Survey Design and Sampling
Survey Design
For the 2018/2019 Export Cargo Report, the target population was yellow commodity corn from the 12 key U.S. corn-producing states representing approximately 95% of the estimated 2018/2019 U.S. corn exports. A proportionate stratified sampling technique was used to ensure a sound statistical sampling of U.S. yellow corn exports. Two key characteristics define the sampling technique for this report: the stratification of the population to be sampled and the sampling proportion per subpopulation or stratum.
Stratification involves dividing the survey population of interest into subpopulations called strata. For the Export Cargo Reports, the key corn-exporting areas in the United States are divided into three geographical groupings, which we refer to as ECAs. These three ECAs are identified by the three major pathways to export markets:
- The Gulf ECA consists of areas that typically export corn through U.S. Gulf ports;
- The Pacific Northwest ECA includes areas that usually export corn through Pacific Northwest ports; and
- The Southern Rail ECA comprises areas that generally export corn by rail to Mexico.
Using data from the USDA, each ECA’s proportion of the total expected annual yellow corn exports for the 2018/2019 corn marketing year was calculated. This average share of exports was used to determine the sampling proportion (the percent of total samples per ECA) and, ultimately, the number of yellow corn samples to be collected from each ECA. The specified sampling proportions for the three ECAs are shown below.
The number of samples collected within each ECA was established so the Council could estimate the true U.S. Aggregate average for the various quality factors with a certain level of precision. The level of precision chosen for the Export Cargo Report was a Relative ME of not more than ±10%. A Relative ME of ±10% is a reasonable target for biological data such as these corn quality factors.
To determine the number of samples for the targeted Relative ME, ideally, the population variance (i.e., the variability of the quality factor in the corn exports) for each of the quality factors should be used. The more variation among the levels or values of a quality factor, the more samples needed to estimate the true mean with a given confidence limit. In addition, the variances of the quality factors typically differ from one another. As a result, different sample sizes for each of the quality factors would be needed for the same level of precision.
Since the population variances for the 15 quality factors evaluated for this year’s corn exports were not known, the variance estimates from last year’s Export Cargo Report were used as estimates of the population variance.
Sampling
The sampling was administered by FGIS and participating official service providers as part of their inspection services. The FGIS field offices indicated that 2018 corn was reaching export points in October 2018. Therefore, FGIS sent instruction letters to the Gulf and Pacific Northwest field offices and to the domestic inspection offices, and the sampling period began in November 2018, for the three ECAs. The FGIS field offices in the respective ECAs responsible for overseeing the sample collection within their region were as follows: Gulf – New Orleans, Louisiana; Pacific Northwest – Olympia, Washington (Washington State Department of Agriculture); and Southern Rail – FGIS Domestic Inspection Operations Office in Kansas City, Missouri.
While the sampling process is continuous throughout the loading of an ocean-bound vessel, a shipment or “lot” of corn is divided into “sublots” for the purpose of determining the uniformity of quality. Sublot size is based on the hourly loading rate of the elevator and the capacity of the vessel being loaded. Sublot sizes range from 30,000 to 120,000 bushels. All sublot samples are inspected to ensure the entire shipment is uniform in quality.
Representative sublot samples from the ports in the Gulf and Pacific Northwest ECAs were collected as ships were loaded, and only lots for which quantitative aflatoxin testing was being performed were sampled. Samples for grading are obtained by a diverter sampling device approved by FGIS. The diverter sampler “cuts” (or diverts) a representative portion at periodic intervals from a moving stream of corn. A cut occurs every few seconds, or about every 200 to 500 bushels (about 5.1 to 12.7 metric tons), as the grain is being assembled for export. The frequency is regulated by an electronic timer controlled by official inspection personnel, who periodically determine that the mechanical sampler is functioning properly.
Sublots ending in zero, three, five and seven from each lot during the survey period were sampled. This was the same sampling frequency for the Pacific Northwest and Gulf ECAs as last year’s export cargo survey. For each sample, a minimum of 2,700 grams was collected by the FGIS field staff and the Washington State Department of Agriculture.
For the Southern Rail ECA, representative samples were taken at domestic interior elevators using a diverter sampler to ensure uniform sampling. A cut is taken about every 200 bushels (about 5.1 metric tons). Only trains of yellow corn inspected for export to Mexico were sampled. Unlike the samples collected from the Gulf and Pacific Northwest ECAs, which collected additional samples at the time of loading specifically for this report, the Southern Rail ECA official service providers submitted file samples. These samples were collected and tested for grade factors and aflatoxin at the time of sampling and then kept on file at the official service providers to be retested in the case of disputes. Each file sample weighed approximately 1,000 grams and represented a composite of five railcars. For each train sampled, three file samples were mailed to the Illinois Crop Improvement Association’s Identity Preserved Grain Laboratory (IPG Lab) when their retention dates were reached, which were generally 30 days after loading. Upon arrival at the lab, the three file samples were composited into a single sample to undergo the chemical composition, physical factor and DON tests. The grade factor results from the three file samples were averaged to represent a single sample. Aflatoxin results were used only if all three file samples were lower than 5 ppb. If one or more of the three file samples tested greater than 5 ppb, then the composited sample was tested at IPG Lab for aflatoxin using the EnviroLogix AQ 309 BG test kits.
Refer to the “Testing Analysis Methods” section for the description of the testing methods employed for the study.