9.0 Additional KEYVARS

1. Age
   
2. Employment
   
3. Family Income
   
4. Family Size
   
5. Highest Grade
   
6. Marital Status
   
7. Number of Weeks Worked
   
8. Sampling Weight
   

9.1 Age

When you begin looking for a variable on 'age', you can begin in two places: 1. The Documentation, i.e., NLSY79 User's Guide, or 2. The Variables list. If you look in the NLSY79 User's Guide, look at the Table of Contents and, in particular, Chapter 4: Topical Guide to the NLSY79. If you want to look for it in the Variables list, begin with the Area of Interest index and find the group KEYVARS because all of the variables discussed in this section are grouped there. You could also look for it in the Any Word in Context index. In this example you are looking for variables about 'age' and, depending on where you started, this is what you would find:

• Area of Interest index. If you did not already know that the 'age' variable is found in the KEYVARS area of interest, you would immediately find nothing helpful about 'age' in this index. Remember the Areas of Interest are groupings of like variables and you would find no group designated 'age.' But as you become more familiar with the NLSY79 you will come to know where to look for a variable. In this section we are looking exclusively at KEYVARS variables that might be of interest to you. You should open the KEYVARS area of interest now (double-click on it in the Contents window).

• KEYVARS area of interest. When you open the KEYVARS list of variables, you might not immediately find the first 'age' variable, "R0216500   AGE OF R AT INTERVIEW DATE  79 INT   R0216500   [*Created] 1979 KEYVARS" in the list. If you sort the variables in the Variables window by description (click on Description), you get like-variables to group together. Sorting makes it easier to scroll down through the list and identify variables groups. After the sort, the first variable about 'age' is "R5081700   AGE OF R AT INTERVIEW DATE   R50817.00  [AGEATINT]  1994  KEYVARS".

  Notice the variable was created in the early years, it was an annual occurrence in 1979-1994, and then every two years subsequently. In 1993 creation of the variable stopped and it became a question named AGEATINT. (The 'survey year' and 'INT' were dropped in 1994 and following active years. There is no difference in the questions other than they sort in a different order due to the omission of 'yy INT'.) You can open the Codebook display of any of these variables by double-clicking on them.

• Any Word in Context index. You would find 1809 variables under the word AGE and the first variable in the Variable list would be "R0000600   AGE OF R  79 INT   R00006.00   [S01Q01B] ... etc. If that was the one you wanted:
  1. Notice it has been assigned to the FAMBKGN area of interest and that it is a result of a question (S01Q01B). (Double-click on it to see its Text/Cdbk display.)
  2. Its Text/Cdbk display adds the question text:
     "AND THAT MAKES YOU (R'S AGE ON HOUSEHOLD ENUMERATION). IS THAT CORRECT ?
     (IF NECESSARY CORRECT HOUSEHOLD ENUMERATION.)
     There is no documentation link to the codebook.

• Documentation. Look at the Table of Contents especially Chapter 4, Area of Interest and you would find the following and much more on 'age'.

  4.1 Age The following section discusses the age and date of birth variables associated with NLSY79 and NLSY79 Child respondents. In addition to the respondent-specific variables discussed in this section, information is also available on the age and/or date of birth for other household members; see the "Household Composition" section for details. NLSY79 NLSY79 Children

  NLSY79 User's Guide Chapter 4.1 Age

  In addition, the documentation differentiates among 'Age of R' and 'Age of R at Interview Date'; 'Age of R at 1st Birth' and 'Age of R at Start of 1st Pregnancy' from the Supplemental Fertility File (area of interest FERTILE); and it weighs the data, provides charts and the actual Reference Numbers to the variables, etc. From the documentation you have expanded your knowledge and scope of the variables available on 'age'. You may now be able to make a more refined choice on 'age'.

9.2 Employment

When you begin looking for a variable on 'employment', you can begin in two places: 1. the documentation, the NLSY79 User's Guide, or 2. the Variables list. If you look in the User's Guide, look at the Table of Contents and, in particular, Chapter 4: Topical Guide to the NLSY79. If you want to look for it in the Variables list, begin with the Area of Interest index or the Any Word in Context index. In this example you are looking for variables about 'employment' and, depending on where you started, this is what you would find:

• Area of Interest index. There is no grouping 'EMPLOYMENT' or anything like it in the Areas of interest. 'Employment' variables are part of the KEYVARS area of interest.

