem with it. ....  Am I being too sticky about this criterion of attribute standardization, or perhaps due to their prevalent nature some chorpleth maps depicting raw counts are accepted by the communities of map creators and readers?
Pyzdra

<{POST_SNAPBACK}>



This is quite true and I see it all the time as well and it hurts. In fact I returned my copy of the book Mapping Hacks published las year becasue of this issue. In one hack they gave specific instructions on how to do this. Shameless. The problem is that a great number of people making maps do not really have a firm grasp of cartography let alone geography.

#3
benbakelaar

benbakelaar

    Ultimate Contributor

  • Associate Admin
  • PipPipPipPipPipPip
  • 658 posts
  • Gender:Male
  • Location:North Brunswick, NJ
  • Interests:maps, information, technology, scripting, computers
  • United States

I think the map skills of many map readers (i.e. saw a map in a newspaper) is fairly low. The reason for standardizing by enumeration area is so that the variables can be compared by enumeration area. However, in the instance of a chloropleth map displaying population counts by state, isn't it true that the larger states have more population, and that is what the map is "saying" to the reader?

In other words, sometimes (often?) a map is simply (mis?)used as a way to visually display statistics, not as an areal representation......... Am I making sense?

#4
Dennis McClendon

Dennis McClendon

    Hall of Fame

  • Validated Member
  • PipPipPipPipPipPipPip
  • 1,083 posts
  • Gender:Male
  • Location:Chicago
  • Interests:map design, large-scale maps of cities
  • United States

many maps produced by reputable institutions do show raw counts on choropleth maps and nobody seems to have any problem with it

You don't say which map in the Census Atlas bothers you. The ones showing raw numbers are in a section with ones showing density and several other "normalized" variables. But if you wanted to know "where are the most populous counties in the US," wouldn't a choropleth map of raw counts be useful?

My pet peeve is that GIS users have been conditioned to display everything as choropleth maps when dot distribution would be only a little more work and often MUCH more meaningful. Someone, I think it was Tufte, points out that had Dr Snow been a GIS user we'd still be looking for the source of that London cholera epidemic. My all-time favorite is the Census Bureau's famous "nighttime" view of US population. Here's a section that I considered using as my avatar on this board:

Attached Files


Dennis McClendon, Chicago CartoGraphics
chicagocarto.com

#5
Pyzdra

Pyzdra

    Contributor

  • Validated Member
  • PipPip
  • 20 posts
  • Canada
Sorry - I thought it was a single map pdf. The map I am refering to is on p. 7: 'Number of People, 2000'.
P


You don't say which map in the Census Atlas bothers you. The ones showing raw numbers are in a section with ones showing density and several other "normalized" variables.

#6
ELeFevre

ELeFevre

    Hall of Fame

  • Associate Admin
  • PipPipPipPipPipPipPip
  • 1,049 posts
  • Gender:Male
  • Location:Louisville, Colorado USA
  • Interests:Cartography, musical instruments, reading, hiking, craft beer
  • United States

My pet peeve is that GIS users have been conditioned to display everything as choropleth maps when dot distribution would be only a little more work and often MUCH more meaningful.


I agree. A dot density map can speak volumes over a choropleth map if your primary intent is to show distribution patterns....especially at small scales and mapped at the county/sub-county level. Choroplethic maps certainly have their place, but they seem to be the "default" thematic map.



#7
Pyzdra

Pyzdra

    Contributor

  • Validated Member
  • PipPip
  • 20 posts
  • Canada
Dear Benbakelaar,
I am not too sure if I understood your argument well, but the example of the lager state with more population being depicted by, I suppose, darker shade on a choropleth map than some other smaller state with less population is exactly what should not be happening on a shaded map. What one should get in my opinion from reading a map like this is the knowledge of only the variable in question, e.g.: population (density), and NOT about the area of the state, which is likely not be be mentioned anywhere on a choropleth map depicting total population counts. In other words the population variable is an artifact of the size of the state so the knowledge that one gets on the whole map is not really a visual representation of population patterns across the study, but of some area size derivative of this pattern. Your argument however is interesting - showing counts on a choropleth map simply shows the statistics for given areas and that's that. I guess it is the matter of cartographic principal, which I guess in this case prevents the less skilled map readers from making incorrect interpretations of choropleth maps depicting raw counts. Again, your agrument cannot be fully disregarded here. I am just as confused now as I was before on this supposingly simple topic, but for now I will play conservatively and normalize my variables for choropleth maps. Thanks.

However, in the instance of a chloropleth map displaying population counts by state, isn't it true that the larger states have more population, and that is what the map is "saying" to the reader?

In other words, sometimes (often?) a map is simply (mis?)used as a way to visually display statistics, not as an areal representation......... Am I making sense?

<{POST_SNAPBACK}>



#8
benbakelaar

benbakelaar

    Ultimate Contributor

  • Associate Admin
  • PipPipPipPipPipPip
  • 658 posts
  • Gender:Male
  • Location:North Brunswick, NJ
  • Interests:maps, information, technology, scripting, computers
  • United States

I am just as confused now as I was before on this supposingly simple topic, but for now I will play conservatively and normalize my variables for choropleth maps.  Thanks.


Pyzdra, I agree, I am confused too! :) I certainly agree that from a cartographic perspective, using choropleth's to show raw counts is incorrect, since a map is by its very nature an areal representation, something you use to extract information from. So I guess it is just a matter of who the audience is... a newspaper will probably use a map to display statistics simply because it is visually interesting. And I guess my only argument is, who is to say that is an "incorrect" use of a map?

#9
Dennis McClendon

Dennis McClendon

    Hall of Fame

  • Validated Member
  • PipPipPipPipPipPipPip
  • 1,083 posts
  • Gender:Male
  • Location:Chicago
  • Interests:map design, large-scale maps of cities
  • United States

What one should get in my opinion from reading a map like this is the knowledge of only the variable in question, e.g.: population (density)


Other maps in the series show density. That doesn't mean that this one showing total population is useless. You wouldn't criticize a table of state populations for showing raw totals instead of population per hectare. Seeing that same information in map form can be useful.

If the reader's question is "where are the most densely populated states?" this map won't help. But if it is "where are the most populous states?" it's a helpful map.
Dennis McClendon, Chicago CartoGraphics
chicagocarto.com




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

-->