Influences on charitable giving for conservation: Online survey data of 1,331 respondents across the US, August 2017

CREATOR(S): Chelsea Batavia, Michael P. Nelson
ORIGINATOR(S): Chelsea Batavia
OTHER RESEARCHER(S): Julia A. Jones, Hannah Gosnell, Jeremy T. Bruskotter, John A. Vucetich
National Science Foundation, award # 1725530 National Science Foundation Long-Term Ecological Research Program at the H.J. Andrews Experimental Forest, DEB 1440409
19 Apr 2018
attitudes and perceptions, environmental ethics, public assessments, public values, conservation
The first and primary objective of this study was to investigate whether people report stronger support for the cause of conservation (as measured by attitudes, hypothetical donations, and donations) when they are approached with an outreach message suggesting nature conservation is good for humans; good for nonhumans; or good for both humans and nonhumans. Data were collected to inform a current debate in the conservation community about ethical motivations for conservation and effective strategies for winning social support. A secondary objective was to test whether commonly observed and politically conditioned patterns of response to messages highlighting different clusters of moral foundations (individualizing or binding; see Methods) are also observed in the context of persuasive communications for conservation. This more theoretical objective served practical purposes as well, by controlling for a message variable (moral foundation) that may have otherwise confounded results related to objective one.
Experimental Design - SS007:
Description: Each study participant was randomly assigned to view one of seven experimentally manipulated messages. In a 3 x 2 factorial design, each flyer featured a verbal message that varied systematically along two experimental factors: beneficiary (who benefits from conservation), and moral foundation (in what sense is conservation presented as a moral issue (Graham et al. 2011). Beneficiaries were either (1) only human; (2) only nonhuman; or (3) both human and nonhuman. Moral foundations were either (1) individualizing (presenting conservation as an issue of care/avoidance of harm and fairness); or (2) binding (presenting conservation as an issue of duty, community loyalty, and the sanctity of nature). The seventh flyer was a control, which did not emphasize specific beneficiaries or moral foundations. Other aspects of the messages (e.g., grammatical structure, length) were kept consistent to the extent possible, and the image depicted in the flyers was held constant across all seven messages. The goal of the message manipulation was to test differences between treatment groups for several response variables, including attitudes, hypothetical donation, moral salience (all measured with survey questions), and donation (measured according to a procedure described in “Field methods”). Citation: Graham, J., Nosek, B.A., Haidt, J., Iyer, R., Koleva, S., and Ditto, P.H. (2011). Mapping the moral domain. Journal of Personality and Social Psychology 101: 366-385. doi: 10.1037/a0021847
Field Methods - SS007:

The survey was designed using the online application Qualtrics, and administered to an online panel of nationwide U.S. respondents. Online panels are comprised of individuals who contract with an organization or corporate entity (in this case, Qualtrics, LLC) to complete surveys for research, marketing, or other analytics purposes, in return for compensation. The survey was administered by a designated management team at Qualtrics, LLC. The sampling procedure was stratified to capture a roughly even admixture of self-identified liberals and conservatives. Panelists were emailed a link to a new survey. If they followed the link, they were directed to the first page of the survey, which asked for consent to participate in a research study. By proceeding to the next page respondents signified their consent. Once the targeted sample size had been achieved, the survey was closed to additional respondents.

To measure donation we employed a procedure documented by Clements et al. (2015). Respondents were told they would receive $5, in addition to their normal panelist earnings, and were offered the opportunity to designate some proportion of the $5 as a donation. After choosing to donate (or not), respondents viewed a de-debrief page informing them that their donations had been recorded as data, a detail that had been withheld from the initial consent document so as to render the donation decision context as naturalistic as possible. Because of the minor deception, in accordance with standards stipulated by the Oregon State University Institutional Review Board, respondents were offered the opportunity to withdraw their data from the sample without forfeiting payments owed to them. Data from respondents who chose to withdraw (N = 269) are not included in the dataset. Upon completion respondents were thanked for their participation and payment was issued according to their contracted arrangement with Qualtrics, LLC. The research team calculated additional payments owed to respondents from the donation incentive (see “experimental design”), and these payments were issued to panelists by Qualtrics, LLC.

Citation: Clements, J.M., McCright, A.M., Dietz, T., and Marquart-Pyatt, S. (2015). A behavioural measure of environmental decision-making for social surveys. Environmental Sociology 1: 27-37. doi: 10.1080/23251042.2015.1020466
Quality Assurance - SS007:
Description: Several attention check questions were included to ensure respondents were actually reading and responding to the survey items. Attention checks followed this or a similar format: “If you are reading this question, please mark your response 1, Strongly Disagree.” Only survey responses from individuals who entered the correct numbers on these items were retained in the dataset. Survey responses with missing cell values (e.g., if an individual chose not to respond to certain survey items) were not included in the final dataset. In addition, data were assessed for quality concerns by comparing each response time to the median response time for the total sample. Individuals whose response time fell outside a reasonable margin of the median time were individually examined for patterns suggesting unreflective or automatic responses. Where strong evidence pointed to poor data quality, those responses were removed from the dataset.
Data Entry - SS007:
Description: Survey responses were recorded directly in an online database, with respondents labeled by unique Qualtrics identifiers. The Qualtrics application automatically compiles data in spreadsheet form and makes it accessible for download by the survey owner in a variety of formats (Excel, SPSS, etc.). Following completion of data collection, additional respondent demographic information was appended to survey data by the Qualtrics management team from information housed in panelists’ Qualtrics user profiles. This information was matched to panel respondents by the Qualtrics management team using unique study identifiers. No personally identifiable information about panel respondents was available to the research team at any time.
Data was collected online from a panel of respondents across the U.S.
one discrete time period