• KEYVARS area of interest. When you open the Variables window (double-click on KEYVARS) you may want to sort the variables by Description in order to get the 'EMPLOYMENT' variables to sort together. Scroll down to 'EMPLOYMENT...' and you will see two groups of 'EMPLOYMENT' variables. One group has the word (COLLAPSED) in its description but that is the only distinction between the group. The Question Names column tells you that most of the variables were created except for the years 1993, 1994, 1996, and 1998 when the variable was the result of a question asked.

  When you look at the Codebook display (double-click on the variable R0214901) for "R02149.01  [*Created]  EMPLOYMENT STATUS RECODE (COLLAPSED)  79 INT" you will notice the note: "NOTE: FOR VARIABLE CREATION, SEE APPENDIX 1: EMPLOYMENT STATUS RECODE" and the codebook for this variable provides a link to more documentation "Documentation Links: Codebook Supplement-Appendix 1".

  The significance of the word (COLLAPSED) in the variable's description means this variable was collapsed from a larger continuous scale. MARITAL STATUS, 'R44183.00', is a variable collapsed from 5-6 codes to 3. In some cases, the uncollapsed version of a variable, for example, unemployment rate, is in the Geocode data for confidentially. "The geocode data files contain the continuous unemployment rate as it was calculated for the area of residence. For confidentiality reasons, the main public use data files contain a collapsed version of the continuous unemployment rate." (NLSY79 User's Guide, NLSY79 APPENDIX 7: UNEMPLOYMENT RATE.)

• Documentation. If you follow the codebook hyperlink (left-click on the link) to the documentation, you will find the following information and, in addition, a list of the variables, and the logic used in creating them.

  NLSY79 APPENDIX 1: EMPLOYMENT STATUS RECODE VARIABLES 1979-1998 Employment Status Recode (ESR) is a standard measure of the respondent's main labor force activity during the survey week (Sunday through Saturday preceding the interview date). What follows is a PL/I adaptation of a Fortran program used by the Bureau of the Census to create ESR for 1979-1993, and for 1994 through 1998, an SPSSX adaptation of the decision rules provided by the Bureau of the Census to create MLR using the new CAPI-generated CPS data.

  NLSY79 User's Guide, Table of Contents, Chapter 4: Topical Guide to the NLSY79, 4.40 Workhistory Data

• Any Word in Context index. Open this index in the Contents window and type 'e', 'm', 'p', and you will see the index jump to the first occurrence of 'EMP'. When you look at the groups of variables whose group name begins with 'EMP' you will see a range of several words from 'EMP' through 'EMPLOYMENT'. Because you are looking for the "EMPLOYMENT' variables, you may be able to disregard 'EMPLD', 'EMPLOYED', 'EMPLOYEE(S)' and 'EMPLOYER(S) and look more closely at the 'EMP', 'EMPL', and 'EMPLOYMENT' groups. Open each one in the Variables window (double-click on the group name in the index) and look at them. Sort the variables by Description to more easily identify sub-groups and scroll down through them. You have to decide if any of these 'EMP' or 'EMPL' variables will be useful to you. If a variable looks useful to you, view it in the Codebook display (double-click on the variable).

  As you peruse the groups of variables (and open individual variables in the Codebook display) you may notice that 'EMP' is usually used as an abbreviation for 'employer' (but now you know where some of them are!) and 'EMPL' is also. Now you need to look at the 'EMPLOYMENT' variables. (Contents window, double-click on EMPLOYMENT.) In the Variables window, sort by Description and scroll down through the employment variables, looking for those that will fulfill your research needs. Remember to open them up in the Codebook display (double-click on the variable) for more detail.

9.3 Family Income

Suppose you are looking for a variable on 'family income', you can begin in two places: 1. the documentation, the NLSY79 User's Guide, or 2. the Variables list. If you look in the User's Guide, look at the Table of Contents and, in particular, Chapter 4: Topical Guide to the NLSY79. If you want to look for it in the Variables list, begin with the Area of Interest index or the Any Word in Context index. In this example you are looking for variables about 'family income' and, depending on where you started, this is what you would find:

• Area of Interest index. If you began in this index looking for 'family income' you would find no area of interest entitle FAMILY, but you would find a large number of variables (3000+) under INCOME to search through for 'family income'. (Double-click on INCOME to open the list in the Variables list window. Left-click on Description to sort the variables whose descriptions are alike.) When you quickly look through the index you find nothing under 'family'. Maybe it will not be fruitful to begin here, unless you know to look under "TOTAL NET FAMILY INCOME ...".

