Third National Health and Nutrition
Examination Survey (NHANES III) Public-Use Data Files Errata
The following items indicate discrepancies found after the release of the Third National
Health and Nutrition Examination Survey (NHANES III)
CD-ROM in 1997. This affects both the NHANES III SETS and ASCII CD-ROMs.
| File Location: |
Exam |
| Section: |
BM - Body Measurements |
| Item Name: |
BMPRECUM |
| Description: |
Additional Note to be added to explain discrepancy of some measurements in
BMPRECUM due to technician error. |
| BMPRECUM: |
This additonal note documents incorrect or highly unusual data values that were
determined for this variable. Analysts may wish to correct or be aware of these
measurements. The sequence number and current value in the data file are given below. To
correct these, 23.1 should be subtracted from each of the values shown below for BMPRECUM. |
| Sequence No. |
BMPRECUM |
| 17350 |
83.6 |
| 1936 |
81.1 |
| 18423 |
88.1 |
| 6939 |
94.6 |
| *7778 |
87.6 |
| 1133 |
90.6 |
| 15844 |
102.3 |
| 1768 |
106.1 |
| 13630 |
108.7 |
| 13510 |
105.6 |
| 3327 |
104.1 |
| 6672 |
106.1 |
| 8517 |
115.3 |
| 20377 |
117.1 |
| 13749 |
115.6 |
| 18556 |
119.5 |
| *This record was found to be correct on July 24,
1997, after the CD-ROM had already been released. Do NOT subtract 23.1 from this record. |
| File Location: |
Exam |
| Section: |
MA - Mec Adult Questionnaire |
| Item Name: |
MAPF32S |
| Description: |
SAS label for this variable incorrectly states
the question as "How many days ago stop taking BC pills." Correct wording for
SAS label is "How many months ago stop taking BC pills." Code book is correct.
Data is correct. Index reflects incorrect SAS label. |
File
Location |
Exam |
| Section: |
Body Measurements |
| Item Name: |
BMPBMI |
| Description: |
Additional note to be added to explain
discrepancy between BMI calculated from variables BMPWT and BMPHT (BMPBMI=BMPWT/BMPHT**2)
vs. use of BMI from the pre-calculated BMPBMI variable on the file. |
| BMPBMI |
This note documents incorrect data values for
the BMPBMI variable that do not satisfy the equation BMI=weight/height**2. Analysts may
wish to correct or be aware of these incorrect values. Another alternative is to simply
disregard the BMPBMI variable and calculate the actual BMI (BMI=BMPWT/BMPHT**2) for BMI
analysis. Discrepancies arose as the result of final edits made to the BMPWT and/or BMPHT
variables for selected subjects, and failing to then recalculate BMPBMI for these subjects
prior to release of the public-use data file. |
| Sequence No. |
BMPBMI |
Actual
BMI |
| 8769 |
12.2 |
14.5 |
| 9967 |
17.2 |
12.3 |
| 13106 |
24.7 |
18.4 |
| 19083 |
12.2 |
14.3 |
| 21348 |
41.5 |
33.4 |
| 37200 |
42.4 |
29.8 |
| 37451 |
20.3 |
27.0 |
| 44949 |
15.8 |
19.9 |
| 48587 |
12.5 |
16.5 |
| 49438 |
16.2 |
16.3 |
| 53487 |
16.5 |
24.7 |
| 53510 |
15.5 |
23.5 |
| File Location: |
Exam |
| Section: |
MA - MEC
Adult Questionnaire |
| Item Name: |
MAPF12R |
Description:
|
Note
lists incorrect ages for administration of question. Corrected note text follows. |
| MAPF12R: |
Pregnancy
status recode This variable
combines responses from self-reported pregnancy status (F12) with the results of urine
pregnancy test conducted in the mobile examination center on women 17-49 years of age. If
either the self-report or the test result was positive, then MAPF12R was coded yes.
Otherwise, if either the self-report or test result was negative, then MAPF12R was coded
no. If the self-report was don't know and the test result was either inconclusive or blank
but applicable, then MAPF12R was coded don't know. If the self-report was blank but
applicable and the test result was inconclusive, then MAPF12R was also coded don't know.
Finally, if both responses were blank but applicable, then MAPF12R was coded blank but
applicable. |
If, when running a frequency on
your variables you notice fill values not specified in the documentation, for instance a
9998, the SAS engine you are running may be different from that used to generate the SAS
input statements. The problem may be corrected by increasing the SAS length by 1 for that
variable.
For instance hyaitmo may appear to have
a fill value of 9998.
Use the SAS length statement to
increase the length from 3 to 4 in the SAS dataset:
length hyaitmo 4;
the data on the
ASCII ROM is accurate.
The following variables may be
affected:
| Variable |
File
Location |
| HYAITMO |
Exam, Lab, Youth |
| HXPH21S |
Exam |
| MAPF51 |
Exam |
| MAPF54R |
Exam |
| MAPF55 |
Adult |
| HAB6S |
Adult |
| HAD9S |
Adult |
| HAS12S |
Adult |
| HAT1S |
Adult |
| HAT3S |
Adult |
| HAT5S |
Adult |
| HAT7S |
Adult |
| HAT9S |
Adult |
| HAT11S |
Adult |
| HAT13S |
Adult |
| HAT15S |
Adult |
| HAT17S |
Adult |
| HAT20S |
Adult |
| HAT22S |
Adult |
| HAT24S |
Adult |
| HAT26S |
Adult |
| HAV1S |
Adult |
| HAV4S |
Adult |
| HAV6S |
Adult |
| HAV7R |
Adult |
Contacts
For publications and electronic data products from NHANES: call (301) 458-4636 or NTIS or
GPO.
Technical questions about
the file contact:
Division of Health Examination Statistics
Centers for Disease Control and Prevention
National Center for Health Statistics
3311 Toledo Road
Hyattsville, Maryland 20782-2003
(301) 458-4567 |
This page last reviewed
July 24, 2008
|
|
|