Summary of Thesis
My thesis explored the influence of animal advocacy movies on behavior (i.e. change in meat and dairy products consumption), attitudes (i.e. carnistic defense) and emotions. For this purpose, I conducted a quasi-experiment with two experimental groups (a fictional movie and a documentary), and a control group. The data on animal products consumption were collected prior to and one month after watching the movie/documentary and the data on attitudes and emotions were collected prior to, right after the movie and in one month. This design should resolve some of the uncertainties of previous cross-sectional studies by producing more reliable data enabling testing causality. I have found no effect of watching either a fictional movie or a documentary on any outcome (animal products consumption or carnistic defence) in one month’s time after watching the movie.
Why is this important
Meat production is increasing rapidly every year. Today's meat production is almost five times higher than in 1961. Currently, over 70 billion animals are slaughtered and eaten every year, which causes a lot of animal suffering, contributes to global greenhouse emissions and increases the risk of cardiovascular diseases in the population. For further discussion please continue to this post or my thesis.
One strategy to alleviate the above mentioned problems is decreasing individual animal products consumption and changing people's attitudes towards animals.. This advocacy strategy is used by many animal advocacy charities and includes interventions like leafleting and online ads (Animal Charity Evaluators, 2018).
Documentaries are one of the most commonly used interventions in the animal rights/welfare movement. There are many ‘must-see lists’ with animal rights documentaries and movies on personal blogs but also animal advocacy organizations’ websites. For example, PETA published a list of Top Animal Right Movies to Show in Class which should change students’ thinking and motivate them to take action for animals.
Although there is some cross-sectional evidence that many vegans and vegetarians say they reduced their meat and dairy products consumption and became veg*ns after watching a movie (Humane League Labs, 2014), there are no experimental studies supporting, describing or explaining the process of this behavioral change. Without such studies, we can not be sure this intervention works as suggested and hundreds of thousands of dollars spent on animal advocacy documentaries might thus not bring about the desired societal change.
Researching effects of advocacy movies is also one of the most recommended research projects by Animal Charity Evaluators.
Strengths and weaknesses
The main strength of this thesis is its experimental design which allows testing for causality. It also provides long-term data, by tracking changes in behaviors and attitudes one month after watching the movie. The results provide new information about predictors of change in meat and dairy products consumption and about the process of this change, suggesting that movies and documentaries seem less effective intervention. There are also some benefits stemming from testing this hypothesis on a sample from the Czech Republic, as to allow for that, I had to translate The Speciesism Scale and The Carnism Inventory to Czech language and partially test its validity in Czech context. This will allow researchers to use these scales on Czech samples in the future and will contribute to research on how these concepts are perceived in various cultures and countries.
The largest weakness of this thesis is its inability to detect smaller effect sizes due to modest sample size (N=99) This study was powered to detect 27% and higher difference between the groups (d=0,38) in their meat consumption. I have found no differences between groups, suggesting that the effect of movies/documentaries is likely not that high. However, even a smaller effect than this (e.g. 15 %) might still make this intervention worthwhile. Future researchers should thus aim to collect large samples, enabling them to detect even smaller effects.
Another problem might be the use of self-reported measure of meat consumption. Although this measure seems to be one of the best of self-reports and was used in other influential studies (e.g. Humane League Labs, 2014), using more objective measure of meat consumption (e.g. tracking amount of animal products on the receipts, measuring what customers purchase using for example customer loyalty card data) provide more reliable results.
Finally, another important factor is the choice of movies. I have used only two kinds of movies (documentary and fictional), both with minimal violence displayed and mentioning only ethical arguments to reduce meat consumption. I can not rule out those same kinds of movies with more drastic/violent scenes would have stronger effects on reduction in animal products consumption or attitudes towards animals. Similarly, if those same kinds of movies used arguments other than (or in addition to) ethical considerations of animal welfare, the effects of such movies could also be different. The optimal approach in movie choice would probably be choosing a type of movie which is most popular (according to the most recent analyses by Kieran Grieg the most viewed documentaries in US in 2019 were Forks Over Knives, What the Health, Cowspiracy and The Ghosts in Our Machine) and addressing the ethical or health arguments for meat reduction (according to this Faunalytics summary majority of current vegans and vegetarians are motivated by ethical or health arguments).
In which ways I have changed my mind since I finished writing my thesis
I started to believe that changing eating habits is a more complex issue than I originally thought. It seems to me that no single factor (e.g. watching a movie) could make the change alone - there have to be more factors involved to allow for a change of eating habits to happen. This kind of thinking is well described by Theory of Planned Behavior, in which the final behavior only appears after factors like personal attitudes, social norms, perceived behavior control, emotions, personal problems and problem awareness work in conjunction. Another potentially important factor which isn’t included in this theory is the price and taste of plant-based substitutes.
Overall I believe that research exploring the effect of movies on meat consumption is difficult to conduct (requires big sample size, it is financially and technically challenging) and it is sensitive to many variables (e.g. the choice of the movie). Also, watching a movie is highly dependent on people's will to do so (in contrast to, for example, public ads). Further, I think that watching a movie can motivate mostly people already reducing their meat consumption to continue doing so or maybe it can motivate people who are already thinking about reducing their meat consumption to give it a try.
Recommendation based on my experience
Doing this research for my bachelor thesis was quite a challenge from its very beginning but it taught me some lessons which might be helpful to other researchers thinking about using similar research design.
To all future researchers I would recommend following:
During choosing a movie, think carefully about its messaging, popularity, availability of subtitles and possibility to stream the movie publicly (for example, Netflix doesn’t allow it). I would also recommend searching for impact analyses which are sometimes available (for example The Ghost in Our Machine Impact Report)
Using objective methods is always better than using self-report methods when measuring meat consumption. Even though objective methods might be more complicated to carry out, they provide more reliable results. What I consider as the worst choice is the self-identification questionnaire (“Which of the following best describes your diet: vegetarian, flexitarian, omnivore,...”). For more information about measuring the diet and the consumption, please follow the Faunalytics or The Humane League Labs.
One of the biggest problems with conducting RCT design of this kind is getting enough participants - especially when you ask for about 3 hours of their time. But it is especially important for reliability of the results to get big enough sample size. From many ways of attracting and motivating people to take part in my study (sharing the screening invitation as a facebook event and offering free entry and refreshments), one thing which worked the best was cooperation with local movie clubs (situated at universities or cities). Those clubs usually have some network of fans, they use their own equipment and space so they can help you a lot with attracting participants and solving the practical challenges (e.g. where and how to do the screenings). Plus they are really nice because they are (usually) very interested in research related to movies. Potentially beneficial can also be cooperation with local animal advocacy groups, which are already organizing public movie screenings.
It is very important to get a big control group (bigger than experimental groups), if you conduct an experiment with time lagged measures. People in the control group have a lower motivation to complete the whole experiment (for example answer the last questionnaire in a one months’ time) than people from experimental groups so you can expect bigger drop out in the control group.
The most important experience was that doing interesting and altruistically-motivated research really pays off. I didn’t receive any grant and I invested a lot of my time and my own money to conduct this research. But I received a lot of support from many people who were really excited about the topic of this study. All professional translators and all movie clubs helped me for free and one anonymous donor even offered me financial help because he was excited about this study.
In case you are interested, all data is free to download at the Open Science Framework.
If you have any questions about this thesis or if you are planning similar research project and struggling in any phase, please, feel free to reach out to me on - [email protected] I would be happy to help.