Study Proposals from Our First Research Workshop on Effective Animal Advocacy
Summary
Last month ACE hosted the 2017 Research Workshop on Effective Animal Advocacy at Claremont Graduate University (CGU) in California. The event was intended to complement our 2016 Symposium on Multidisciplinary Research in Effective Animal Advocacy, held at Princeton University last fall. The Symposium was open to all and featured presentations by academics and advocates on cutting-edge research as well as avenues for further research. This year’s workshop was a way to bring together those most likely to work on research in animal advocacy in the near future and give them the tools for conducting rigorous, practical studies—as well as opportunities to work together on developing research proposals.
This event brought together 36 academics and advocates who worked collaboratively in small groups to develop ideas for empirical research that will advance our understanding of effective animal advocacy. Workshop attendees were placed into groups of 3–4 and spent the weekend developing concrete research ideas and soliciting feedback. Each group included at least one attendee with an academic background and at least one attendee with a background in animal advocacy—we hoped that this matching system would help academics better understand the type of research that is most useful for animal advocates, and that it would help to yield study ideas that were both academically rigorous and highly actionable. At the end of the workshop, each group presented their research proposal to all attendees. Abstracts from these proposals are included below.
The group work sessions were supported by networking time, updates on current research in animal advocacy, and several presentations. Our Director of Research Allison Smith presented on intervention research, Catherine Amiot (University of Quebec—Montreal) presented on foundational research, Joshua Tasoff (Claremont Graduate University) presented on study design, Krystal Caldwell (Mercy For Animals) presented on survey data, and Harish Sethu (Humane League Labs) presented on observational data.
It was exciting to see so much collaboration taking place at the workshop, and we hope that the ideas generated there—and the connections made among attendees—will inspire future projects. We intend to hold the workshop at somewhat regular intervals, and expect to host further symposiums a bit more intermittently so as to allow time for new research to develop. If you are interested in attending future research workshops, please get in touch!
We would like to thank Joshua Tasoff, Associate Professor of Economic Sciences at CGU, for his help facilitating our partnership with the university. We would also like to extend our utmost gratitude to Jorge Lugo for making this groundbreaking workshop possible.
Research Proposal Abstracts
The following research proposals were generated at the workshop:
Occidental College panel dataset of actual vegetarian/non-vegetarian individualized consumption data
Kieran Greig, Animal Charity Evaluators
Jacob Peacock, Humane League Labs
Joshua Tasoff, Claremont Graduate University
Occidental College’s dining hall presents a rich dataset of individual-level food consumption data of three meals per day, 20+ weeks per year for the ~2,000 students at the college. This data provides a unique opportunity in intervention research; with one quarter of students (~500) cycling through the institution each year, new interventions could be tested routinely. Even without intervention, such a detailed dataset at a crucial time for dietary change can help us understand long-term changes toward a veg*n diet. Consenting students could be further subjected to a battery of psychological and economic measures which are then correlated with consumption and intervention outcomes. Supplementing with administrative data would provide avenues for correlation analysis, while social network data would allow measurement of social interaction effects. Currently, animal- and plant-based meals are equally priced; by intervening with the price, we could measure cross-product elasticities. Furthermore, we would like to perform a systematic search for similar dining service data at other institutions using either The Humane League’s interns at colleges across the United States or hired research assistants.
Review of the formation of academic fields to establish the welfare biology field
Ozy Brennan, Wild-Animal Suffering Research
Mark Budolfson, University of Vermont
Zachary Groff, Innovations for Poverty Action
Advocates concerned with wild animal suffering lack essential research to guide wild animal advocacy, and the academic study of welfare biology would be a potentially productive source of rigorous research on the subject. In order to research how we can establish a welfare biology academic field, we propose to study the formation of novel academic fields. We will conduct a qualitative study of fields with a normative focus on others who are not in a position to advocate on their own behalf. Our study will begin with a scoping literature review on the history of eight relevant academic fields: conservation biology, animal welfare science, environmental engineering, artificial intelligence safety, environmental ethics, environmental economics, development economics, and positive psychology. In the initial review, we will identify questions and key figures in the history of the relevant fields. Next, we will conduct interviews with each of the key figures in the history of the relevant fields. We will aim to understand what worked in establishing new fields, what failed, and what the context was so that we know how much it can be extended to the context of welfare biology. We will ask experts for further sources to look into on the establishment of their fields and review those sources following the interviews. After conducting interviews and concluding the literature reviews, we will synthesize common trends across interviews and write up a qualitative report on effective strategies to establish new, other-oriented academic fields.
