Justification for change

SUPPORTING STATEMENT

Economic Value of Puerto Rico’s Coral Reef Ecosystems for Recreation/Tourism

Non-substantive Changes Request

0648-0713

Summary of Changes

Based on the pretest of the Visitor’s Surveys, we are requesting the following changes:

1. Due to lack of funding, we are not able to do the pre-test or the full survey for the Residents of Puerto Rico, so we are dropping this from the current request. This results in a decrease in our overall burden hour request for the full survey from 2,200 hours to 1,319 hours.

2. The Internet Survey was not successful and will be replaced with an On-site Survey. The On-site Survey asks all the same questions, except the non-market economic value choice questions were reduced from four choices per respondent to two choices to reduce time burden. Also, the expenditure and importance-satisfaction questions were moved to the Airport Survey as mailbacks. They were already mailbacks as secondary back-ups to the Internet Survey for visitors that did not join the Internet Panel; now they are the primary source of this information.

3. Total burden hours for the Visitor Survey changes were due to a combination of decreases for the on-site airport survey and increases in the on-site survey outside the airports. We reduced the number of airport surveys required to obtain enough recruits into the Internet Survey from 1,500 per season to 750 per season.  With two seasons that is a reduction from 3,000 to 1,500 of these 5-minute responses. For the mailback surveys, we had estimated a 45% response for the expenditure survey (675 responses at 35 minutes = 394 hours, and 60% for the satisfaction survey (900 responses at 20 minutes = 300 hours).

4. However, we had previously estimated the need for 500 per season for the Internet survey or 1,000 for both seasons; this now had to increase from 500 15-20-minute responses per season to 750 per season, so a total increase from 1,000 to 1,500, for the onsite survey.  This was because we had to go from four choices per respondent in the Internet survey to two choices per respondent for the onsite survey.  We did not need to double our sample since the optimal design indicated we would have sufficient sample size with 750 per season instead of 1,000. By expanding to 18 versions with two choices per respondent, we were able to achieve the 36 choices required for the optimal design, while also assuring statistically reliable estimates. All the Visitor Survey questionnaires and supporting materials are in Appendix E: Visitor 00

Overall, the full visitor survey increased from 900 to 1,319 hours.

Total hours for the visitor pretest (150) added to the 1,319 hours for the full survey equals a final total of 1,469 hours.

5. Per approval 0648-0713 we addressed two objectives: 1) design the dollar bid amounts for the non-market economic value choice questions and 2) test for item non-response. All answered the non-market economic value questions with only three protest bids. Four respondents did not answer intensity of reef use questions and 29% did not answer the household income question. This latter non-response is twice as high as the airport survey, so additional interviewer training is required to correct this problem. See Appendix D: Pre-test results.