analyze the national survey of oaa participants (nps) with r

the united states administration on aging (aoa) spends about a billion dollars on community services and meals for elderly americans every year.  enacted as part of lyndon johnson's great society program, title iii of the older americans act funds state agencies to provide rides to the doctor, light housework, adult day care, and piping-hot dinner deliveries to help seniors stay in the neighborhood (and out of institutions).  and to keep federally-funded programs like meals on wheels accountable for those taxpayer dollars, lawmakers and researchers have a series of nationally representative surveys that ask program beneficiaries (those older americans) about their experience in receiving the services.

despite their modest agenda, these surveys include smart, straightforward documentation worth reading if you care about more than the sample frame.  but in a sentence: the aoa samples about three hundred area agencies on aging, and within those sampled agencies, randomly selects beneficiaries (grandmom and grandpop) to include in the survey microdata just like any good cluster sample.  before you worry through the technical documentation, browse the questionnaire to make sure they're asking whatever you're trying to answer.  i've done my best to get your tires off the road asap.  this new github repository contains three scripts:


download all microdata.R

analysis examples.R
  • load an example single year of data
  • construct a fay's adjusted replication survey design
  • demonstrate the beauty, simplicity, and power of survey analysis with the r language

replication.R



click here to view these three scripts



for more detail about the national survey of oaa participants, visit:


notes:

the administration on aging's agidnet website hosts quite a lot of easy-to-access statistics (and not just from this survey), including their custom data table-maker.  before you sink your teeth into the microdata, play around with what's online.  after all, online query tools will be overtaking all non-r statistical languages any minute now.


confidential to sas, spss, stata, and sudaan users:  why are you bragging about the great deal you got on your new granite kitchen countertops when the best restaurant in town is free?  time to transition to r. :D