An exploration of animal advocacy messaging that appeals to political conservatives
Lisa Kramer, University of Toronto
Shiva Pauer, University of Vienna
Jacy Reese, Sentience Institute
We set out to explore how animal advocacy messages can appeal to conservatives, a demographic group that may currently be relatively neglected by advocacy campaigns. We plan to employ methods that have a theoretical foundation from the psychology literature and have been used in other applied fields of research such as climate change advocacy. We aim to better understand how receptiveness to different advocacy messages varies across political identifications. We plan three exploratory studies to explore this research question, using Amazon Mechanical Turk to recruit participants. In all three studies, we will have two treatment groups and a control, and we will measure how our treatments differentially influence attitude measures and behaviors relevant to animal advocacy. In Study 1, we explore moral foundations, using statements incorporating values that prior research suggests are embraced by conservatives. In Study 2, we investigate whether the framing of an op-ed piece influences the degree to which the reader is swayed by the arguments therein. In Study 3, we explore system-sanctioned change.
Comparing environmental, animal welfare, and health arguments’ efficacy in reducing animal product consumption
Katie Cantrell, Factory Farming Awareness Coalition
Adam Feltz, Michigan Technological University
Julia Hormes, State University of New York—Albany
Andrew Jalil, Occidental College
This study will examine the efficacy of three different messages to reduce consumption of animal products. We will recruit college students to watch one of four webinars, each presenting a different message. The three messages will focus on the consequences of over-consuming animal products along (1) environmental, (2) animal welfare, and (3) health dimensions. A fourth group will watch a control webinar.
Our target population is college undergraduates; two hundred undergraduates from one midwestern university and one university on the east coast will be recruited. Because our sample size is constrained by total number of participants available, we conducted a sensitivity analysis. The sensitivity analysis assumed a repeated measures design, a power of 0.8, four groups with three measurements, and correlations of 0.7 among the measures. Given these values, the f = 0.08 (d = 0.17).
A baseline questionnaire will capture demographics, relevant family characteristics (e.g., veg*n relatives, diet-related chronic illness, pet ownership), involvement in activism, and previous exposure to animal advocacy messages. We will administer the Animal Rights and Social Desirability Scales, and ratings of political orientation and religiosity/spirituality. Outcomes of interest, captured at all three timepoints, include: readiness for and intention to change, the 4Ns, and a comprehensive food frequency questionnaire. As a behavioral outcome measure, participants will be offered a Veg Starter Kit, recipes, and coupons (at baseline) and the choice of vegan or meat-based snacks (at follow-up).
While this project will help us to identify which messages are most effective at raising an intention to reduce consumption of animal products, it does not allow us to test various combinations of messages (e.g., animal welfare and environmental; animal welfare and health), nor does it allow us to distinguish between intentions and actual behavior. For future research, we would like to test the efficacy of combining the various messages among a larger sample of students, and to develop a methodology for tracking actual behavior.
How is consumer acceptance of vegan products affected when a product is labeled vegan, vegetarian, or plant-based, compared to no label?
Leah Edgerton, ProVeg International
João Graça, University Institute of Lisbon
Matthew Ruby, La Trobe University
In Germany alone, more than 6,000 products from 400 companies are labeled through ProVeg’s V-Label program. The labels currently specify “vegan,” “vegetarian,” or a specific form of vegetarian (e.g., “ovo-lacto vegetarian”). Evidence gathered to date on the program indicates that the label is viewed positively by vegans and vegetarians and suggests that the label goes unnoticed by most omnivores. Our study would seek to determine how the presence of V-Label impacts consumer acceptance. Study participants will be recruited via Facebook ad, targeting Germans who follow major German grocery store chains on Facebook. People will be incentivized to participate by being offered a chance to win a gift card to the grocery store. Participants who click on the ad will be redirected to Qualtrics, where they will participate in a survey. They will be shown nine products: five vegan target products and four non-vegan distractor products. These were chosen based on the best-selling German grocery store products. They will be randomly assigned to one of four conditions: “vegan,” “vegetarian,” “plant-based,” and no label, in order to reflect the four label conditions being tested. They will be shown the products one at a time, and asked to rate each based on taste and their interest in buying each product. All products will be from the same German grocery store brand, with the logo blurred out. Afterwards, participants will be asked to rate their perception of the brand as “ethical” and “healthy.” At the end, participants will be asked demographic questions.
