C. Statistical Analysis
The sample test results for the grade factors, moisture, chemical composition, and physical factors were summarized as the U.S. Aggregate and also by three composite groups that supply corn to each of three major export channels, labeled Export Catchment Areas (ECAs), as follows:
- The Gulf ECA consists of areas that typically export corn through the U.S. Gulf ports;
- The Pacific Northwest (PNW) ECA includes areas that export corn through Pacific Northwest and California ports; and
- The Southern Rail ECA comprises areas generally exporting corn to Mexico.
In analyzing the sample test results, the Council followed the standard statistical techniques employed for proportionate stratified sampling, including weighted averages and standard deviations. In addition to the weighted averages and standard deviations for the U.S. Aggregate, weighted averages and standard deviations were estimated for the composite ECAs. The geographic areas from which exports flow to each of these ECAs overlap due to available transportation modes. Therefore, composite statistics for each ECA were calculated based on estimated proportions of grain flowing to each ECA. As a result, corn samples could be reported in more than one ECA. These estimations were based on industry input, export data, and evaluation of studies of grain flow in the United States.
In some instances, the elevator recruiting process resulted in surplus samples in the ASDs. These extra samples were tested to provide greater density in the sampling. However, the U.S. Aggregate and ECA averages were still weighted by the original sampling proportions. The Relative ME was calculated for each of the quality factors for the U.S. Aggregate and each of the ECAs.
The Relative ME was less than ±10% for all the quality attributes except for SCI for the U.S. Aggregate, and stress cracks and SCI for the Gulf, Pacific Northwest and Southern Rail ECAs. The Relative ME for stress cracks and SCI was as follows: While the lower level of precision for these quality factors is less than desired, these levels of Relative ME do not invalidate the estimates. A footnote in the summary table for “Physical Factors” indicates the attributes for which the Relative ME exceeds ±10%.
References in the “Quality Test Results” section to statistical differences between 2012 and 2011 test results were validated by two-tailed t-tests at the 95% confidence level. The t-tests were calculated between results in the Harvest Report 2011/12 and the Harvest Report 2012/13.