Wednesday, July 10, 2013

Wednesday search challenge (7/10/13): How can you map world literacy rates?

In many sensemaking or analysis tasks, the hardest step is just getting started.  You might have a clear image of what needs to be done, and even what information sources you have at hand... but a huge barrier for many people is making the first move. 

Here's a scenario:  Suppose you're trying to understand the way in which literacy rates vary around the world.  You might find it easy to locate a table of world countries and their literacy rates (say, for people over the age of 15).  But THEN what do you do?  How do you convert this giant table into something that you can actually use to understand what's going on around you?  

You might have in your mind's eye something like this: 
Interactive world literacy map.  2 minutes to generate, including finding data, etc. 

This is an interactive world map I generated in a couple of minutes.  (Obviously, this is just a static image of it--but when you click on any of those dots, you'll get the pop-up with literacy rate by country, male, and female literacy rates as well, if available.)  Note that I'm NOT recommending this as a great visualization method--I'm including it here just to give a suggestion of the things you might do.  I'm sure you can do a better job!  

I know that some of you will find this an easy problem--but in my classes, I find that lots of people don't even begin an analysis like this because they have no idea how long it will take.  So the theme of this week's challenge is How to get started in visual data analysis!   

The challenge is simple:  

     Can you quickly create a visual representation
    of world literacy rates by country?  


And, of course, let us know HOW you did it and what you did to make it work.  Please include an estimate of how long it took you to go through your steps.  This is your chance to share a bit of how-to knowledge with the SearchResearch readers!  

I'll be really interested to see what kind of visual maps and charts you create.  Share them with us as well!  

Search (and visualize) on! 


25 comments:

  1. Fascinating question!

    I don't have time to answer today, so I sort of cheated. ;)

    The first result of [ literacy rate by country ] is, as I was expecting, the Wikipedia article List of countries by literacy rate. The map is already there, with the kind of visualization I had in mind: rated colors.

    So here's the map: http://goo.gl/N1Wik (is it just a very unlikely coincidence that this shortened url came out ending with the first letters of Wikipedia?)

    Had I not have a readily available map I would probably try Google Charts (maybe better suited for this than the amazingly easy to use Google Maps Engine Lite).

    ReplyDelete
    Replies
    1. Luis

      I tried with Google Maps Engine Lite. It only works with max. 100 rows and for this SearchResearch Challenge we need much more. I agree with you this new tool is very easy to use and very helpful.

      Delete
    2. Ramón, Maps Engine Lite works with much more than 100 lines, you just have to add them by hand, which is not practical for big chunks of data. Check for example this map I made (close to 300 lines at the moment): http://goo.gl/Bs4G6.

      Delete
    3. Hello, Luis, RoseMary, Debbie and fellow SearchResearchers. Thanks for your comments. It is like RoseMary says I am learning new tools and improving my knowledge in other tools, thanks!

      Luis, I did not know about the extra rows in GMEL. Your map is excellent, thanks for sharing it. I was thinking last night, that maybe adding a new layer could do the trick for the rows. Now that I read your post, I know how to do it by hand. Do you believe adding layer will do the job automatically? I'll try it!

      Maybe Google Fusion Tables are better suited for big projects and GMEL for small ones.

      Google Charts it is for everyone or just for enterprises?

      Have a great day, everybody.

      Delete
    4. Well, in fact when I wrote "Google Charts" I was actually thinking about Google Fusion Tables… :$ When I tried to add a link I found it weird not to recognize the website. Now I know why…

      On the first glance, Google Charts seems to be amazingly powerful to use on your website or blog but it requires at least some stamina dealing with all those code stuff (either that or being a programmer, which I'm not).

      It seems to be for everyone, not just for enterprises.

      Delete
  2. Good day, Dr. Russell, fellow SearchResearchers

    Searched:

    [literacy rates database]

    Found: http://www.uis.unesco.org/literacy/Pages/default.aspx

    Then remebered about past SearchResearch Challenges [literacy ]Google Tables. Searched recent data.

