Appendix D2. National Agricultural Statistics Service (NASS) Comments

Appendix D2.
National Agricultural Statistics Service (NASS) Comments

Review of OMB Part B: Collection of Information Employing Statistical Methods Supporting Statement for OMB Clearance for

FDPIR Paraprofessional Nutrition Training Assessment for Indian Tribe Organizations

Review performed by:

Sarah Goodale, Mathematical Statistician

USDA, NASS

Methods Division/Summary Estimation and Disclosure Methodology Branch/Census Data Section

T: 202-720-0791

Part A:

I felt that this section was well worded and provided a lot of detail on why the survey is being performed. It was very comprehensive and well written.

Part B:

It appears that this survey is intended to be a fact finding mission to determine what the current training is, where training is needed, and the interest in the training for ITO paraprofessionals. Based on the objectives of the study, there are no estimates that are produced as a final product. All the data is qualitative (mentioned in the docket), which will be summarized into a written recommendation based on the answers to the survey. It should be mentioned that the survey is asked in question with a freeform answer as the data collected. It is possible that all the data collected will be different from one another.

Response

The purpose of this study is not to produce a generalizable estimate, or catalogue all of the ITOs’ thoughts about a paraprofessional nutrition training, but rather, to gauge interest in offering a paraprofessional nutrition training program from select ITO Directors, ITO Staff, and Key Stakeholders. This information will help inform if and how the paraprofessional nutrition training program should be created and delivered. The respondents will be asked about their opinions on the format/delivery of a paraprofessional nutrition training, potential topics to cover, and motivations for participation, for example. Section B.1 was revised to better communicate this purpose.

I think that it needs to be clear that the data collected is for research. What will be published is not actually data estimates but recommendations on what has been seen to work well and where there is a need for future improvement. To get the best results possible a different sampling procedure is needed. The current sampling design is a sample of convenience which will introduce bias into your survey results. Asking people who should participate will not lead to a clear picture on the intended study. This could lead to people non-proportionally responding in a favorable or unfavorable manner. Adding a randomized sample design will remove this bias from the sample. To randomize the sample a clear population (or multiple) populations need to be clearly defined. Based on the information provided here is an example on how the sampling design can be done: For example Group 1: ITO directors: All the ITO directors are considered the full population. These directors can be broken or sorted into groups based on the geographic area and/or the size of the program. Then a simple random sample (or systematic sample) should be taken from each group to reach the needed sample size. By randomly taking a sample from the population you eliminate the bias of a sample of convenience. The next group is FDPIR Staff: These names are coming from the ITO directors. The ITO directors can bias the data coming in from the survey. They will give you the names of people to support their point of view (good or bad). Instead a random sample of all the FDPIR staff should be taken or a random sample of the people working under the ITO director. This will help provide a more well-rounded study.

Response

The purpose of this study is not to produce a generalizable estimate, or catalogue all of the ITOs’ thoughts about a paraprofessional nutrition training, but rather, to gauge interest in offering a paraprofessional nutrition training program from select ITO Directors, ITO Staff, and Key Stakeholders. This information will help inform if and how the paraprofessional nutrition training program should be created and delivered. The respondents will be asked about their opinions on the format/delivery of a paraprofessional nutrition training, potential topics to cover, and motivations for participation, for example. We have not changed our data collection methodology, as the purpose is to obtain opinions from select individuals. Additionally, a random sample based on geographic and size strata, for example, may underrepresent cultural communication differences which is a critical consideration for a useful assessment with ITO Directors and Staff. Section B.1 was revised to better communicate this purpose.

The response rates and adjustment to the sample appear reasonable and appropriate for the study.

For the testing it might have been better to test on more than one person to catch any potential flaws. While the actions taking in the testing phase looked appropriate it is always better to get more than one opinion.

Response

The feedback received from the four pretest respondent types was highly similar; thus, we chose to only conduct pretests with these four so as not to add burden.