Errata for NLSY97 Round 5 Release

Errata for NLSY97 Round 5 Release

R5 Errata affecting the August 2003 Release: Round 5 Schooling Created Variables

(CV_SCH_ATTEND_YR). A programming error in the code that creates this variable inadvertently counted the number of schools the respondent attended as of June 2000 rather than the correct date of June 2001. Updating this variable to reflect the correct date affected the number of schools attended by 140 respondents. Most (130 respondents) attended 1 additional school, while five respondents attended two additional schools. The data in the R5M_NumSch.dat file fixes this problem. Researchers should use the CV_SCH_ATTEND_YR variable in this file rather than the variable presently contained on the August 2003 version of the NLSY97 Event History data.

The R number that corresponds to this variable is the following: R7237600 [CV_SCH_ATTEND_YR].


R5 Errata affecting the August 2003 Release: Schooling Event History

Two mistakes in the round 5 NLSY97 Schooling Event History Arrays have been fixed. The round five public release has a large number of illegal skips (-3) in the enrollment arrays. These illegal skips values were set to unknown by the program that creates the schooling histories but were accidentally converted to illegal skip upon public release. Researchers using the fixed data will see, instead of illegal skips, respondents flagged as not enrolled.

The second error is that some of the semesters respondents spent in college were not included in the event histories. The semesters that were not included were those whose start dates were not explicitly stated. The NLSY97 questionnaire tries to elicit the starting date of every college semester, but it is possible for respondents to indicate that they are still in college and by-pass the start date questions. Semester start and stop dates are now estimated for respondents who by-pass this portion of the questionnaire.

Fixed data monthly schooling data will be integrated with the NLSY97 round 6 event history release.


R5 Errata affecting the August 2003 Release: Round 5 Employment Event History (EMP_HOURS_2000.10 to EMP_HOURS_2001.22)

An event history file that updates the hours worked per week did not load into the data set properly which caused data for select respondents to be missing from the tenth week of 2000 (EMP_HOURS_2000.10) to the twenty-second week of 2001 (EMP_HOURS_2001.22). Previously, these data were available only for respondents whose round 4 interview date occurred after the 22nd week of 2001; from the 23rd week of 2001, the data were available for all respondents, regardless of interview week. The files in the R5EH_hrs_update file fix this problem. Researchers should use the EMP_HOURS_2000.10 to the EMP_HOURS_2001.22 variables in this file rather than the variables presently in the August 2003 version of the NLSY97 Event History data.

The Rnumbers that correspond to these variables are the following:

R8814700 [EMP_HOURS_2000.10] to R8819000 [EMP_HOURS_2000.53] and R8909400 [EMP_HOURS_2001.01] to R8911500 [EMP_HOURS_2001.22]


R5 Errata for May 2003 Release: Created Age at Interview Date Variable (CV_AGE_INT_DATE)

In the NLSY97 data released May 2003, the codebook page for the variable CV_AGE_INT_DATE incorrectly lists the frequencies for this variable. The data for this variable are correct and the documentation has been updated on the Round 5 Event History Release.