• KEYVARS area of interest. Because this section is introducing you to the KEYVARS area of interest, this is the best place to start looking for a 'family income' variable. But finding this variable will not be obvious. In the Contents window when you open the KEYVARS group in the Variables window (double-click on KEYVARS) you will need to sort the variables by Description (left-click on Description). When you scroll down to FAMILY ... you find variables whose names begin with "FAMILY POVERTY LEVEL", "FAMILY POVERTY STATUS", AND "FAMILY SIZE", but no 'family income'. If you open the Codebook display (double-click on the variable) of one of the variables in the "FAMILY POVERTY LEVEL" or "FAMILY POVERTY STATUS" groups, you will find a hyperlink, Documentation Links: Codebook Supplement-Appendix 2. If you link to this reference (NLSY79 Codebook Supplement, Main File 1979-1998, NLSY79 Appendix 2, you will find reference to "TOTAL NET FAMILY INCOME VARIABLE CREATION: 1979-1996" and the creation methodology.

  Look at the Documentation section below and read the quote. It also refers to a created variable entitled "TOTAL NET FAMILY INCOME? The best place to find a created variable is in the KEYVARS variables list. When you try to find it there, you will. (Sort the KEYVARS variable list by Description and scroll down to 'TOTAL ...'.) You will find, for example, "R0217900   TOTAL NET FAMILY INCOME IN PAST CALENDAR YEAR *KEY* 79 INT   R02179.00 *Created 1979 KEYVARS" and if you open it in the codebook (double-click on the variable), you will find a "NOTE: FOR VARIABLE CREATION, SEE APPENDIX 2: TOTAL NET FAMILY INCOME" and the hyperlink to the "Documentation Links: Codebook Supplement-Appendix 2".

• Any Word in Context index. If you began in this index looking for 'family income' you would find grouping of variables for both words, FAMILY and INCOME. If you look under FAMILY you would not find FAMILY INCOME among the Descriptions of in the list of variables. If you look under INCOME you would have to carefully scrutinize the Descriptions of the extensive list of variables to find FAMILY INCOME. This is a time consuming task. Better try a different approach.

• Documentation. If you begin with the NLSY79 User's Guide, Table of Contents, Chapter 4: Topical Guide to the NLSY79, you would find the topic 4.21. Income. When you open this topic (double-click on Income) the only information on 'family income' is a reference near the end, that points you to a variable's name.

  Created Values and Summary Statistics Each year, the CHRR staff creates an NLSY79 variable entitled "Total Net Family Income." This variable is designed to provide researchers with a summary variable of all the income received in the household. The actual computer code used to create this variable is found in Appendix 2 of the NLSY79 Codebook Supplement. Appendix 2 shows that each year a slightly different program is used to compute net family income.

  NLSY79 User's Guide, Chapter 4.21 Income

9.4 Family Size

Suppose you are looking for a variable on 'family size', you can begin in two places: 1. the documentation, the NLSY79 User's Guide, or 2. the Variables list. If you look in the User's Guide, look at the Table of Contents and, in particular, Chapter 4: Topical Guide to the NLSY79. If you want to look for it in the Variables list, begin with the Area of Interest index or the Any Word in Context index. In this example you are looking for variables about 'family Size' and, depending on where you started, this is what you would find:

• Area of Interest index. If you began in this index looking for 'family size' you would find no groupings for FAMILY of SIZE. FAMBKGN is the closest grouping and if you peruse this variable list it appears nothing pertaining to 'family size' is there. Check the KEYVARS area of interest.

• KEYVARS area of interest. "R0217502 FAMILY SIZE 79 INT R02175.02 *Created 1090 KEYVARS"
  Codebook - "R02175.02 [*Created] FAMILY SIZE 79 INT". No hyperlink to doc in codebook, just a reference in the note.

  NOTE: VARIABLE CREATION PROCEDURE IS COMPARABLE TO COMPUTATION OF
  "FAMILY SIZE" USED TO DETERMINE "FAMILY POVERY STATUS". SEE APPENDIX
  2: TOTAL NET FAMILY INCOME.
  

  and the reference to "SEE APPENDIX 2: TOTAL NET FAMILY INCOME". (Read: NLSY79 Codebook Supplement Main File 1979-1998, Appendix 2.)

