NASS Review of FAS OMB Docket_Trade Shows Surveys

NASS Review of FAS OMB docket for

Trade Missions and Shows Revised Applications and Post- Event Surveys

Lindsay Drunasky

Mathematical Statistician

Summary, Estimation, and Disclosure Methodology Branch

Background:

This docket describes a data collection effort by FAS from participants of USDA endorsed trade shows, trade missions, and virtual trade events. The data collection instruments include two applications for the events and three post-event surveys to gain feedback. It is described as a new information request, but the docket also states that the survey instruments have been used previously.

Part A:

There are no statistical concerns for the two applications submitted. However, Part A states that the post-event surveys are used to measure the Key Performance Indicator (KPI) by asking 12-month projected sales and on-site sales. In addition, the response to question A2 indicates that only 70% of participants are expected to respond. If the KPI is considered an official estimate, NASS recommends accounting for the nonresponse in some way, and those details should be included in Part B.

Part B:

As noted above, the current responses in Part B are lacking details if the KPI is considered an official estimate. Specifically, question B2 should have more information on how the KPI will be calculated/estimated. All participants are asked to respond to the post-event survey so there are no concerns related to sampling. However, in question B3, there are no methods described to maximize response rates. At least one attempt of nonresponse follow up would be recommended. Once data collection is completed, then the remaining nonresponse should be accounted for in some way.

Remarks and Suggestions:

It is not clear to the reviewer if the post-event surveys will be presented in the exact same way as they are included with the docket. If that is true, perhaps there is a better way for the respondent to report the numerical data such as projected 12-month sales, on-site sales, and number of products and/or meetings, etc. Currently, there are multiple pages of options for each question which greatly increases the length of the survey. It is possible that the electronic version of the survey looks much different to the respondent.