1 / 19

Angler Surveys: Internet Samples Can Be Representative

Angler Surveys: Internet Samples Can Be Representative. Pathways 2014 Conference Estes Park, CO October 5-9. Kjetil Henderson and Dr. Larry Gigliotti. Why Use Internet Surveys. Mail Surveys $25-30,000. Internet Surveys

Download Presentation

Angler Surveys: Internet Samples Can Be Representative

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Angler Surveys: Internet Samples Can Be Representative Pathways 2014 Conference Estes Park, CO October 5-9 Kjetil Henderson and Dr. Larry Gigliotti

  2. Why Use Internet Surveys Mail Surveys $25-30,000 Internet Surveys • Cost savings particularly relevant for a regularly occurring monitoring survey.

  3. Types of Internet Surveys Open invitation Invitation sent by mail Invitation sent by email (w/ link to website)

  4. 2 Potential Problems w/ Internet Surveys Incomplete Email Coverage Relatively Low Internet Response Rate

  5. Email Coverage – 44% 2012 Licenses

  6. Internet Response Rate – 32% 2012 Licenses

  7. Incomplete Coverage Are anglers who provide an email address similar or different from anglers not providing GFP with an email address for the types of variables measured in a statewide monitoring survey? Procedure All Anglers Email 44% No Email 56%

  8. Internet Nonresponse Are anglers who respond to an Internet survey similar or different from nonrespondents for the types of variables measured in a statewide monitoring survey? Procedure Email Anglers 44% Non-Respondent 68% Respondent 32%

  9. Research Can an Internet survey of anglers collect representative data similar to a mail survey? Mail Survey of No Email Group 56% Internet Respondents 14% Mail Survey of Internet Nonrespondents 30%

  10. Analyses – 2011 & 2012 Variables Licenses (2011 & 2012) Resident Combination Resident Annual Resident Sr. Annual Resident Jr. Combination Resident Sr. Combination* Nonresident Annual Nonresident Family Nonresident 3-Day Nonresident 1-Day • Rating of Fishing Quality • Satisfaction • Importance of Fishing • Days of Fishing • Boat Fishing • Percent Fishing • Age • Sex

  11. Results – Incomplete Coverage Significant Variables Summary 130 statistics tests 22 significant (17%) 100% of the significant variables were considered negligible or weak as measured by effect statistics Cramer’s V or Cohen’s d Age……………………………….…9 Sex……………………………….….6 Percent Fishing…………….....4 Boat Fishing……………………..3 Rating of Fishing Quality….0 Satisfaction…......................0 Importance of Fishing………0 Days of Fishing…………………0 Bonferroni correction (α ≤ 0.00313) was used to control for type one error as multiple comparisons were performed per license.

  12. Results – Internet Nonresponse Significant Variables Summary 130 statistics tests 11 significant (8%) 100% of the significant variables were considered negligible or weak as measured by effect statistics Cramer’s V or Cohen’s d Age……………………………….…4 Sex……………………………….….5 Percent Fishing…………….....2 Boat Fishing……………………..0 Rating of Fishing Quality….0 Satisfaction…......................0 Importance of Fishing………0 Days of Fishing…………………0 Bonferroni correction (α ≤ 0.00313) was used to control for type one error as multiple comparisons were performed per license.

  13. Summary – Internet Data Sex Bias • All License Types: Females under-represented in the Internet Data (8% overall). Summary – Internet Data % Fishing Bias All License Types: Percent fishing over-represented in the Internet Data.

  14. Sex & Age Bias in Internet Data • Sex & Age were significantly related to many variables: 64 significant (34%) out of 187 tests • 94% of the significant variables were very small or small based on effect size statistics.

  15. Correcting Internet Survey Data for Coverage and Nonresponse Bias • The small differences due to sex and age biases in the Internet data can be corrected by weighting (population values know from license purchase information). • The percent fishing bias in the Internet data can be corrected using the values identified in the 2011 and 2012 study (Percent fishing varies by license type but is relatively stable from year to year).

  16. Proportion by Numbers Days Fished

  17. Conclusion The Internet survey (corrected for sex, age & percent fishing biases) can provide accurate & reliable data similar to a mail survey for the variables normally measured in the statewide angler monitoring survey.

  18. Acknowledgements • SDSU, Department of Natural Resource Management • South Dakota Cooperative Fish and Wildlife Research Unit • South Dakota Game, Fish & Parks

More Related