• Any Word in Context index. If you began here, you would find a grouping of variables called FAMILY and if you open the group of variables (double-click on FAMILY) in the Variables window and sort them on Description (left-click on Description), you would find a group whose description begins with "FAMILY SIZE ...". If you opened "R0217502   FAMILY SIZE  79 INT  R02175.02 *Created 1979 KEYVARS" in the codebook display (double-click on the variable), you would find the

  NOTE: VARIABLE CREATION PROCEDURE IS COMPARABLE TO COMPUTATION OF
  "FAMILY SIZE" USED TO DETERMINE "
  FAMILY POVERY STATUS". SEE APPENDIX
  2: TOTAL NET FAMILY INCOME.

  and the reference to "SEE APPENDIX 2: TOTAL NET FAMILY INCOME". (Read: NLSY79 Codebook Supplement Main File 1979-1998, Appendix 2.)

• Documentation. If you began in this index looking for 'family size' you would find only Chapter 4.13 Family Background. If you linked to it, you would not find information on 'size' but some interesting things to note on 'family background'.

  However, if you followed one of the links from the codebook (NLSY79 Codebook Supplement Main File 1979-1998, Appendix 2) mentioned above, you would find two choices:

  • Variable Creation: Total Net Family Income
  • Family Poverty Status and Family Poverty Level Variables

  "Variable Creation: Total Net Family Income" contains only the 'raw' coding for the "Total Net Family Income" variable (no textual explanation). "Family Poverty Status and Family Poverty Level Variables" contains a textual explanation of the variables and Poverty Guideline tables.

If you followed the hyperlink to "Family Poverty Status and Family Poverty Level Variables" you would find the following quote and you would have to conclude that more information might be found in the documentation for the FAMILY POVERY STATUS variable.

1979 FAMILY POVERTY STATUS CREATION The 1979 NLSY household interview schedules did not use the Poverty Income Guidelines. As a result, when the family poverty status and the family poverty level variables were created in 1990, the actual federal Poverty Income Guidelines were used. Since the 1979 family poverty status variable was not created until 1990, a modified version of the family poverty status variable creation was used. First, if the created total net family income variable R(2179.) was greater than zero then income was compared to the Poverty Income Guideline amount for the respondent's family size, farm/non-farm distinction, and state of residence. Second, if the total net family income was less than zero, an abbreviated version of income was calculated using selected income variables (primarily earned income variables and other income variables that did not require computation of the number of months that the income was received). This abbreviated income amount was only used to determine if the family was not in poverty. Third, if the family poverty status was not computed based on either of the previous scenarios, then the screener income R(1916.10), if it was greater than zero, was compared to the Poverty Income Guideline amount for the respondent's screener family size, farm/non-farm distinction, and 1979 state of residence. A special flag variable (R2179.30) was created so that users could determine which income value was used to create the poverty status and the poverty level.

NLSY79 Codebook Supplement Main File 1979-1998, Appendix 2

9.5 Highest Grade

When you begin looking for a variable on 'highest level of education', you can begin in two places: 1. the documentation, the NLSY79 User's Guide, or 2. the Variables list. If you look in the User's Guide, look at the Table of Contents and, in particular, Chapter 4: Topical Guide to the NLSY79. If you want to look for it in the Variables list, begin with the Area of Interest index or the Any Word in Context index. In this example you are looking for variables about 'highest level of education' and, depending on where you started, this is what you would find:

• Area of Interest index. You would have to look in the KEYVARS area of interest to find the 'educational level' variables. When you open the KEYVARS group (double-click on KEYVARS) in the Variables window sort it by Description. Scroll down and peruse the sub-groups of variables and you will easily find "HIGHEST GRADE COMPLETED AS OF...". One group is (REVISED), the other is not. A 'REVISED' variable is typically created because it is an improvement on what came before. Given a choice to use a variable or its REVISED version, use the REVISED. The "HIGHEST GRADE COMPLETED..." variable was originally oriented for High School and not for 20-30 year olds in college as non-traditional students. You may want to look at the variables in the Codebook window (double-click on the variable). When you do you will find hyperlinks to more documentation.
  The Codebook display for "R02167.   [*Created] HIGHEST GRADE COMPLETED AS OF MAY 1 SURVEY YEAR  79 INT" provides the logic for the creation of the variables in the note field, and it provides a hyperlink, Documentation Links: Codebook Supplement-Appendix 8.

