National Survey of Family Growth, Cycle III, 1982 (ICPSR 8328)

1. The original hierarchical data file consisted of 21,737 records. The original hierarchical file has, however, been replaced by two rectangular files of which 7,969 are respondent records (one record for each sample woman) and 13,768 are interval records (one record for each closed pregnancy (completed pregnancy) interval reported by sample women).

2. Data for CYCLE I, 1973 (ICPSR 7898), CYCLE II, 1976 (ICPSR 7902, 8181), CYCLE IV, 1988 (ICPSR 9473), CYCLE IV, 1990 TELEPHONE REINTERVIEW (ICPSR 6643), CYCLE V, 1995 (ICPSR 6960), and CYCLE VI, 2002 (ICPSR 4157) are also available from ICPSR.

3. For certain variables, the variable names that appear in the data differ slightly from the variable names given in the original codebook. For example, variables A_1 and A_5B_1 (Respondent File) in the original codebook appear as A1 and A5B1, respectively, in the data.

4. Codes for variable E26 are undocumented in the original codebook. However, the codes for values of 1 (Yes) and 2 (No) are defined in the original questionnaire. Since yes/no questions follow the same coding scheme, ICPSR has added labels for codes 7 (Refused), 8 (Don't know), and 9 (Not ascertained) based on the assumption that E26 also follows the standard yes/no coding scheme.

5. Please see the employment codes listed in the 1980 U.S. Census section of the original codebook) for the value labels for the following variables found in the Respondent File: F7, F9, F32_33, F63_64.

6. For variables MC_2A through MC_9A, there are not value labels for codes greater than 500. Codes between 616 and 720 represent cases for which two birth control methods were used during the same time period. Codes between 816 and 920 represent cases for which two birth control methods were used in sequence during a particular month.

7. For variable F22_1CM there exists a code which appears to be out of the normal range: '9999999'. There are 6 cases for which a code of '9999999' is given. Code '9999' is a valid code for this variable, representing "data not ascertained". The original codebook has 31 cases for which '9999' is given as a response. In the data, the code '9999' is given as a response only 25 times. It is possible that the codes '9999999' were incorrectly coded and were meant to be '9999' (the sum of the two would provide the 31 cases that responded '9999' according to the codebook).

8. The following variables have an implied decimal point: AGEDISS, AGEREMAR, AGEDD1, AGECON1, AGEBABY1, FINALWT, FMAR1AGE, MAR2P, and NONRWT (Respondent File) as well as AGEPREG (Interval File). Please consult the codebook for additional information regarding the placement of the decimal point for these data.

9. Inapplicable responses were originally coded as blanks. However, due to certain limitations in SAS, SPSS, and Stata these have been converted to system missing. As such, the inapplicable statements, originally found in the value labels have been moved to the question text that accompanies the ICPSR Variable Descriptions and can be found in the codebook.

10. Apparent inconsistencies among related variables result from the use of different items on the questionnaire to define certain concepts. See the section of the original study documentation entitled "Data Quality" for additional information.

11. Data variables are often given as century months. For additional information on how century months are calculated please refer to the "Date Codes" section of the codebook.

12. In the processing of this study, the original column locations (from the NCHS version) were changed. Throughout the original ASCII data file there were blank columns that served as "Fillers". These blank columns were not maintained resulting in different column locations for the majority of variables. Please disregard the column locations in the original codebook as they no longer apply to the ASCII data file being released with this study;

13. In the latest update to ICPSR 8328, version 3, the widths of the following variables were augmented in order to accommodate the missing data codes; none of the data was affected as there were no cases with missing data: B12_2, B16_2, B16_3, B17_2, B18_2, B18_3, PCT_A, B54_5, B55_5, B54_6, B55_6, B54_7, B55_7, MC_1A, D6_2_3, D8_2_2, D8_2_3, D12_R, D12_H, CH_5, and POVERTY.

14. The variable F9 has been changed from a 2-digit occupation code to a 3-digit occupation code in this dataset.

15. In the update to ICPSR 8328, version 4, the variable POVERTY was corrected and replaced.

personal interviews

• United States Department of Health and Human Services. National Center for Health Statistics. National Survey of Family Growth, Cycle III, 1982. ICPSR08328-v4. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2010-04-13. http://doi.org/10.3886/ICPSR08328.v4

2010-04-13 The variable POVERTY in the Respondent File has been corrected and replaced. Updated ready-to-go, ASCII, and tab-delimited files are being provided. The PDF codebook has also been updated with a new frequency distribution for POVERTY.

2009-05-27 The widths of the following variables were augmented in order to accommodate the missing data codes; none of the data was affected as there were no cases with missing data: B12_2, B16_2, B16_3, B17_2, B18_2, B18_3, PCT_A, B54_5, B55_5, B54_6, B55_6, B54_7, B55_7, MC_1A, D6_2_3, D8_2_2, D8_2_3, D12_R, D12_H, CH_5, and POVERTY. The variable F9 has been changed from a 2-digit occupation code to a 3-digit occupation code in this dataset.

2008-11-12 This study has been updated to include SAS, SPSS, and Stata setup files as well as SAS transport, SPSS system, and Stata system files. An English version of the questionnaire has also been added to the study documentation.

2000-05-17 The codebook and data collection instruments for this study have been converted to Portable Document Format (PDF).

1985-05-24 ICPSR data undergo a confidentiality review and are altered when necessary to limit the risk of disclosure. ICPSR also routinely creates ready-to-go data files along with setups in the major statistical software formats as well as standard codebooks to accompany the data. In addition to these procedures, ICPSR performed the following processing steps for this data collection:

• Performed consistency checks.
• Standardized missing values.