Atlantic Wind Survey Revised Supplement B (clean) OMB 2015-12-11

Supporting Statement B

Atlantic Offshore Wind Energy Development–Public Attitudes, Values, and Implications for Tourism and Recreation

OMB Control Number 1010-XXXX

1. Describe (including a numerical estimate) the potential respondent universe and any sampling or other respondent selection method to be used. Data on the number of entities (e.g., establishments, State and local government units, households, or persons) in the universe covered by the collection and in the corresponding sample are to be provided in tabular form for the universe as a whole and for each of the strata in the proposed sample. Indicate expected response rates for the collection as a whole. If the collection had been conducted previously, include the actual response rate achieved during the last collection.
   
   

Our analysis will use separate samples: a General Population Sample and a Beachgoer-Only Sample. The same survey will be administered for both samples. GfK Custom Research (GfK) (formally Knowledge Networks) will conduct the survey using their Knowledge Panel -- a nationally representative and probability-based online panel.

The response rate for the sample used to form the Knowledge Panel is near 10%. The details of the how that screening is done is covered in the GfK attachment on sampling & weighting procedures included along with the submission of this supplement. Given the efforts made to make the sample representative of the population along numerous demographic and other lines in pre- and post-sampling procedures, we believe the Knowledge Panel is the best on-line sample available for our purposes. Also, there are numerous examples of past projects approved by OMB, which use GfK’s Knowledge Panel. Some of these are reported in the GfK Documentation sent here. There are also many published peer-review articles in the economics literature using the Knowledge Panel, which is testimony from the profession on the quality of the data. We have listed a several articles based on GfK data at the end of this section.

The population for the General Population Sample is all individuals over 17 years old in the 20 States (including Washington DC) shown in Table 1 below. The population for the Beachgoer-Only Sample is all individuals over 17 years old in the same 20 states who have visited a beach on the east coast in the last 12 months. Beachgoers, of course, are of special interest in our analysis. Our definition of an “east coast beach” is an ocean beach in any state from Massachusetts to South Carolina. In both samples, respondents are randomly drawn from their respective populations.

The target sample size for the General Population Sample is 500. Since we expect a Knowledge Panel response rate of 85%, we need to solicit around 588 people (500 = .85 x 588). The response rate of 85% may seem high, but recall that we will be using a prescreened sample.

The target sample size for the Beachgoer-Only Sample is 1,600. Since we expect about 35% of the population will be beachgoers based on past surveys and again expect a Knowledge Panel response rate of 85%, we will need to solicit around 5,378 people (1,600 = .85 X .35 x 5,378). The past survey used to estimate the beach-going rate was a Knowledge Panel survey done by the University of Delaware on beach use in the Mid-Atlantic from a similar set of east coast states in 2005.

As noted above, the same survey is administered for both samples. The survey protocol, however, is different for the two samples. The question on page 14 of the survey (included here along with the supplements) is used to identify beachgoers and non-beachgoers. For the General Population Sample, the response to this question directs beachgoers and non-beachgoers to different survey questions tailored to their beach use. For the Beachgoer-Only Sample, the response to this question is used as a screen, wherein non-beachgoers are dropped from the sample and beachgoers are included. Otherwise, the administration and construction of the survey for the two samples is the same.

Since there are beachgoers in the General Population Sample and the Beachgoer-Only Sample, we expect a total of approximately 1,775 beachgoers in total (.35 x 500 from the General Population Sample plus 1,600 from the Beachgoer-Only Sample). Recall, the expected number of beachgoers from the population is about 35%. That leaves 325 (.65 x 500) non-beachgoers for the analysis. Table 2 shows the expected response data from the two samples.

General Population Sample
Completed Surveys based on 85% response rate	500*
Non-response or dropped from survey	88
Total Number of People Solicited	588
Beachgoer-Only Sample
Completed Surveys based on 85% response rate and 35% of the population being beachgoers	1,600
Non-response, dropped out, or not qualified	3,778
Total Number of People Solicited	5,378

Table 1: States Included in the Sample

States
Maine	Connecticut	Maryland	Ohio
New Hampshire	New York	Virginia	West Virginia
Vermont	Pennsylvania	North Carolina	Tennessee
Massachusetts	New Jersey	South Carolina	Kentucky
Rhode Island	Delaware	Georgia	Washington DC

Table 2: Expected Response Data from Two Samples

* Of these 325 are non-beachgoers and 175 are beachgoers.