  The Codebook display for "R02167.01   [HGCREV79]  HIGHEST GRADE COMPLETED AS OF MAY 1 SURVEY YEAR  (REVISED)" provides the same hyperlink, Documentation Links: Codebook Supplement-Appendix 8.

• Any Word in Context index. Suppose you are looking for a variable dealing with the "highest level of education". If you began in the Any Word in Context index you would look for a grouping of variables entitled "HIGHEST" or "LEVEL" or "EDUCATION". In this example, you would find a group under each of the above terms. To be sure you would need to look at the variables in each group (double-click on the group 'word').

  If you began by opening the "HIGHEST" group and sorted it by Description, you would find "HIGHEST GRADE ATTENDED", "HIGHEST GRADE COMPLETED", and "HIGHEST GRADE EVER COMPLETED" and you would have to look at these variables in the code book for more detail and perhaps links to documentation in the NLSY79 User's Guide.

• Documentation. If you begin here looking for a variable on "educational level" or "highest grade", you should start with the "NLSY79 User's Guide, Table of Contents, Chapter 4: Topical Guide to the NLSY79". There you would find "4.12 Educational Attainment & School Enrollment" and if you hyperlinked to that topic, you would find the following information that introduces the scope of the 'educational' variables. There is a wealth of information in this section, be familiar with it if you plan on using educational variables. The third paragraph in the quote below not only contains the information on 'highest grade completed', but also contains additional information. References to supplementary material is included.

  NLSY79 Information on NLSY79 respondents' educational experiences has been collected during each survey year. In addition, three special data collections conducted during 1980-83 gathered the following for select universes: (1) supplementary information on degrees and certifications received as of the 1980 survey; (2) detailed information on the amount of time spent at school on each of the past seven days and the amount of time spent in various school-related activities, e.g., attending classes, studying, participating in other activities; and (3) high school course information gathered (directly from school records) during the 1980-83 transcript surveys. This section will review the primary types of enrollment and attainment data collected during the main NLSY79 surveys. More information on the special school/transcript surveys can be found in the "School & Transcript Surveys" section of this guide. Descriptions of the various standardized test scores available for NLSY79 respondents can be found in the "Aptitude, Achievement & Intelligence Scores" section. [Paragraph deleted} Two sets of variables have been created that summarize each respondent's school enrollment status and highest grade completed as of May 1 of each survey year. Codes for the names and locations of recent colleges attended, i.e., Federal Interagency Committee on Education (FICE) codes, are available for some years on the restricted-release geocode files. Finally, data on highest grade completed are available for household members (available all years), all siblings (1993), mother/father (1979), and current/most recent spouse (1979-82). Table 4.12.1 summarizes the major types of NLSY79 educational status and attainment variables and identifies the survey years during which such data were collected.

  NLSY79 User's Guide, Table of Contents, Chapter 4.12 Educational Attainment & School Enrollment

  If you followed the codebook links for the variables R02167. and R02167.01 discussed above, "Documentation Links: Codebook Supplement-Appendix 8", you would find the following additional information on the revised versions of the Higest Grade Completed (HGC) and Enrollment Status variables:

  Introduction The series of programming statements contained in this appendix were used in the standard computations for the created Highest Grade Completed (HGC) and Enrollment Status as of May 1st Survey Year variables from 1990-1998. In addition to the standard created variables, revised versions of the HGC and Enrollment Status variables for all survey years (1979-1998) have been added to the NLSY79 main data file. The general sources of error in the standard created variables were: grade "reversals", in which a respondent completed a lower grade in a later year, rather than staying in the same grade or advancing: The programs included below did not account for these cases, the largest source of which have been respondents enrolled in college; respondents with incomplete or ambiguous school information, preventing the computation in a given year of a HGC or Enrollment Status variable (a problem which can then continue through subsequent years, even if the respondent reported attending school in later survey years).