Evaluating and improving conversational effectiveness
Jen Birstler, University of Wisconsin—Madison
Lucas Freitas, Google
Dhruv Khurana, Claremont Graduate University
Harish Sethu, Humane League Labs
In a variety of outreach events—such as tabling, pay-per-view, and street actions—activists engage in conversations with passersby. The goal of this two-phase research project is to improve the percentage of positive conversations at these events. In the first phase, we will identify features of conversations that are positively or negatively correlated with improving the receptivity of the activists’ messages. The results will inform the curriculum of a workshop to train activists on how best to engage with passersby. In the second phase, we will conduct a randomized controlled trial to evaluate the effectiveness of the workshop.
Audio-recorded activist-passerby conversations will be rated by untrained curators selected among omnivores, conflicted omnivores, and partial vegetarians. The rating will employ a Likert scale from 1 (strongly disagree) to 5 (strongly agree) with the statement: “the passerby is more likely to try veganism as a result of the conversation.” Features from the recordings will be extracted both computationally and manually, each validating the other for robustness. The Likert scale ratings will be used in a multi-level mixed model to ascertain the most significant features.
Variables of interest will include omnivorous status of the curator, demographics of the activist and of the passersby, conversation length, percentage of time spoken by each person, selected keywords and talking points, and the number of questions asked by the passerby.
Audio recordings of the activist-passerby conversations from Anonymous for the Voiceless outreach in Madison, WI, and additional cities will serve as the data source. We will use a sample size of 100 conversations by at least 5 different activists for a pilot study. The first-phase study sample size and the subsequent randomized controlled trial will be informed by the results of the pilot study.
The role of religious messaging and authority on members of the Catholic church
Bob Fischer, Texas State University
Sabrina Grela, Albert Schweitzer Stiftung für unsere Mitwelt
Steven Rouk, Mercy For Animals
Bianca von Wurzbach, University of Mannheim
Members of religious communities are often underrepresented in the animal rights movement, and their communities have been neglected when it comes to education and outreach. In order to help gain more knowledge about how religious individuals relate to nonhuman animals, we aim to discover the role of religious messaging and religious authority in positively influencing people’s attitudes and behaviors toward farmed animals.
We will do this by recruiting participants from the service MTurk and asking them to complete a simple survey in a 3 x 2 between subjects experimental study, with the two treatments being (1) the treatment text, and (2) who “signs” the text (i.e., the cited source of the text).
Participants will be randomly divided into six different groups. Two groups will read a standard argument for veganism, similar to those currently used by many farmed animal protection organizations. The argument will be signed by either a Catholic religious authority (e.g., “Father John Smith of such-and-such church,” for one group) or a generic authority (e.g., “Dr. John Smith,” for the second group). Two groups will read an argument for veganism that appeals to religious reasons, once again with the argument signed by either a Catholic religious authority or non-religious authority. Finally, two groups will read text unrelated to veganism, once again signed by the two different sources.
After this text treatment, we will ask participants standard questions about attitudes and behaviors toward farmed animals. We will pull these questions from ACE’s question bank.
