Appendix 7: Continuous Month Scheme and Crosswalk

Appendix 7: Continuous Month Scheme and Crosswalk


As described in Appendix 6, the various event history arrays present month-by-month or week-by-week information on the respondent's status with respect to various activities. Information about marriage/cohabitation status, program participation, schooling, and arrests/incarceration is provided on a monthly basis using a continuous month timeline. Employment histories are presented on a weekly basis using a continuous week timeline. This appendix explains the continuous month/week system and provides the actual dates corresponding to the numbered weeks and months.

Continuous Month Timeline
The non-employment event history arrays are presented using two different timelines. The first, a continuous month approach, labels January 1980 as month 1, February 1980 as month 2, and so on. Thus, a respondent born in month 4 might start receiving public assistance in month 193 and leave the program in month 198. Key events occurring during the life of each respondent can be indexed within this month-by-month structure. To aid users, a number of variables other than the event history arrays have already been created using the event history format. For example, the respondent's age in months at interview date is presented in continuous month format. The event history arrays also can be constructed using actual month and year dates.

The data set includes variables for the event history arrays for every month number under the continuous month scheme. However, actual month and year dates are only included for a given respondent in months in which transitions occurred. That is, the respondent in the example above would have continuous month variables relating to program participation status for every month beginning with 172 (the month of the respondent's 14th birthday) but would only have actual date variables for January and June 1997. To aid users in moving between the two dating schemes, this appendix contains a crosswalk between actual dates and continuous month numbers (as well as continuous week numbers).

The continuous month system allows researchers to easily compare the times at which various events occurred in the respondent's life. The following example illustrates a case in which a respondent turns 17 in February 1997 and, in the same month (month 194 in the continuous timeline), begins receiving Aid to Families with Dependent Children (AFDC). She remains on AFDC for 2 months (month 194 and month 195). The following table illustrates this information using a month-by-month timeline.

Non-Employment Event History Array Incorporating the Month-by-Month Timeline

Question Name Participation Status Month Number
(created by the user)
Birth date:
AFDC_STATUS_1997.01 0 193  
AFDC_STATUS_1997.02 1 194 194
AFDC_STATUS_1997.03 1 195  
AFDC_STATUS_1997.04 0 196  

In addition to linking the dates through the crosswalk, any user who selects an event array will be able to extract additional characteristics that relate to the event. The following table presents the additional information that a user may chose to link to the non-employment event history arrays. A user is able to create these and other variable combinations for any status array.

Age (birth year) Variable 1 Variable 2 Variable 3
Enrollment Highest grade attended Highest grade completed Highest degree earned
Cohabitation/Marital Partner's race Partner's religion Partner's grade level
AFDC Received during month Benefit amount People benefiting

Continuous Week Timeline
In the event history section of the data set, an employment history of the number of weeks worked is presented in a continuous week-by-week status array. This array is very similar to the month-by-month scheme used for the rest of the event history variables. In this format, the first week of January 1980 is numbered week 1, the second week of January 1980 is numbered week 2, and so on. Weeks are listed by exact date, as well. This week-by-week array covers a respondent's employment status for each week since the month of the respondent's 14th birthday.

As with the month-by-month arrays, respondents have data available for every week in the continuous week arrays, but only for transition weeks in the actual date arrays. However, users should note that, due to technical considerations, the variable titles number weeks by calendar year rather than from week 1 through week 1253. For example, a variable might read "1994 Employment: Employment Status in Week 07" instead of "Employment: Employment Status in Week 737."  This appendix also provides a crosswalk between the actual beginning date of each week, the continuous week number, and the week number for each week in a calendar year (as well as continuous month numbers).

Naming Conventions
The question names of the NLSY97 event history variables incorporate the dates to which the variables apply. The marriage/cohabitation status, program participation and arrests/incarceration array titles include the month and year. (For example, AFDC_AMT_1997.12 corresponds to December 1997.) Likewise, employment array variables are listed for each week and year. (For example, EMP_STATUS_1997.52 corresponds to the 52nd week in 1997.) The schooling variables are yearly and employ a slightly different system. In general, these variables refer to the school year rather than the calendar year. That is, 1991 in a variable title or in the data for a variable generally indicates the school year starting in fall 1991 and ending in spring 1992.

Continuous Month and Week Crosswalk

The below crosswalk document provides actual dates corresponding to numbered weeks and months. The document is provided as an Excel spreadsheet, but if you need to access this information in another form, please contact NLS User Services.

Download the Crosswalk Spreadsheet

Notes on the Excel file:

  • All weeks start on Sunday and end on Saturday. The only exception is the first week in the file, which is listed as starting on January 1, 1980 (the earliest possible birthdate of any NLSY97 respondent, and thus the starting point for the event history arrays). January 1, 1980 was a Tuesday.
  • The file contains week and month numbers starting in January 1980 (the birth month of the oldest respondents in the cohort) and currently extends through 2022.