USING THE CURF
ABOUT THE NNS CURF FILES
MERGING DATA FILES
The 1995 NNS CURF has been provided as three sets of data files:
- NNS Day 1 data files. These contain 13,858 confidentialised respondent records from the survey. Each record contains food and nutrition data from the survey, together with demographic, socioeconomic and geographic items. These files contain some information collected in the NHS which was considered integral to the NNS (e.g. breastfeeding, exercise and self-reported height and weight).
- NNS Day 2 data files. These contain 1,489 respondent records for those people who provided food intake data for a second 24-hour period (Day 2). The ABS has used these files to calculate factors which can be applied to nutrient intake data on the NNS Day 1 data files, with the aim of reducing within-person variation. The NNS Day 2 data files are intended for people wishing to undertake similar analyses, and can be merged with NNS Day 1 data files through a common identifier.
- NHS data file. This contains a full set of NHS information for the 13,858 NNS respondents. This can be merged with the NNS data file through a common identifier.
These data are released under the Census and Statistics Act 1905
, which enables release of data in the form of unit records where the information is not likely to enable the identification of a particular person or organisation. Accordingly, there are no names or addresses of survey respondents on the CURF and some other steps have been taken to protect confidentiality, such as omitting selected data items collected in the survey and reducing the detail of other data items. As a result, it may not be possible to exactly reconcile some tabular/aggregated data produced from the CURF with data published by the ABS.
A list of all the data items on the NNS Day 1, NNS Day 2 and NHS files is available on the Downloads
tab. This information is also available on electronic files provided as part of the NNS CURF. The NNS Day 2 data file has an identical structure to the NNS Day 1 data file, but the data item list for the Day 2 file contains only items which were collected in the Day 2 interview.
ABOUT THE NNS CURF FILES
The services of a computer programmer may be required to re-format the data for some analysis software. However, the NNS CURF contains confidentialised survey unit record data in SAS for Windows format, SPSS format and Stata for Windows format for each level (person and nutrition). Confidentialised survey unit record data is also provided in hierarchical comma separated file format (CSV). A SAS program for each file is provided and can be used to translate the CSV file formats to SAS formats. These SAS programs also create a formats library and attribute the formats to relevant variables in the associated files (for further information see below or the PDF provided: Important Information 1995 National Nutrition Survey CURF).
|NNS Day 1 data - person and nutrition levels|
|The following files specifically relate to the NNS Day 1 data:|
|NNS95PD1.CSV - contain the raw confidentialised survey data in hierarchical comma delimited ASCII text format - person level.|
|NNS95ND1.CSV - contain the raw confidentialised survey data in hierarchical comma delimited ASCII text format - nutrition level.|
|NNSBD1E2.SAS - contains a SAS program to run the SAS formats.|
|NNS95PD1.SAS7BDAT - contains the data in SAS format - person level.|
|NNS95ND1.SAS7BDAT - contains the data in SAS format - nutrition level.|
|NNS95PD1.SAV - contains the data in SPSS format - person level.|
|NNS95ND1.SAV - contains the data in SPSS format - nutrition level.|
|NNS95PD1.DTA - contains the data in Stata format - person level.|
|NNS95ND1.DTA - contains the data in Stata format - nutrition level.|
|FORMATS_DAY1.SAS7BCAT - SAS library containing formats for the Day 1 files.|
|NNS Day 2 data - person and nutrition levels|
|The following files specifically relate to the NNS Day 2 data:|
|NNS95PD2.CSV - contain the raw confidentialised survey data in hierarchical comma delimited ASCII text format - person level.|
|NNS95ND2.CSV - contain the raw confidentialised survey data in hierarchical comma delimited ASCII text format - nutrition level.|
|NNSBD2E2.SAS - contains a SAS program to run the SAS formats.|
|NNS95PD2.SAS7BDAT - contains the data in SAS format - person level.|
|NNS95ND2.SAS7BDAT - contains the data in SAS format - nutrition level.|
|NNS95PD2.SAV - contains the data in SPSS format - person level.|
|NNS95ND2.SAV - contains the data in SPSS format - nutrition level.|
|NNS95PD2.DTA - contains the data in Stata format - person level.|
|NNS95ND2.DTA - contains the data in Stata format - nutrition level.|
|FORMATS_DAY2.SAS7BCAT - SAS library containing formats for the Day 2 files.|
|NHS data - person level only|
|The following files specifically relate to the NHS data:|
