Sample Design



Introductory Files



Web-Accessible Data




Content Outlines

Wave 1

Wave 2

Wave 3

Codebooks and Data Access

Wave 1

Wave 2

Wave 3





Working Papers










In an effort to facilitate the process of downloading and analyzing our data, we have compiled a list of the most frequently asked questions sent to us through NSFHHELP. Please look over the FAQ list before contacting NSFHHELP for assistance.

I'm using NSFH2 and/or NSFH3 data and can't find data on the respondent's race.

This information was collected at wave 1 (variable M484). Merge your data with this variable by joining case ids.

I can't find the respondent's gender, educational attainment, or marital status in NSFH3.

Gender and marital status variables for the main respondents are SEX_A and MS_A respectively and they are located in the household roster file. There is a constructed variable, COMPLED3 for educational attainment for the main respondents in the main interview file. No educational attainment variable was constructed for the spouses.

What variable should I use to find the sex of the spouse/partner in NSFH1?

R2SEX is the best measure for sex of the spouse/partner. There should be no same sex spouses/partners in the NSFH data.

How can I tell which wave the union history information is from in the "Marriage History" or "Union History" data files on the NSFH3 page?

Each wave asks for information that occurred after the last wave and up to the current wave. You can therefore identify the wave the information came from by selecting union dates that occurred before or after the interview dates of other waves.

What variable is used to merge data from NSFH1 with data from NSFH2 or NSFH3?

The ID variable (which is also referred to as case number, or case ID) in columns 1-5 of each data file. In many files the ID number is available in both character and numeric form. Please note that in order to successfully merge your files using the ID number, you must check that all of your ID variables are in the same format (i.e. either all character or all numeric).

Within NSFH2, what variable is used to merge data from the main respondent data file with data from the spouse or ex-spouse file?

The ID variable in columns 1-5 of each data file. Column 6 of these files indicates which respondent. Note that some of the variable names are the same in both files and you may need to change them before you merge.

How can I obtain NSFH1, NSFH2, or NSFH3 geographical information on respondents? 3/11/2021 Geomerge service for NSFH is temporarily suspended due to limited staff time at the Center for Demography of Health and Aging.

A mandatory confidentiality agreement must be signed and returned to NSFH before this information can be released to the user (GeoMergeOnline). The NSFH1 codebook can be found at (nsfh1gdt.doc), the NSFH2 codebook can be downloaded from: (nsfh2gdt.doc) and the NSFH3 codebook can be downloaded from: (nsfh3apl.doc). Once you have signed the form you can either mail it or fax it to the address listed on our home page. You can then write in to : and request the data. To view an example of the files you need to prepare for your geomerge request, please download this zip archive. To limit the possibility of disclosure, we reserve the right to recode data to missing if we believe the frequency counts for a geographical area are too small.

What is the best way to access variable frequencies?

The codebook files (c1*.* for wave 1 and c*.* for wave 2) contain the frequencies for the various categories or ranges of categories for each variable.

Given the person number for an individual in a household, how can I tell the relationship of this person to the respondent?

To find the relationship variable for the person indicated by a person number, look in the household roster and identify the relationship variable for the person with the same person number as given in the initial answer. To see how we constructed our household composition variables please see: (c1app_k.001).

Where are the survey instruments?

The survey instruments for NSFH1 and NSFH2 are on the web pages labeled "Codebook and Data Files" under a heading called "Interview Schedules". All interviews for NSFH3 were conducted over the telephone using CATI (computer-assisted telephone interviewing) technology. These files will not be publicly available. For more information on NSFH3 please read the field report located at: (fieldreport.doc).

I'm having problems with age variable discrepancies between NSFH 1 and NSFH2.

We have recognized errors in the respondent's age in NSFH1. The NSFH2 age data is more reliable.

How can I obtain skip maps?

Skipmaps are now available for NSFH1 and NSFH2 on the "Codebook and Data" web pages. We do not have them for wave 3.

I need citation information for NSFH1 and NSFH2.

The wave one citation is:
"The Design and Content of The National Survey of Families and Households." (NSFH Working Paper #1) James Sweet, Larry Bumpass, and Vaughn Call 1988. 

The NSFH2 citation is:
James A. Sweet and Larry L. Bumpass, The National Survey of Families and Households - Waves 1 and 2: Data Description and Documentation. Center for Demography and Ecology, University of Wisconsin-Madison (, 1996

The NSFH3 citation is:
James A. Sweet and Larry L. Bumpass, The National Survey of Families and Households - Waves 1, 2, and 3: Data Description and Documentation. Center for Demography and Ecology, University of Wisconsin-Madison
(, 2002.

Where are the interview date variables for NSFH2?

The interview date variables are in MR33, MR34 and MR35, columns 5255-5260.

Where can I find constructed variables on income measures for NSFH3?

No variables on income measures have been constructed for NSFH3. Only waves 1 and 2 have such variables.

Does NSFH have Marlowe-Crowne Social Desirability Scale?

No, NSFH does not have this scale.

How to calculate the century months if the event dates are after January 2000?

Conversion Formula for an event date after January 2000:

cm = ((year-1900) * 12) + month
For example:
June 2001 would be calculated like this
(2001-1900) * 12 + 6= 1218

For a table of century months from 1900 to 1999, see Appendix G in wave 1.

Are there variables available for use with SUDAAN, and if so, how can I get them?

Yes, these variables are available. The file, (sudaan.dat)contains Stratum, PSU and LA information for each of the NSFH respondents. sudaan.doc (sudaan.doc)contains format information for sudaan.dat. Please send e-mail to NSFHHELP ( if you have questions concerning these files.

I've read the instructions for retrieving files and obtaining data, but I still don't understand how to download them.

We now have a link, "Web-Accessible Data", which takes you to Nesstar. Through Nesstar, a data publishing program, you will be able to download data in the format of your choosing.

1. From our webpage: ( click on the link that says "Web-Accessible Data".

2. Click on the tab on the left side of the screen that says "National Survey of Families and Households".

3. Select the wave you are interested in and click on the tab directly to the left of that wave.

4. Once you have selected the wave of data you are interested in, click on the icon that looks like a computer disk located in the top right corner of the screen. From this page you will be asked to sign in as a user. Once you are recognized as a user you can then download data and/or a codebook, in the format of your choosing. It can take a couple days for your user name and password to be recognized.

Here's a quick-step method for downloading data files in text format via our web site.
1. Once you have read the introductory files, simple go to the "Codebook and Data Files" web page and save the data file you are interested in.

2. You can use either the codebooks or the "layout" files to find variable names and column locations to read in the data

How do I read the final weights in NSFH2?




Last modified on September 16, 2009