IX. HARVEST SURVEY AND STATISTICAL ANALYSIS METHODS
The key points for the survey design and sampling and statistical analysis for this 2015/2016 Harvest Survey are as follows:
- The harvest samples were proportionately stratified according to Agricultural Statistical District (ASD) across nine key sorghum-producing states, which represented more than 98% of U.S. sorghum exports. Additionally, the samples were classified according to two Harvest Areas – Early Harvest and Late Harvest.
- A total of 200 harvest samples collected from the nine states were targeted to achieve a maximum ± 10% relative margin of error (Relative ME) at the 95% confidence level for the grade factors.
- There were a total of 207 unblended sorghum harvest samples tested. These samples, received from local elevators, were pulled from inbound farm-originated trucks from August 28, 2015 through January 5, 2016.
- A proportionate stratified sampling technique was used for the mycotoxin testing across the ASDs in the nine states surveyed for the other quality factors. This sampling resulted in 58 harvest samples being tested for aflatoxins and DON.
- Weighted averages and standard deviations following standard statistical techniques for proportionate stratified sampling were calculated for the U.S. Harvest Aggregate and each of the two Harvest Areas.
- To evaluate the statistical validity of the harvest samples, the Relative ME was calculated for each of the quality attributes at the U.S. Harvest Aggregate and the Harvest Areas. The Relative ME for the quality factor results was less than ± 10%, except for three attributes for the U.S. Harvest Aggregate and the Early Harvest Area: BNFM, foreign material, and total damage. While the lower level of precision for these quality factors is suboptimal, these levels of Relative ME do not invalidate the estimates.