E R R A T A


This version of the National Health and Social Life Survey (1992) 
is not identical to a previously released version.  A few variables 
have been added, minor formatting changes have been made, the wording 
of the variable and value labels differ in style, and minor changes 
have been made to the data to reflect a small number of coding errors 
that have come to the researchers' attention.

All but two of the previously unreleased variables correspond to
items in section 11 ("Interviewer Comments") of the questionnaire.
These two variables are LEDATE (the date at which the respondent
last had sex) and INCCATZ (an alternate measure of family income,
discussed below).

A few of the items in the questionnaire are not available for public 
release for reasons of confidentiality.  Most of these variables are
included in some aggregated form.  For example, items that were originally
recorded using specific (month, day, and year) dates have been converted 
to "century months", where a century month is calculated as the number of 
months since January 1900.  City and state variables have been recoded
to the census region level.

In addition, this dataset contains variables which are not taken directly 
from the questionnaire.  They have been constructed from various sources, 
including the interview screener, sampling information, and rosters discussed 
in sections 2 and 4 of the questionnaire.

The NHSLS sample is comprised of two components:  a cross-section sample 
of 3159 cases (designed to give each household an equal probability of 
inclusion) and a supplementary sample of 273 cases drawn to increase the 
number of Blacks and Hispanics available for analysis.  The sample components
can be distinguished by using the variable:

     SAMPLE	Coded 1=Over-sample, 2=Cross-section

For household level measures, the cross-section sample is self-weighting.  For 
individual level measures, the following item can be used to construct a 
weight to adjust for individual level differences in selection probability:

     HHSIZE 	Number of eligible adults (ages 18-59) in the household.
		
This is done by weighting each case proportionately to the number of eligible 
adults in the household.  This is usually done in such a way that the 
constructed weights sum to the number of cases.

We also provide a weight (constructed by Martin Frankel, NORC's chief
sampling statistician) that allows for combining the oversample and 
cross-section cases while simultaneously adjusting for household size and
for nonresponse (post-stratification adjustment):

    RWEIGHT	SAMPLE WEIGHT FOR RESPONDENTS

This weight is scaled to sum to the total combined sample size of 3432.


For further discussion of the NHSLS sample and the measures contained 
in this dataset, please see:

  Laumann, E.O., Gagnon, J.H., Michael, R.T., and Michaels, S. (1994)
	THE SOCIAL ORGANIZATION OF SEXUALITY:  SEXUAL PRACTICES IN THE
	UNITED STATES.  Chicago: University of Chicago Press.

S E X U A L    P A R T N E R S

The following measures are based upon information provided by the respondent
in the "Spouse/Cohab" (section 2) and/or "Partner Last Year" (section 4) 
rosters, as well as the section on their most recent sex partner.

    PNAME1	This set of items corresponds to questions 2.2 and 4.2 
      to	in the questionnaire.  PNAME1-PNAME4 are reserved for 
   PNAME28	the respondent's two most recent sexual partners, the 
		respondent's first spouse/cohab partner, and most recent 
		spouse/cohab partner as applicable.  PNAME5 through PNAME28
		have been allocated to hold information on any other sexual
		partners.  The actual names of each partner have been replaced 
		by a three digit partner identification number.  These
		identification numbers will be used to match other measures
		(e.g. LNAME) to specific partners.

   SPTYPE1	This set of items indicate whether the corresponding
     to		partner is a spouse/cohab, a recent sexual partner
  SPTYPE28	(within the past year), or both.  Therefore, this item 
		can be used to determine whether or not section 2 
		("Marriage and Cohab") and/or section 4 ("Partner
		Identification and 1 Year Sexual Activity") contain
		additional information on this particular sexual partner.
		
   SPSCFR1	This set of items indicate whether or not the corresponding
     to		partner is a spouse/cohab and if so, if this partner is the
  SPSCFR28	respondent's first spouse/cohab, last spouse/cohab, or neither.
		
   SP2PLY1	This set of items indicate whether the corresponding partner
      to	was identified as the respondent's primary sexual partner
 SPS2PLY28	during the last year (see items 4.12-4.13), secondary
		partner (items 4.54-4.55), or neither.

     LNAME	This item (question #5.1) is a three digit code linking 
		partner idenfication number to the respondent's most recent 
		sexual event.  All items in section 5 ("Last Event") of the 
		questionnaire correspond to the sexual partner who has been
		assigned this three digit code.  