Localizing animal advocacy messages: A test with rural identifiers
Dana Barre, Humane Advisers
Jared Piazza, Lancaster University
Michael Webermann, Better Eating International
For many people, small-town local farming is a salient part of their identity. Yet this group of people is often ignored by animal advocacy outreach efforts. The current project will contrast animal advocacy messages aimed broadly at more traditionally urban populations with those specifically tailored to individuals identifying with rural or small-town communities. Two experimental treatment messages will frame the negative consequences of industrial farming: its global impact on the environment, animals, and health, or more local impact on rural communities. A neutral, non-message control condition (about crop yields) will be included as a baseline comparison group. 335 young adults (ages 18-30) will be recruited from the University of Wisconsin to participate in a three-stage online data collection procedure which will take place (a) two weeks prior to the administration of the message treatment, (b) immediately after the administration of the message treatment, and (c) two weeks after the administration of the message treatment. After stage (a), participants will provide the following: their level of group identification with rural and urban communities, basic demographic information including age, gender, and political orientation, their level of commitment to a meat-based diet via the Meat Commitment Scale (Piazza et al., 2015), their attitudes toward factory farming (MacDonald et al., 2016), and their perception of vegetarianism as a threat to ingroup values (Dhont & Hodson, 2014). At stage (c) participants will be randomly assigned to read one of the three messages. They will then respond to attention checks, evaluate the message (or control passage), and, once again, provide their level of commitment to a meat-based diet, and provide their attitudes toward factory farming and vegetarianism.
Social contagion of food choices
Catherine Amiot, University of Quebec—Montreal
Krystal Caldwell, Mercy For Animals
Michelle Rojas-Soto, Better Eating International
Social contagion is a well-established phenomenon. We will examine whether an individual’s diet can be influenced by learning about their friends’ diet. We will recruit “influencers” on FitBit and send push notifications to their friends. We will randomly assign influencers’ friends to one of the following conditions:
- Control A: friends who don’t receive push notifications
- Control B: friends who receive push notifications about the influencer, non-diet related content, e.g. ‘Maggie was active today’
- Intervention: friends who receive push notifications about the influencer’s diet
- Intervention A: ‘Maggie had a vegan lunch. Take a look!’
- Intervention B: ‘Maggie had steak for lunch. Take a look!’
We will assess friends’ diets in the days following the push notification, and compare it to their diet before the intervention, and across interventions.
- Changes in amount of animal product consumption
- Changes in amount of plant-based product consumption
Both the pre- and post-push notification diet data is self-reported by FitBit users.
We hypothesize that: (1) social contagion will follow the same pattern observed in the running social contagion study: men influence other men, but not women; women influence men and women equally, but the effect is subtler than among men above; strict vegans’ eating habits have almost no influence on their friends, while people who are experimenting and/or transitioning to vegan eating have the most influence on friends; (2) there will be a lag in any observable contagion behavior; (3) the effect of the contagion behavior diminishes over time for almost everyone.
Optimizing municipal campaigns for nonhuman rights
Greg Boese, Animal Charity Evaluators
Matthew Dominguez, The Nonhuman Rights Project
Kelly Witwicki, Sentience Institute
The Nonhuman Rights Project (NhRP) works to secure legally recognized fundamental rights for nonhuman animals through litigation, advocacy, and education. In the next 1–2 years, the NhRP will organize the introduction of ~100 municipal ordinances in cities and counties across the U.S. which, if passed, would secure fundamental rights for certain nonhuman animals. To inform the work of the NhRP, we propose a series of survey experiments centered around three questions: (1) Which species should the NhRP use as the target of these campaigns? (2) What language should the NhRP use when defining the concept of fundamental rights? (3) Which individual characteristics should the NhRP focus to justify the case for fundamental rights?
We will commission surveys in several moderate-to-large liberal cities where these campaigns will be held. Participants will be asked to read a hypothetical news article describing a municipal campaign seeking fundamental rights for nonhuman animals. In Study 1, to understand the effect of species, we will vary whether the campaign is aimed at recognizing rights for (i) cetaceans, (ii) great apes, (iii) elephants, or (iv) companion animals. In addition to manipulating the species featured in the campaign, we will also manipulate whether the campaign is described as aiming for fundamental rights of “bodily integrity and liberty” or “freedom from harm and imprisonment.” After reading the news article, participants will be asked whether they support the campaign (strongly favor, favor, oppose, strongly oppose, undecided/don’t know), and whether they would support a law to grant farmed animals such as chicken, pigs, and cows fundamental rights. In Study 2, we will again manipulate the type of species featured while also varying whether (i) perceptions of sentience or (ii) perceptions of intelligence are emphasized to justify the case for fundamental rights. After reading the news article, participants will again be asked whether they support the campaign (strongly favor, favor, oppose, strongly oppose, undecided/don’t know), and whether they would support a law to grant farmed animals such as chicken, pigs, and cows fundamental rights.
