3. Results and discussion
Although an effective quality assurance program during
data processing is time consuming, it is essential for
producing accurate and usable food and nutrient intake
information. In recent years, this effort has been lightened
somewhat by the introduction of automated collection and
coding procedures, which have eliminated many of the
types of errors once found during data processing. For
example, the USDA Automated Multiple Pass Method,
used in the integrated What We Eat in America since
January 2002 and by other studies involving FSRG,
minimizes problems such as collection of ambiguous
information or insufficient data by guiding the interview
through standardized questions about each food reported.
Food coding has also been enhanced. Once a potentially
error-prone operation, over 50% of foods collected by the
AMPM are now coded automatically by a computer
program. Still, quality assurance continues to be important
throughout data processing. For one thing, it provides the