FFQ Error Report

Check error reports after FFQs are scanned. Note that all FFQs are processed regardless of missing or invalid values. A nutrient analysis record is generated using default values for any missing or incorrect fields. The following guidelines were developed to help users judge when error correction and rescanning may be useful.

(1) Review the Error Report (either the printout or output file xxx####.err)

The error report lists the locations and types of all errors. The report is sorted by FFQ booklet ID #. Note that the food item questions are numbered starting with FFRQ1 for "cold cereals" through FFRQ103 for "Other candy such as Lifesavers®, licorice, and jelly beans". Beverages are numbered starting at BFRQ1 for "Milk, all types as a beverage" through BFRQ17 for "Liquor and mixed drinks." See Survey File Variable Dictionary for a list of all variables.
The following types of errors are reported:

• Multiple marks when only one is allowed
• Missing values when a value is required
• Food Sections with less than half the foods marked

Note that incomplete erasures can trigger the scanner to show a multiple mark error. Covering erasures with correction tape prevents this error.

(2) Consider the Following Error(s) for Correction and Re-Processing

• Missing a summary question (FFQ page 10). NOTE: missing summary questions are serious and will affect estimates of fat, fruits and vegetables.
• Multiple marks in a summary question.
• More than 5 of the following types of errors on adjustment questions (FFQ pages 2 & 3):
  "Illegal multiple marks (exceeding____) for adj. question(s) ___."
  "Illegal missing value(s) for adjustment question(s) ___."
• Too many food items missing in a section:
  "Illegal blank fraction > .5 for food frequency group".

Evaluate these errors and how they may affect your research focus to determine if post-scanning edits are necessary. To do this, examine how the questions with errors default and whether these could have a meaningful impact on your results (see Technical Documentation: Adjustment Question Defaults). Note that a missing food item is assumed not to be eaten, and missing portion sizes are assumed to be medium. In general, these types of errors do not require correction. As noted above, missing summary questions are serious and will affect estimates of fat, fruits and vegetables.
**It is recommended that new users send smaller batches until they become familiar with the quality assurance measures necessary for successful processing.**

	Using FFQs