     MARST	Marital Status
     COHAB	Cohab Status

    SCOMIT	Indicates omission of Section 2 "Spouse/Cohab".
   PLYOMIT	Indicates omission of Section 4, "Partner Last Year". 
  OMITPRIM	Indicates omission of "Primary Partner Choice".
   OMITSEC	Indicates omission of "Secondary Partner Choice".
    LEOMIT	Indicates omission of "Last Sex Event".

C O L L A P S E D    M E A S U R E S 

Responses to the following items have been collapsed as indicated by
their assigned value labels.


     RES14	Question 1.7 -- Type of place respondent was living when R was
		14 years old.

  FAMILY14 	Question 1.8 -- With whom R was living at age 14.

    FAMWHY	Question 1.8a -- Why R was not living with own mother or father

    NUMBRO	Question 1.14 -- How many brothers did R have.
  & NUMSIS	        1.14a -- How many sisters did R have.

  LEFTHOME	Question 1.19 -- Age at which R first left home.

  FIRSTJOB	Question 1.22 -- Age at which R got his/her first job?

  WORKLAST	Question 1.25a -- Number of hours R worked for pay last.
  WORKTYPI	         1.26a -- Number of hours R would typically work.
 & WORKHRS	         1.30  -- Hours R works in a normal week.

    SPRES1 	Questions 2.10 and 2.12 combined for each of 28 possible
     thru	sexual cohabitation partners -- Did cohabitation period
   SPRES28	end?

   NUMPREG	Questions 3.1 and 3.2 combined -- Number of pregnancies 
		(women) R or (men) R's partners had by R in the past.

    BIRTHS	Question 3.3 -- Number of children ever had (live births).


  EDLEVEL1	Highest grade/year of schooling completed (DEGREE, collapsed).

  MILSTCAT	These measures are the collapsed (categorical) 
  MILEDCAT	alternatives to items #1.20A, 1.20B, and 1.21B as
  JAILACAT	indicated by their value labels.  They represent 
		the start and end dates of the respondent's 
		military experience (if any) and the age at which 
		the respondent was first confined, if ever.

   INCCATT	Information on household income (section SAQ No. 1)
   INCCATZ	has been collapsed into income level categories as 
  INCCAT1A 	indicated by their value labels.  INCCATT and INCCATZ
    to 		correspond to total household income, where INCCATT
  INCCAT6A	is non-missing only if _every_ subcategory of income
		(INCCAT1A-INCCAT6A) contains a valid response and 
		INCCATZ is non-missing if _any_ subcategory of income 
		is valid.

A D D I T I O N A L   V A R I A B L E S


    CASEID	Unique case identifier

    GENDER	This measure was recorded by the interviewer and
		was not asked of the respondent in this survey.

    FAMTOT	Number of family members in household. 
   FAMKIDS	Number of related kids (under 18) in the household.

      SIZE	These measures are based upon the respondent's address
    REGION	as recorded in item 1.6 of the questionnaire.  They 
   SRCBELT	were constructed using the procedure used in NORC's
  XNORCSIZ	General Social Survey (GSS).

       AGE 	Age in years was derived from the interview date and the 
		respondent's date of birth. 

  PRESTIGE	These measures are calculated from the respondent's
  PAPRES14	occupation and father's occupation responses.  They 
    		are based upon procedures used in NORC's General Social 
		Survey (GSS).

  CURRELIG	These measures represent the respondent's current religion
     REL14	(composite of RELIG and DENOM) and religion at age 14 
		(based on RELIG14 and DENOM14).  The Protestant category is 
		broken down into two subcategories, "Protestant Type I" and 
		"Protestant Type II", with Type I Protestestants constituting 
		more mainstream denominations and Type II Protestants
		representing more conservative/fundamentalists groups.  A
		discussion of this breakdown can be found in Appendix 3.1a
		of THE SOCIAL ORGANIZATION OF SEXUALITY (Laumann, et. al.,
		1994).  
		
    PART12	These measures are based upon the items in SAQ No. 2
       P18	and sections 2 and 4 of the questionnaire.  Construction
      P18S	of these measures is discussed in Appendix 5.1A of
    PRT18S	THE SOCIAL ORGANIZATION OF SEXUALITY (Laumann, et. al.,
		1994).  P18S (discrete) and PRT18S (categorical) are 
		constructed from question 1 of SAQ No. 2 and represent the
		number of sex partners since age 18.  P18 is an additional 
		discrete measure which was constructed from both the
		responses in SAQ No. 2 and section 2 of the face-to-face
		questionnaire ("Marriage and Cohab").

      POOR	These measures are based on the 1992 census defined poverty 
      RICH 	line, as calculated from both their income and household
  POVINDEX	composition information.

    ETHNIC	This is a measure which combines the respondent's
		race and ethnicity information into one measure.