    Google fusion tables with data obtained

    Answer

    https://www.google.com/fusiontables/DataSource?docid=1GfPhXr2UU5_0d5l_EMg_5CLlxSyjWAJmRL6ueak

    ReplyDelete
    Replies
    1. Watching my post, noticed that my link shows data not map. So here is the correct link.

      https://www.google.com/fusiontables/embedviz?q=select+col0+from+1GfPhXr2UU5_0d5l_EMg_5CLlxSyjWAJmRL6ueak&viz=MAP&h=false&lat=21.589395831079568&lng=-59.87924962500006&t=1&z=3&l=col0&y=2&tmplt=2

      Delete
  3. I think so. I turned over SEARCH terms in my head first while I had to do something else. That done I settled on

    [chart country by country literacy rates]

    This produced lots of hits including the usual suspects that I thought would have this info: UNESCO, CIA Factbook and Wikipedia. There were others I have never heard of.

    http://www.tradingeconomics.com/ has an enormous amount of data of hundreds of factors which can be sifted and sorted in many ways. Five or Six chart options are available. Data can be downloaded to a spreadsheet program (?cost)

    http://www.indexmundi.com/g/r.aspx?v=39 has a nice green horizontal bar chart lifted from CIA Factbook

    http://world.bymap.org/LiteracyRates.html has a nice red vertical bar chart also lifted from the CIA book

    http://www.uis.unesco.org/Pages/default.aspx?SPSLanguage=EN has loads of charts

    and so it goes.

    Yes I did this quickly. Couple of minutes.

    There sure is a lot of work to be done in SE Asia and sub-Saharan Africa

    jon



    ReplyDelete
  4. I totally agree that starting is tough especially if you haven’t worked in the area of statistics, tables or charts. So step one was to do a google search to find data that could be uploaded into a chart such as

    Filetype: xls world literacy rates

    I quickly found this link http://unstats.un.org/unsd/demographic/products/indwm/default.htm

    United Nations Statistics gives me the database Table 4A an xls document of adults 15+ last updated Dec 2012.When I look at the footnotes I get some idea of how this data was compiled. I downloaded the document, saved it so I can find it, and upload to Tables/Google/Fusion

    I created a table showing all the data provided by the U.N. Then I took this table and created charts but I felt the results were unsatisfactory. So I tried to reduce the date and selected only countries below 75% literacy rate and in 2010. My charts were not what I wanted.

    So I imported the table into Google Earth. At least in Google Earth you can use Satellite View. Not really much more than the flat map. Not sure how to show you this.

    There is a website(Google Earth Outreach) that seems to be able to do what I want but it will take me some time to figure out the process. Here’s the link: http://earth.google.com/intl/en/outreach/tutorial_odk_visualize.html#viewdataearth

    I decided an interesting comparison might be to look at literacy rates of oil producing countries so I merged two tables giving us

    https://www.google.com/fusiontables/data?docid=1T5vZh1K6mGrMI3hTR3H5K82xEZ0BuuzRxQWY3jY#rows:id=1

    I wanted to show on the map how oil producing countries have a much higher literacy rate but quite frankly I have no idea where to go from here. I would like to have countries with low literacy rates and low oil production be highlighted on the map but that’s beyond me at the moment.If I have time and figure it out I will post later

    Of course we always have Wikipedia http://en.wikipedia.org/wiki/List_of_countries_by_literacy_rate

    ReplyDelete
    Replies
    1. I was able to use the Google Maps Lite Engine and produce a really rough map showing my merged data of literacy and oil production. It just a few lines but did produce the map. First exposure to this tool and I can see how it would be useful. Here's the map I think I've created..

      https://mapsengine.google.com/map/edit?mid=zhKTIzZOLROc.kQSjsCDEp2HM

      Debbie just saw your map and it looks very good. Well done.

      Delete
    2. Rosemary, unfortunately that map isn't shared with me. Please try tweaking the settings, changing the Visibility options ("Who has access") to either Public on the web or Anyone with the link.

      Delete
    3. Thanks for letting me know about the link. I had it "anyone with the link" but now changed it to Public (I hope it works). As well I fixed a few data errors.
      I mentioned in my first response I imported the data into Google Earth. If anyone can tell me how I could show that as well. What I think is probably possible is to have selected countries in indexed colors as shown in the Wiki image. I would hope as well it could be interactive so you could change based on data selected.

      https://mapsengine.google.com/map/edit?mid=zhKTIzZOLROc.kH2AuXB0DJmA

      I found this challenge really interesting because I have no experience in this area. I have learned so much from Ramon, Luis and Debbie. Thanks guys.

      Delete
    4. The link is working fine now, thanks. The colored markers are a very good visual aid.

      As to your hypothesis that "oil producing countries have a much higher literacy rate", I found it truly weird. If anyone told me there is such a correlation between oil production amount and literacy rate I would have answered 1. I truly doubt there is such a correlation and 2. if there's indeed one, I would guess the exact opposite (big oil = low literacy), as a matter of fact based solely on my (not that far from reality) prejudice that a state with vast amounts of natural resources is more prone to be ruled by the kind of people who believe that in order for them to personally keep accumulating riches, it's better for the population not to be very opinionated, which can be controlled by means of keeping them uninformed and uneducated.

