NASS Review Comments

2/13/2012

Michael Jacobsen

OMB DOCKET FOR THE 2012 MCKENZIE RIVER AND TRAIL VIEW SURVEY AND FLATHEAD WILD AND SCENIC RIVER VISITOR SURVEY

I have reviewed this OMB docket and have the following comments.

This is a very well-organized document. The authors provide a strong justification for this survey and a fairly detailed plan for capturing the data. However, I have questions about some of the parts of this survey program. I will use the following abbreviations for convenience:

• FL = Flathead River and Trail Survey

• MKR = McKenzie River Survey

• MKT = McKenzie Trail Survey

Will the data from the McKenzie River and Flathead River surveys be compared to each other? This was not made clear in the docket but I am assuming not.

Why do the McKenzie and Flathead surveys ask different questions? Are the visitor populations to each park so different that different questions are needed? For example, zip code is asked for on FL but not MKR and MKT. MKR and MKT, but not FL, have questions on user experience. The MKR and FL ask for different types of boats. Finally, MKR question 22 and MKT question 19 ask the respondent to specify a birth year, but FL question 1 provides ranges of ages. If the populations are very different and the three surveys will not be compared, then this is a moot point. However, it seems strange that the same questions are asked in different formats.

What is the observational unit, individuals or groups? I am asking this because I can see the situation in which one of your enumerators receives a completed questionnaire from two different members in the same group. How will you know if they are part of the same group? Furthermore, in MKR question 25 and MKT question 27 you ask for expenses for both the respondent and the respondent’s group, which seems to imply a total expense for the group. What if the respondent does not know? This particular question would be adequate for an individual but not for a group.

Will you accept multiple responses from the same respondent? I am thinking of the situation where a visitor visits a park on two different occasions that happen to coincide with your randomly selected days. Will you accept both responses? This could affect data analysis, especially when it comes to visitor counts.

In Part B, you mention that the MKR and MKT have similar questions with different “… activity- and experience-specific items …” However, the same items are present on both questionnaires. For example, on MKT question 13, answers 1, 18, 20-24, and 26 seem specific to rivers users, while on MKR question 15, answers 3, 7, and 27-31 seem specific to trail users. Do these activity-specific answers need to be on both questionnaires?

In conjunction with my previous question, can visitors use both the river and the trail? If so, will they answer both the MKR and MKT? I am assuming that very few people will have the energy to do both so this should not be a problem.

Finally, I am confused about your “non-response” analysis. What do you consider to be a non-respondent? For example, if there is a very busy day at one of the surveying locations and the enumerators are not able to approach all groups, does that mean that those people are non-respondents? What if they refuse because of the weather? What is the Survey Log that you mentioned? You are going to include it with the OMB Docket but I did not see it. However, its description implies that much of the data to be collected about non-respondents may be hard for the enumerators to actually capture, such as group size and composition. I think simple counts of refusals would provide more information than haphazardly capturing hard-to-get information.