  NLSY79 User's Guide NLSY79 User's Guide, Table of Contents, NLSY79 Codebook Supplement Main File 1979-1998

9.6 Marital Status

Suppose you are looking for a variable on 'marital status', you can begin in two places: 1. the documentation, the NLSY79 User's Guide, or 2. the Variables list. If you look in the User's Guide, look at the Table of Contents and, in particular, Chapter 4: Topical Guide to the NLSY79. If you want to look for it in the Variables list, begin with the Area of Interest index or the Any Word in Context index. In this example you are looking for variables about 'marital status' and, depending on where you started, this is what you would find:

• Area of Interest index. If you looked for a 'marriage' variable in the index, you would find the group MARRIAGE. When you open it (double-click on MARRIAGE) and sorted it on Description, you would eventually find variables on MARITAL STATUS, for example, "R0010600  MARITAL STATUS R00106.00 [S02Q01] 79 INT" assigned to the MARRIAGE are of interest. If you opened it in the Codebook window (double-click on the variable), you will see the codebook information and the hyperlink to Documentation Links: User's Guide-Marital Status, Marital Transitions & Attitudes. Because we know some of the 'marriage' variables are in the KEYVARS area of interest we will also look there next.

• KEYVARS area of interest. When you open the KEYVARS variable list (double-click on KEYVARS) and sort the variables by Description, you will find two groups of 'marital status' variables. From 1979-1992 they variables in both groups were created, then became questions in 1993, 1994, 1996 and 1998.

  If you open the codebook (left-click on the variable), for example, "R0217501 MARITAL STATUS R02175.01 [*Created] 79 INT", you would find the possible answer categories but you would find no links to documentation from the codebook.

• Any Word in Context index.

• Documentation. If you followed the codebook link from variable "R0010600  MARITAL STATUS" to the documentation topic "User's Guide-Marital Status, Marital Transitions & Attitudes" (NLSY79 User's Guide, Chapter 4.27 Marital Status, Marital Transitions & Attitudes) you would find the information below on marital status. If you first began in the NLSY79 User's Guide, Table of Contents, you would find in "Chapter 4: Topical Guide to the NLSY79" a section, 4.27 Marital Status, Marital Transitions & Attitudes, on the variables of interest.

  This section of the User's Guide is important to understanding the 'marriage' variables and should be read thoroughly. It points out inconsistencies in the variables and refers to related topics and variables collected in other sections of the survey: for example, "General information regarding creation of the Supplemental Fertility File (area of interest FERTILE), including marital transition data, age at first marriage, and months between first marriage and first birth, is found in "Appendix 5: Supplemental Fertility Files" in the NLSY79 Codebook Supplement."

  Marital Status Marital status information for NLSY79 respondents is available from: (1) responses to questions fielded during the annual or biennial surveys; (2) two sets of created variables specifying marital status as of the interview date; and (3) an item on the marital status of each respondent as of the previous interview derived from the interviewing aid called the Information Sheet. The marital status of each respondent, i.e., whether he or she was married, widowed, divorced, separated, or never married, was collected during the 1979 survey and is available as a single variable, 'Marital Status.' The 1980-87 interviews collected change in marital status information. During the 1988 and subsequent surveys, 'Current Marital Status' interview checks are included in the questionnaire to verify separately the marital status of respondents who report a change in status since the date of last interview and respondents who do not. Two created variables provide data on the respondent's marital status, incorporating any changes, as of each interview date. The first set of yearly created 'Marital Status' variables is constructed with coding categories of "never married," "married," "separated," "divorced," and "widowed." Although two additional categories, "remarried" and "reunited," are present within the "Marital History" section of the questionnaire, those respondents who are remarried or reunited are simply coded as "married" in the created variable series. A collapsed version of this variable that codes the respondent's status as "never married," "married spouse present," or "other" is also available. Prior to 1988, marital status was created based on the last actual stated change in marital status. Marital status for 1988 and subsequent survey years has been created from the interview checks mentioned above. Since 1980, a marital status variable has been available from each interview's Information Sheet. This variable reflects the respondent's current marital status as of the date of the last interview. Coding categories are similar to those for the created marital status variables but differ slightly across years. In addition to these data, information is available on respondents' age at first marriage, the presence of opposite-sex partners, and the marital status of household members. The 'Age Began 1st Marriage' variable series has been created for 1982 through the present from the created 'Month/Year Began 1st Marriage' variables (see "Marital Transitions" below) and from the 1979 date of birth.

  NLSY79 User's Guide, 4.27 Marital Status, Marital Transitions & Attitudes

  There is a wealth of information on marital status, with definitions, and several user notes that should be looked at. For example, one note documents "spouse vs. partner" with the user caution "to carefully check the wording...of this type that are of interest".