|NHS95BE2.CSV - contain the raw confidentialised survey data in hierarchical comma delimited ASCII text format - person level.|
|NHS95BE2.SAS - contains a SAS program to run the SAS formats.|
|NHS95BE2.SAS7BDAT - contains the data in SAS format - person level.|
|NHS95BE2.SAV - contains the data in SPSS format - person level.|
|NHS95BE2.DTA - contains the data in Stata format - person level.|
|FORMATS_NHS.SAS7BCAT - SAS library containing formats for the NHS file.|
|The following additional files are also provided:|
|NUTADJ.TXT and LOGADJ.TXT contains age and sex specific factors that can be used to adjust nutrient intakes derived from the 24-hour recall questionnaire for within-person variation. These can be applied to the NNS Day 1 data files to derive adjusted nutrient intakes.|
|ANSURS technical support files which were used to code the food intake data from the 24-hour recall. The files provided are ACBDES.TXT, ACBGMWT.TXT, ARECHEAD.TXT, ARECING.TXT, AMODHEAD.TXT, AMODING.TXT, ARETFACT.TXT, ARETDES.TXT and |
|ANUTDES.TXT. These files are discussed in detail in Appendix 1.|
|COPRIGHT1.BAT describes the copyright obligations for people using ABS products and services.|
|IMPORTANT INFORMATION 1995 NNS CURF.pdf - provides links to all the CURF supporting documentation and a brief description of all of the files provided with the NNS CURF.|
Person and nutrition records for NNS Day 1 and Day 2 files have been provided in separate files, as outlined above. The NHS data file includes only person records.
The first record type is the person record, with each person having one of these. This record type includes socio-demographic information, summary information on food and nutrient intake the previous day, physical measurements, eating habits and attitudes, and results from the food frequency questionnaire (note: many of these fields are zero filled in the Day 2 person level files).
The second record type is the nutrition or food record, with each person having one or more nutrition records. Each nutrition/food record contains details of the individual food and beverages consumed by the person on the day prior to interview, including derived nutrient content.
Using the different record types
The ID field on the NNS data files identifies whether a record is a person level record or a food/nutrition record. In the survey documentation files, the ID field is called IDP on the person record file and is called IDN on the nutrition/food record file. The field is in the same position for both record types. The type of record can be identified by which file you are using (person or nutrition level) but also by using the following criteria:
- if IDP=4, then it is a person record;
- if IDN=5 then it is a food/nutrition record.
Each individual's person and nutrition records are linked through the field ABSHID (which identifies the household) and ABSPID (which identifies people within the household).
MERGING DATA FILES
The procedure to merge files depends upon what files are being merged.
Within NNS data files
Examples of merging within the NNS data files include merging the NNS Day 1 and Day 2 data files or, the person and nutrition records, that is merging food level data onto the person record.
All data on the NNS Day 1 and Day 2 data files have been sorted by the random identifier (ABSHID) and person number within the household (ABSPID). Merge the required data on ABSHID and ABSPID.
Between NNS and NHS data files
The main example of this is merging the NNS Day 1 and NHS data files. The NHS data file has been sorted on the random identifier (ABSHID), family number (ABSFID), income unit number (ABSIID) and person number within income unit (ABSPNID). Person number within income unit (ABSPNID) is different to the person number within household discussed above (ABSPID). This is because there can be multiple income units within a household.
As described in the section above, the NNS Day 1 and NNS Day 2 data files have been sorted on ABSHID and ABSPID. Therefore, these files need to be sorted in ascending order on the fields ABSHID, ABSFID, ABSIID and ABSPNID, and then merged with the NHS data file on these fields.
- All numeric data is right justified and contains leading zeros.
- Fields of types FILL only have the first four lines of documentation.
- Character data contain bother lower and upper case characters.
A CURF for the full NHS sample (53,828 records) can be separately purchased.
Note that the NNS CURF files contain approximately 260 person records which are not on the NHS CURF. This is because the NHS CURF only includes people from fully responding households (those households where all in-scope people completed the NHS). However, the NNS did not seek to enumerate all people within the household, so has retained the records of people who individually completed the NHS but were members of partially responding households (those households where one or more in-scope person(s) did not complete the NHS).