description of datasets?

[nicholasjhorton]

It appears as if both datasets in the package have the same description:

NHANES NHANES 2009-2012 with adjusted weighting
NHANESraw NHANES 2009-2012 with adjusted weighting

This isn't true for raw, is it?

[rpruim]

It's one document for both data sets and includes the following:

NHANES and NHANESraw each include 75 variables available for the 2009-2010 and 2011-2012 sample years. NHANESraw has 20,293 observations of these variables plus four additional variables that describe that sample weighting scheme employed. NHANES contains 10,000 rows of data resampled from NHANESraw to undo these oversampling effects. NHANES can be treated, for educational purposes, as if it were a simple random sample from the American population.

I don't really want to duplicate the help file -- it's better to document these together -- but we could adjust the description if you want to propose a change.

[rpruim]

One other note from the help file is this

Weighting Variables (NHANESraw only)

WTINT2YR, WTMEC2YR, SDMVPSU, SDMVSTRA
Sample weighting variables. For more details see one of the following.

http://www.cdc.gov/Nchs/tutorials/environmental/orientation/sample_design/index.htm

http://www.cdc.gov/nchs/nhanes/nhanes2009-2010/DEMO_F.htm#WTINT2YR and

http://www.cdc.gov/nchs/nhanes/nhanes2011-2012/DEMO_G.htm#WTINT2YR

So in NHANESraw the data includes the weighting information and in NHANES the weights have been used to create a sample that mimics a random sample from the underlying population by taking the weights into account.

[nicholasjhorton]

The documentation looks great: I have no issues there. It's just the default output of:

data(package="NHANES")

that is not optimal.

Nick

On Jan 29, 2016, at 11:24 AM, Randall Pruim notifications@github.com wrote:

It's one document for both data sets and includes the following:

NHANES and NHANESraw each include 75 variables available for the 2009-2010 and 2011-2012 sample years. NHANESraw has 20,293 observations of these variables plus four additional variables that describe that sample weighting scheme employed. NHANES contains 10,000 rows of data resampled from NHANESraw to undo these oversampling effects. NHANES can be treated, for educational purposes, as if it were a simple random sample from the American population.

I don't really want to duplicate the help file -- it's better to document these together -- but we could adjust the description if you want to propose a change.

—
Reply to this email directly or view it on GitHub.

Nicholas Horton
Professor of Statistics
Department of Mathematics and Statistics, Amherst College
Box 2239, 31 Quadrangle Dr
Amherst, MA 01002-5000
https://www.amherst.edu/people/facstaff/nhorton

[nicholasjhorton]

Here's the output of: data(package="NHANES")

Data sets in package ‘NHANES’:

NHANES NHANES 2009-2012 with adjusted weighting
NHANESraw NHANES 2009-2012 with adjusted weighting

Should there be a different title for NHANESraw?

[rpruim]

What you are seeing is the title of the (common to both data sets) help page. So as long as we document them together (which I think makes sense) that text will be the same for both.

We can, however, change the text to anything (short) we like.

[nicholasjhorton]

perhaps the weighting (and the raw) could be referenced in the package description?

"Body Shape and related measurements from the US National Health and Nutrition Examination Survey (NHANES, 1999-2004). See http://www.cdc.gov/nchs/nhanes.htm for details."