  User Notes: Collection of information on partners as distinguished from spouses has varied over time in the NLSY79. To some extent, the term partner is used more interchangeably in sections of the questionnaire such as "Fertility" and "Childcare" than in sections such as "Marriage" and "Income and Assets," where items may specifically include or exclude partners. Users are cautioned not to assume the interchangeability of terms but to carefully check the wording of all the questionnaire items of this type that are of interest.

  NLSY79 User's Guide, 4.27 Marital Status, Marital Transitions & Attitudes, Spousal Characteristics

9.7 Number of Weeks Worked

Suppose you want to begin looking for a variable on 'weeks worked'. The NLSY refers to this variable as NUMBER OF WEEKS WORKED... In fact, always take a look at 'NUMBER OF ...' under the variable Description (Variables window) for variables that might contain phrases such as 'weeks worked', 'weeks unemployed', 'weeks in service', 'weeks since last', and 'weeks in or out', 'hours worked', etc. In this example you are looking for variables about 'weeks worked' and, depending on where you started, this is what you would find:

• Area of Interest index. If you started here you would find no area of interest for 'week', 'work', 'number of ...', 'employ...', etc. But you now know this variable was assigned to the KEYVARS area of interest.

• KEYVARS area of interest. When you open the Variables window (double-click on KEYVARS) you may want to sort the variables by Description in order to get the 'NUMBER OF ...' variables to group together. Sorted, they are easier to identify in groups. Scroll down to the 'NUMBER OF ...' area of the list and you will find several groups that begin with 'NUMBER OF WEEKS WORKED...'. You can double-click on any of the variables to view the codebook display to find out more information. In the codebook for R02157 you will find a note inserted:
  R02157. [*Created] NUMBER OF WEEKS WORKED IN PAST CALENDAR YEAR *KEY* 79 INT
  

  AUGUST 1983  DENNIS GREY
  NOTE: THIS KEY VARIABLE CONTAINS A VALID CODE FOR EACH RESPONDENT EVEN
  IF ALL TIME CANNOT BE ACCOUNTED FOR. IT SHOULD BE USED IN CONCERT WITH
  VARIABLE R(2157.01) THAT TELLS WHAT PERCENTAGE OF TIME IS MISSING FOR
  EACH RESPONDENT.
  

  The significance of the word *KEY* in a variable's description means it is a KEYVAR, a key variable, that was a deliverable by contact and typically created.

  In the Codebook note you find a correlation with another variable that complements the R02157 variable: "R02157.01  [*Created]  % OF WEEKS UNACCOUNTED FOR IN CALCULATING WEEKS WORKED IN PAST CALENDAR YEAR  79". The correlated variable is described in the note and its Reference Number is given. You have found another variable that can be used at your discretion.

• Any Word in Context index. If you started here, you would find the word NUMBER (1456 entries) in the index and when that group of variables was opened in the Variables window (double-click on the word NUMBER in the index), you would want to sort the variables by Description to get the 'NUMBER OF WEEKS WORKED' variables sorted together. You could eventually find that group of variables by scrolling down through a long list. You would then have to discriminate among several groups of 'NUMBER OF WEEKS WORKED' variables: ...BY SPOUSE..., ...IN PAST..., or ...SINCE LAST... Use the codebook display (double-click on the variable in the Variable list) to differentiate among the choices.

• Documentation.If you begin in the NLSY79 User's Guide, look at the Table of Contents and, in particular, Chapter 4: Topical Guide to the NLSY79. Look down the list of topics for something to do with 'work'. You will find "4.39 Work Experience" and "4.40 Workhistory Data". Read up on these topics so that you can distinguish among the 'work' variables that you need for your research. When you read the section in 4.39 about the "Cumulative Labor Force Experience", it will refer you by hyperlink to the Workhistory section if you are looking for week-by-week labor force status.

  Cumulative Labor Force Experience [Paragraph deleted] The detailed collection of dates of employment and gaps in employment over the history of the NLSY79 allows construction of a cumulative picture to be constructed of a respondent's labor force activity over the course of the survey. A large number of summary variables are created based upon the week-by-week labor force status arrays produced by the Workhistory program; see also the "Workhistory" and "Labor Force Status" sections of this guide for more information.

  NLSY79 User's Guide, Table of Contents, Chapter 4: Topical Guide to the NLSY79, 4.39 Work Experience

  Later in this section of the documentation, 4.39 Work Experience, you will find a list of variables on 'cumulative labor force experience' that can be helpful in determining the range of variables available to you on 'weeks worked'. You will have to following the same 'process' of investigation in order to find out about each of these variables.

