• Their race is "Asian Indian"
• They speak one of the top 10 Asian Indian languages
• They were born in India
• They are a son or daughter in a household where either the householder or spouse is classified as "Asian Indian", either originally or because of the previous three conditions

Note that our 4 conditions do not handle the problem entirely. For example, there are people whose race is Pakistani and are classified as having "American Indian" ancestry. But the conditions we have already applied make a tremendous difference, and before we narrow it down any further, we want to hear from those who are more knowledgable regarding the implications.

Next, we need to test our new data item. To do that, we create an 8 dimensional analysis, using all the pertinent data items at the appropriate (household or person) levels.

• Dimensions
  • Household Level Dimensions (we made a change based on the householder or spouse, so we need to know which households had Asian Indian householders or spouses, based on the new ModAncstry data item)
    • The number of Asian Indian householders in each household (based on the new "ModAncstry" data item). Potential values are '0-0' and '1-1'. We named the dimension HHAsInd.
    • The number of Asian Indian spouses in each household (based on the new "ModAncstry" data item). Potential values are '0-0' and '1-1'. We named the dimension SpsAsInd.
  • Person Level Dimensions (we need to look at each field that causes us to make a change, and also the result of the change)
    • Ancstry1 (Primary Ancestry). Potential values for this analysis are "American Indian", "Asian Indian", and "Other".
    • ModAncstry (Modified Ancestry). Potential values for this analysis are "American Indian", "Asian Indian", and "Other". This had better be the same as Ancstry1 in all cases except where our specified criteria requires it to be different, in which case it had better be different.
    • Pob (Place Of Birth). Potential values for this analysis are "India" and "Other".
    • Lang2 (Language). Potential values for this analysis are "AsIL" and "Other". AsIL indicates that the person speaks one of the top 10 Asian Indian languages.
    • Race. Potential values for this analysis are "Asian Indian" and "Other".
    • Relat1. This is the person's relationship to the householder. Potential values for this analysis are "Son/Daughter" and "Other".
• Volumes (numbers that are sliced-and-diced - none of the volumes are very important in this analysis, since it is to check our data modification. But we included the usual ones just because someone might find them interesting.)
  • Household Level Volumes
    • (none in this case)
  • Person Level Volumes
    • Pwgt1 (Weighted Person Count)
    • People (Unweighted Person Count)
    • PHouseholds (Weighted Household Count at the person level)

Click here to see the results of the 8 dimensional analysis. Note that (for example) in lines 33-45 our tools have converted the "American Indian" ancestry into an "Asian Indian" ancestry, based on the specified criteria.

Line 185 is interesting - this represents 54,226 people who were born in India, speak the Asian Indian languages, have an Asian Indian race, and yet they were categorized with an "American Indian" ancestry. There are also other lines that are similarly interesting.

The purpose of the above analysis is to check our data modification, but we've included it on the site because it might be interesting to see how many people were changed from American Indian to Asian Indian for each reason. If you check each line, you'll see that for each line where there was a change between Ancstry1 and ModAncstry, it was done in the cases which we had specified earlier.