I have shapefiles representing a country, states of that country or towns etc... For each shapefile there is associated data with them.

The country shapefile has 100 rows and maps to .csv files with 100 rows, same goes for towns, states etc. Basically it gives me a flexibility at how much level I can experiment with.

Now, let's say I chose the states data and I wanted to find whether folks with education are more likely to get jobs than those who don't, the first and easy part was to join the shapefile to its .csv file that contains data about employment.

After that I simply got confused.. the CSV files contains raw data and how to use them confuses me, an example (fake data):

S1  STA     Alabama      4345   1003    100000    2000       1000
S2  STA     Alaska       1456   8343    8888      4000
S3  STA     Arizona      90000  1247    1000
S4  STA     Arkansas     18831  1213    0

How do I use the data above? I don't think a select by attributes will do me any good as it shows just a blue border, and I don't know how I could possible query that? It doesn't give me any specific data, it is just an aggregate.

I think my best option is to use either categories or quantities symbology.

However, I'm not sure how to get meaningful data. If I was to use quantities (I think quantities is the way to go), do I simply normalize people who are employed against people having education? How would I display the output.. density dots or gradient symbology/color?

  • Lews, You cannot (validly) answer questions like yours with data like this. Mere counts of employed and unemployed along with a single "Education" field (whatever that might mean) tell you nothing about the experiences of people looking for jobs and how they might have varied by educational level. You need far more detailed data. In the US, start at the BLS website for instance.
    – whuber
    Commented Dec 16, 2012 at 21:54
  • @whuber I know.. but that's the data I was given and can only use. Everything is just count. Employed, Unemployed, At Work, Education etc.. all counts. Hence my confusion. I just used graduated colors and normalized certain fields to show the data... but as you say it isn't valid. Commented Dec 16, 2012 at 23:27

1 Answer 1


Esri have a good PDF Publication on Normalizing census data

Good Advice comes from the document:

Know data before normalizing it. Normalizing unrelated data is like mixing apples and oranges. It makes fruit salad, not good analysis.



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.