  Related Created Variables (Cumulative Labor Force Experience) Number of Weeks Worked since Last Interview Number of Weeks Worked in Past Calendar Year Number of Hours Worked since Last Interview Number of Hours Worked in Past Calendar Year Number of Weeks out of Labor Force since Last Interview Number of Weeks out of Labor Force in Past Calendar Year Number of Weeks Unemployed since Last Interview Number of Weeks Unemployed in Past Calendar Year % Weeks Unaccounted for since Last Interview % Weeks Unaccounted for in Past Calendar Year Weeks since Last Interview Weeks in Active Military Service since Last Interview Weeks in Active Military Service in Past Calendar Year

  NLSY79 User's Guide, Table of Contents, Chapter 4: Topical Guide to the NLSY79, 4.39 Work Experience, Related Created Variables

  When you link to "4.40 Workhistory Data" (a.k.a. 'Workhistory') in the documentation you find this summary on 'labor force variables'. From this summary you can find out the scope of the various variables created on 'work' activity and the time period(s) referenced "...since last interview and the past calendar year" in these variables.

  Summary Labor Force Related Variables Variables are constructed summarizing different aspects of a respondent's labor force activity, including total number of hours worked, weeks worked, weeks unemployed, weeks out of the labor force, and weeks in active military service. There are two sets of these variables referencing each of two time periods—the period since the last interview and the past calendar year (see the "Labor Force Status" section of this guide). Variables are also created indicating the number of weeks since the previous interview and the percent of weeks which cannot be determined in constructing the summary variables discussed above. See the "General Work Experience" part of the "Work Experience" section for further notes on these variables.

  NLSY79 User's Guide, Table of Contents, Chapter 4: Topical Guide to the NLSY79, 4.40 Workhistory Data

9.7 Sampling Weight

Sampling weights are constructed in each survey year. These weights provide the researcher with an estimate of how many individuals in the U.S. each respondent's answers represent. The interested user should consult the NLSY79 Technical Sampling Report (Frankel, Williams, and Spencer 1983) for a step-by-step description of the adjustment process. A cursory review of the process is given in the documentation. The most important thing about the sampling weight is to know the "implied decimal point" of the datum, and you get this from the codebook.

• Area of Interest index. You would find no grouping resembling "SAMPLING" or "WEIGHT". It is in the KEYVARS area of interest.

• KEYVARS area of interest. Open KEYVARS in the Variables window by double-clicking on it in the index. Sort the variables by Description and scroll down to 'SAMPLING...'. Open "R0216100 SAMPLING WEIGHT 79 INT" in the codebook by double-clicking on it. See the sample codebook display below in the following section.

• Any Word in Context index. If you begin here, look for 'SAMPLING' in the index and open it (double-click on it) in the Variables window. You will see all the sampling variables. Double-click on one to open it in the Codebook window.

  R02161. [*Created] SAMPLING WEIGHT 79 INT Survey Year: 1979 SEE R(1376.) 1 THRU 9999999 SAMPLING WEIGHT (2 IMPLIED DECIMAL PLACES) Refusal(-1) 0 Don't Know(-2) 0 TOTAL =========> 12686 Valid Skip(-4) 0 Non-Interview(-5) 0 Lead In: R02160.20[Default] Default Next Question: R02162.00

  Notice the phrase: "SAMPLING WEIGHT (2 IMPLIED DECIMAL PLACES)". This instructs the user to shift the decimal point two places to the left when using one of the statistical packages.

• Documentation. If you began here, NLSY79 User's Guide, Table of Contents, you would find "Chapter 2: Sample Design & Fielding Procedures. If you followed that link you would find:

  2.8 Sample Weights and 2.9 Practical Usage of Weights In each survey year a set of sampling weights are constructed. These weights provide the researcher with an estimate of how many individuals in the U.S. each respondent's answers represent. Weighting decisions for the NLSY79 are guided by the following principles: a) Individual case weights are assigned for each year in such a way as to produce group population estimates when used in tabulations, and b) The assignment of individual respondent weights involves at least three types of adjustment, with additional considerations necessary for weighting of NLSY79 Child data. The interested user should consult the NLSY79 Technical Sampling Report (Frankel, Williams, and Spencer 1983) for a step-by-step description of the adjustment process. A cursory review of the process follows.

  NLSY79 User's Guide Chapters 2.8 & 2.9

Next section, 10.0 The FERTILE Variables.