Update and Revision History
National Household Food Acquisition and Purchase Survey (FoodAPS) public-use data files and codebooks were released in November 2016. This page contains information about all updates and revisions to the data, including changes to the data files and codebooks.
April 7, 2022
On April 7, 2022, the FoodAPS User Guide was republished, with updated processes for accessing restricted-use data.
April 28, 2021
On April 28, 2021, the Meals and Snacks data file and codebook were republished without the variables SNACKEVE1-7. The variables were removed following the discovery of a data processing error.
April 26, 2017
On April 26, 2017, the Household, Food-at-Home (FAH) Event, and Food-Away-From-Home (FAFH) Event codebooks were updated to include a new appendix section with definitions of SNAP store types provided by USDA's Food and Nutrition Service. These definitions are used to classify SNAP stores for the PRIMSTORESNAPTYPE and ALTSTORESNAPTYPE variables in the Household codebook, and the PLACESNAPTYPE variable in the FAH Event and FAFH Event codebooks.
February 21, 2017
On February 21, 2017, the Food-Away-From-Home (FAFH) Nutrient data file and codebook were republished to the ERS website to reflect the reclassification of 19 records in the FOODGROUP variable and the addition of a description for the variable FLUIDOZGRAMS in the codebook. The Household codebook was also republished with revised descriptions for the variables INCHHREPORTED, INCFAMREPORTED, and JOBCHANGE; however, note that the corresponding data file did not change.
January 18, 2017
On January 18, 2017, a revised FoodAPS individual-level public use file was republished to correct a small number of variables that were incorrect due to a data processing error. Twelve variables that identify an income source (with corresponding flags for imputed values), affecting 2 records each, were revised. The other corrected variables are RELATION_R, MARITAL, SPOUSEHHMEM, AGE_R, EDUCCAT, YEARSRESIDENCE, INITINTRVMON, INITFINALDAYS, FAMMEMBER, ANYINCOMEIND, and EMPLOYMENT, affecting 34 to 150 records each. Preliminary analyses indicate that this correction would probably have a minimal effect on analyses from the original file. The corresponding codebook was not affected.