      I downloaded the data from the Wikipedia article "List of countries by oil production", fed it to the already downloaded table for Literacy rate by country. Then I ran the CORREL statistical function to compare both data sets and found that, as I suspected, there is no correlation whatsoever (R=0.054596, for whoever knows this better than I do). You may check the data on a table and on a graphic here: http://goo.gl/YvH8K.

      Delete
  5. I think I got it! I had no trouble finding the literacy rates. Wikipedia had a good list of statistics and I also found them at several other sites including CIA World Factbook. The hard part for me was how to visualize the data. I remember Google maps engine lite from the mapping class and first tried to use that. I did remember that you could only have 100 cells of data (i had saved the info to an excel spreadsheet) so first I copied and pasted the first 100 cells into a new spreadsheet. I tried to upload it into maps engine lite and had no success. I went back and rewatched the tutorials and just wasn't getting it. I had never tried Google Fusion and made an attempt to use it. I didn't watch the tutorial and then realized that Dan Russell always says to look for a tool to help solve your problem. I did a search on visualize data on a map and got to the site batchgeo.com. I uploaded my excel file and in a matter of seconds had this map! I would use this tool again it was simple and easy to use! http://batchgeo.com/map/5798724e03fb0a2d7c81222ca5681f43 Search on!

    ReplyDelete
    Replies
    1. Wow, thanks for letting me know a new tool!

      As I already wrote on an answer to Ramón González, Maps Engine Lite does deal with way more than 100 rows (not cells) of data, but in a not so practical way, since above 100 you have to fill them by hand.

      Delete
    2. Luis thank you for letting me (us)know about maps engine lite. I need to go over the tutorials again because I remember thinking it had amazing capabilities when I took the mapping class. But like anything else if you don't use it you lose it!

      And, Rosemary have to comment that I love reading your responses. They really help me to see how a problem should be laid out. Ramon I also enjoy reading your comments.

      Delete
  6. The challenge is simple: Can you quickly create a visual representation of world literacy rates by country?
    agreed, this is a daunting task to visually present such a large amount of amorphous data… enjoyed Rosemary's effort and creativity in her approach and sourcing.
    I found these useful:
    bymap
    indexmundi
    It also occurred to me that sources like the UN & CIA Worldbook would be useful… also found Wolfram Alpha precise and relevant in its breakdowns (another "tool" example - as Debbie G suggested) -
    Wolfram|Alpha
    In the end, it seemed to me that to have any real value and for efficiency, each country should be looked at individually or much smaller comparative blocks…
    all that said, I ended up throwing the towel ("quick" evaporated long ago) and finally came up with this graphical representation - NOT in line with Dan's academic desires, but I travelled a little different route from the pack… ;•)
    global map/graphic


    ReplyDelete
  7. I do not find the maps to be useful.

    Spots and dots and clicking is not useful information display in this context for me.

    The bar charts I found I think are much more easily understood. That said I intend try your nifty mapmaking tips.

    Cheers

    jon on sunny Vancouver Island

    ReplyDelete
    Replies
    1. clearly data is ingested in subjective ways by different individuals with varying degrees of success - I concur that the visuals in this case often add a layer of obscurity rather than simplifying or enlightening - eye candy doesn't always illuminate.
      In practical, pragmatic terms, I found this a useful format - while acknowledging visual information formatting is a growing field…
      UNESCO
      see this example of data visualizations techniques in the academic realm -
      Börner

      Delete
    2. Remmij thanks for your perspective. Jon I failed to give some examples of what is doable with new mapping software. Here's a link that gives you a library of already existing maps which are dynamic interesting and make good use of today's technology. Have a look and see if you find these more useful. I would like to know more about creating them.

      http://www.google.com/earth/outreach/stories/showcase.html

      Delete
    3. @Unknown -- As I said, I wasn't advocating this as a great visualization. It was just one I put together quickly for demo purposes. The only reason to use this particular viz would be to be able and zoom in on tightly packed countries (e.g., the Benelux) and see what the variations are.

      Delete
    4. Why not use Wolfram Alpha? It's an excellent tool to plot down geographical data.

      http://www.wolframalpha.com/input/?i=literacy+rate&dataset=

      Delete
    5. interesting contrast in the results between yours and the one I got previously by adding 'world' to the query:
      w-l-r
      93.5% vs 84.07%…? Wolfram|Alpha is a great source/tool but susceptible to the JI;JO axiom.
      junk in;yada, yada, yada

      Delete
    6. Yes, I've noticed a variance between Alpha's data and Google's. Part of it is that they have different dates of collection -- but sometimes, I worry about Alpha's data curation. It suggests that we all *double check* our data integrity!

      Delete