[acs-r] accessing margin of error in acs.r

[arilamstein@gmail.com]

Thanks Ezra.

Also, it looks like I made a typo in my email. When I wrote:

"Also, it would be great if someone could provide guidance about which of
these metrics (margin of error vs *standard deviation*) is better to report
along with the esimate."

I meant to write "standard error".

I will plan to do what you suggested, and report the 90% margin of error.


On Mon, Feb 22, 2016 at 5:12 PM, Ezra Haber Glenn <eglenn at mit.edu> wrote:

>
> The package keeps acs-class objects with geographies, estimates, and
> standard.errors (not standard deviations -- lookup both to see the
> difference).  When printing out such an object, the package
> automatically converts the standard error to a 90% margin of error, by
> multiplying it by 1.645.  (For other MOEs, you'd need to use a
> different multiplier.)
>
> So for your table, you can use 1.645*standard.error(county_data) to
> get the 90% MOEs.
>
> --Ezra
>
> At Mon, 22 Feb 2016 17:03:13 -0800, arilamstein wrote:
> >
> > I am trying to create a data.frame of acs data with 3 columns: region
> (for example county
> > fips), estimate and margin.of.error. I'm starting with this code:
> >
> > counties = geo.make(state="*", county="*")
> > county_data = acs.fetch(geography=counties, table.number="B19301")
> > head(county_data)
> >
> > I can construct the county fips code from county_data at geography
> >
> > I can construct the estimates from estimate(county_data).
> >
> > I am not sure how to get the margins of error. If I type
> "head(county_data)" the estimates
> > are printed along with the margins of error. However, there does not
> seem to be a
> > "margin.of.error" function to strip out just the margins of error:
> >
> > >head(county_data)
> > ACS DATA:
> >  2007 -- 2011 ;
> >   Estimates w/90% confidence intervals;
> >   for different intervals, see confint()
> >                         B19301_001
> > Autauga County, Alabama 25035 +/- 916
> > Baldwin County, Alabama 27217 +/- 591
> > Barbour County, Alabama 15899 +/- 883
> > Bibb County, Alabama    18462 +/- 1267
> > Blount County, Alabama  21185 +/- 862
> > Bullock County, Alabama 20678 +/- 3797
> >
> > However, there does seem to be a standard.error function which provides
> the  standard
> > errors:
> >
> > > head(standard.error(county_data))
> >                         B19301_001
> > Autauga County, Alabama   556.8389
> > Baldwin County, Alabama   359.2705
> > Barbour County, Alabama   536.7781
> > Bibb County, Alabama      770.2128
> > Blount County, Alabama    524.0122
> > Bullock County, Alabama  2308.2067
> >
> > Can someone help me understand this discrepancy?
> >
> > Also, it would be great if someone could provide guidance about which of
> these metrics
> > (margin of error vs standard deviation) is better to report along with
> the esimate.
> >
> > My situation is that I have a mapping package that uses the acs package
> to get census data
> > estimates as a data.frame and then maps it. I would like to improve the
> package by making
> > the data.frames also have an error metric. But I'm confused about
> whether margin of error
> > or standard error is the best metric to include.
> >
> >
>
> --
> Ezra Haber Glenn, AICP
> Department of Urban Studies and Planning
> Massachusetts Institute of Technology
> 77 Massachusetts Ave., Room 7-337
> Cambridge, MA 02139
> eglenn at mit.edu
> http://dusp.mit.edu/faculty/ezra-glenn |
> http://eglenn.scripts.mit.edu/citystate/
> 617.253.2024 (w)
> 617.721.7131 (c)
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/pipermail/acs-r/attachments/20160222/8d386344/attachment.html