Question

Status:Closed    Asked:May 15, 2015 - 09:44 PM

Probability of duplicates if using multi-year data from the ACS?

I apologize if this seems like a pretty basic question, but I'm using ACS data from 2008 to 2013 in order to expand a specific subsample that is rather rare. My question is when researchers use multiple years of the survey to do statistical analysis and estimations, do they worry about households that may be a part of multiple surveys?

 
Do you have the same question? Follow this Question
 

Staff Answer

avatar

Tim_Moreland

Staff

The short answer is not to worry about repeated households. Once a household has been selected for the ACS that household is removed from the sampling frame for at least 5 years. Since your sample includes 6 years, it is technically possible that a household could be repeated; however, the probability of a household being selected in the 1% national random sample for 2008 and then selected again in 2013 is only about 0.01%, which is much too small to affect your analysis. Even if there were a small number of repeated households in your sample, no variables exist to identify such households.


Hope this helps.

 

May 18, 2015 - 01:42 PM

0
0
Report it

OTHER QUESTIONS NEEDING ANSWERS

what weights to use if I merge multiple years of ACS data?
Login   |   Register

Recently Active Members

View More »

Share |