A sampling of recent published studies using GfK data:

Wallmo K. & Lew, D. (2011). Valuing improvements to threatened and endangered marine species: an application of state preference choice experiments. Journal of Environmental Management 92(7).

Hiller, R., Savage, S. & Waldman, D. (2015). Market structure and media diversity. Economic Inquiry 53(2).

Petrolia, D., Landry, C. & Colbe, K. (2013). Risk preferences, risk perceptions, and flood insurance. Land Economics 89(2).

Mansfield, C., Finkelstein, E., Wood, D. & Rowe, B. (2013). Evaluating welfare improvements from changes in homeland security policies. Institute for Homeland Security Solutions.

Cameron, T. & DeShazo, J. (2013). Demand for health risk reductions. Journal of Environmental Economics and Management 65.

Brown, D., Poulos, C., Johnson, F., Chamiec-Case, L. & Messonnier (2014). Adolescnet girls’ preferences for hpv vaccines: a discrete choice experiment. Preference Measurement in Health, Advances in Health Economics and Health Services Research, 24.

Smith V, Mansfield C, & Klaiber A. (2013). Terrorist threats, information disclosures, and consumer sovereignty. Information Economics and Policy 25(4).

2. Describe the procedures for the collection of information including:

* Statistical methodology for stratification and sample selection,

* Estimation procedure,

* Degree of accuracy needed for the purpose described in the justification,

* Unusual problems requiring specialized sampling procedures, and

* Any use of periodic (less frequent than annual) data collection cycles to reduce burden.

As noted above, we are interested in beachgoer and non-beachgoer populations, but have a special interest in beachgoers. For this reason, we oversample the beachgoer population. Our sample sizes are sufficient to allow us to have good inference on both groups, but some finer regional specificity for beachgoers (e.g., impacts in New England versus the Carolinas). Since we randomly draw from both of these populations, there is no specialized estimation procedure needed. Also, GfK provides sample weights as needed to make the samples representative of the population.

3. Describe methods to maximize response rates and to deal with issues of non-response. The accuracy and reliability of information collected must be shown to be adequate for intended uses. For collections based on sampling, a special justification must be provided for any collection that will not yield "reliable" data that can be generalized to the universe studied.

We have taken the usual steps used to keep response rates high: short survey, brief but interesting questions, and so forth. The single largest adjustment made to the survey to increase response rate is to ask people detailed trip information on one site only. This is instead of asking for data on all trips over a season, which is common in travel cost studies. This streamlines the survey, allows us to obtain more detailed trip-specific data, improves recall, and keeps the respondent focused on the beach where the wind project will be introduced.

Since we are working with a prescreened sampling frame, nonresponse issues are kept to a minimum. Also, GfK has demographics on all panel members so what we need to gather on individuals is limited. We will compare respondent and non-respondent characteristics for statistically significant differences, report these results, and use them to reweight the data if necessary.

4. Describe any tests of procedures or methods to be undertaken. Testing is encouraged as an effective means of refining collections of information to minimize burden and improve utility. Tests must be approved if they call for answers to identical questions

from 10 or more respondents. A proposed test or set of tests may be submitted for approval separately or in combination with the main collection of information.

We have conducted a small pretest (< 10 respondents), which led to refinement and simplification of the survey including the focus on last-trip data mentioned above, reducing the number of choice questions each respondents faces, and other wording changes. These were one-on-one pretests where respondents were interviewed following the survey.

5. Provide the names and telephone numbers of individuals consulted on statistical aspects of the design and the name of the agency unit, contractor(s), grantee(s), or other person(s) who will actually collect and/or analyze the information for the agency.

George R. Parsons

Professor of Economics and Marine Policy

University of Delaware

[email protected]

(302) 831-6891

Jeremy Firestone

Professor of Marine Policy

University of Delaware

[email protected]

(302) 831-0228

Wendy Mansfield

Vice President

GfK Custom Research

[email protected]

(202) 686-0933