Evaluation of USDA Study on Wheat Protein Measurement
RCED-88-50, Nov 19, 1987
Pursuant to a congressional request, GAO analyzed the Department of Agriculture's (USDA) Federal Grain Inspection Service's (FGIS) study which resulted in a change in the way FGIS conducts wheat protein measurement.
GAO found that: (1) FGIS used appropriate statistical techniques and sampling methods in its study; (2) FGIS compared wheat protein content measurements using two methods, the Kjeldahl chemical process and the Near-Infrared Reflectance (NIR) method; (3) FGIS concluded that its NIR instruments had become biased in the direction of overstating the protein content of hard red spring wheat; (4) to correct the bias, FGIS adjusted its instruments so that they would measure the protein content of hard red spring wheat at 0.2 percent less than before the adjustment; and (5) FGIS initiated a study to recalibrate its instruments. GAO also found that FGIS plans to improve the accuracy of wheat protein measurement by: (1) standardizing grain coarseness by retrofitting grinders; (2) standardizing the time elapsed between grinding the wheat sample and placing it in the NIR instrument; and (3) recalibrating its NIR instruments on an annual basis. In addition, GAO found that, although FGIS has the authority to test wheat for protein content, no federal agency is responsible for routinely ensuring that wheat protein measurements by other entities, such as grain merchandisers, are accurate.