Scientific Activism

For most of its history, Western medicine was a bit of a crapshoot.¹ With treatments involving leeches, bloodletting, and mercury (among other things), going to the doctor was as likely to harm you as your actual illness. Doctors had a tendency of taking the patients who recovered as proof of the treatments’ success and patients who didn’t as inevitable losses to the original disease. A doctor could have their pet theories and no system would hold them accountable for the validity of those theories. It wasn’t until we started bringing in the tools of science and systematic study that modern medicine really began to take off and start doing more good than harm.

Animal activists aren’t running around giving people mercury (thank goodness), but we are guilty of the same kind of motivated reasoning. We will leaflet or protest, hear people say they’ll change, and assert that our work is done. When we win, it’s an indicator that we did the right thing. When we lose, well, the system is just too entrenched for us to have done anything. While this doesn’t always happen, there’s a dangerous lack of accountability to results in activism. Incorporating scientific principles and practices into our activism is a big way of closing that accountability gap.

The challenge is that few of us are actually familiar with these principles. The average American’s exposure to doing science is a class in high school involving black composition books and prewritten experiments. For folks who did science in college, we may have done some research that involved writing incredibly dry papers which we submitted to a journal and never thought about after the fact.² These experiences portray science as formal, clinical, boring, and worst of all, predictable. In truth, they’re watered-down versions of the real thing.

If you strip away the trappings of science (the lab coats, the papers, the equipment, the degrees), what you’re left with is the willingness and ability to test our stories about the world. Humans are born storytellers - we’ve been telling stories for as long as we’ve been human. Some would even argue it’s what made us human. What we have not been doing is rigorously testing those stories to see how they hold up.³ That’s what science is about - wading into the uncertainty of the world and testing to see which of our ideas survive under pressure.

The Hypothesis: The Beating Heart of Science

The heart of science is making and testing predictions. If I do X, then Y will happen.⁴ Making predictions is easy. We make predictions all the time: “People like happy hours and they’ll be more likely to join our activist efforts if they go to a fun event first”; “If we tell people about the horrors of factory farming, they will stop eating meat”; or “if we send a bunch of emails to a CEO, they’ll get really annoyed and change their policy”. Formally, these predictions become our hypotheses. Once we have a hypothesis, we have to figure out how to test it.

The most important feature of a hypothesis is that you have to be able to prove it wrong.⁵ If we can’t demonstrate that the prediction failed, then it’s all too easy to keep on believing what we already want to believe. Then we have faith, not science. In order to be able to fail a hypothesis, we have to be able to observe the results of our experiment. For example, if you’re doing vegan street outreach, it would be a bad prediction to say that lots of people go vegan after talking to a street activist. What does “lots” mean? How would we know that they actually followed through? It would be a much better prediction to say that 10% of people put down their email or that 10% of people say they’re eating vegan when we call them six months later. These claims are more specific, which makes it possible to definitively say whether they happened.

Making specific, testable hypotheses benefits from a little quantification. It’s nice to be able to count something: donations received, event attendance, postcards written, etc. Counting removes a lot of subjectivity - however you feel about your street outreach conversations, it’s hard to argue with emails registered or signatures on a petition. Quantification can also help us come up with hypotheses. You can always ask what you want to see more of and then look for a way to count that. An easy structure for a lot of hypotheses is “if I do X, I will get more of Y”.

If you’re going for something a little more intangible, there’s room for surveys or interviews. Surveys can be on a scale (“On a scale of 1 to 10, how much do you like vegan food?”) or use freeform answers (“What are your associations with veganism?”). An easy way to fold surveys into an experiment is by giving people a before and after survey, ex. asking how good they think vegan food is, giving them a sample to taste, then asking again. Surveys are a great option if we’re trying to sway people’s opinions or improve people’s experience.

Thinking in terms of counting emails or writing surveys can be counterintuitive. Hypotheses don’t follow the dramatic arcs that our story-loving brains crave; instead, they focus on picking over details for exactly what happened (or didn’t). But that’s where the magic comes from. By forcing us into the weeds and holding us accountable to observing the results of our actions, science enables us to build an understanding of what works and what doesn’t. This understanding is what gives us the power to reshape the world and it all starts with a well-made hypothesis.

Campaigns are Experiments

Once you have a hypothesis, the rest of the scientific process can fit seamlessly with activism we already do. We don’t have to go out of our way to run experiments; our events and campaigns are our experiments. Every event we host or campaign we run can be built to test a specific hypothesis about what works. The biggest change is being consistent about recording the results of our events and campaigns. You may already be doing this - we collect sign up lists, count postcards, and take attendance all the time. If you’re not, keeping records just takes a few minutes of time during or after the event.

Analyzing the results of our experiments meshes with the reflection we already (hopefully) do about how our events and campaigns went. Most activists I know spend plenty of time thinking about what they’re doing and if it works. With a hypothesis in hand and good records, we’ll have the data to discover what works. Picking up some stats can be useful if you want to formally analyze the results, but there’s also room to eyeball things.⁶ We’re not trying to get published in Nature; we’re looking for interventions that can make a big difference.

After we have our results, we share them. We discuss activism constantly. Having some actual data to make a point is a great way to spark discussion. Maybe you’ll get some valuable feedback, ideas on what to test next time, or inspire someone to try your experiment for themselves. Science advances on discussion, on people looking at each others’ experiments; proposing and debating theories; and then going out and doing it all again. Science, like activism, is a collaborative project that needs all of us.

Science in the Wild

As a local organizer for The Humane League (THL), I’ve spent a lot of time worrying about whether my work made a difference. My main responsibility was to host events that supported THL’s national cage free campaigns and drive activist turnout. Given the tiny size of our movement, growing our community was one of my chief preoccupations. As a math major, I promptly started recording all of the data I could get my hands on. Eventually, I realized I could run experiments and use the data to see how they affected turnout.

I tried a lot of things to get people to attend digital action meetups and protests over the years. I posted on Facebook, Meetup, and other social media. I tabled at farmer’s markets and Veg Fests. I sent personalized texts. I went to socials for unrelated groups and talked to people there about THL. I put up posters around local coffeeshops. When the meetups and protests rolled around, I recorded everyone who showed up. That gave me the data I needed to know who came when and as a result of which experiments.

Take a moment. Which of these experiments do you think would work? Why? I’m sure you can come up with all kinds of reasons for why each idea would either succeed or fail.

As it turns out, most of them accomplished very little. Social media posts and posters resulted in new volunteers once every couple months. Tabling and going to social events were more effective, yielding a new volunteer per two to four hours of effort. By far the most consistent way to get people to show up was the personalized texts - attendance fell from 6-8 people when I texted people to 1-2 people when I didn’t.⁷ These results are not what I would have expected; I would have expected tabling to be a lot more effective and personalized texts to not matter so much. But they’re what I got and what I worked with.

From these results, I’ve come to the conclusion that there’s no substitute for a human touch when it comes to event invitations and the most reliable way to build a community is a lot of legwork by someone more social than me, putting in the raw hours to network and table. Am I 100% confident in these conclusions? Of course not. Maybe I’m just bad at writing social media advertisements and that’s why no one showed up. Maybe Seattleites crave personal communications more than other folks, which is why the texts worked so well. We won’t be able to tell without more experiments and more data. This is why we need a range of activists running experiments and seeing what works across different contexts, so we can find the more general principles behind good organizing.

Three Things to Try

1. Make a Prediction. Next time you attend or host an event, write down something you’re expecting will happen. Be specific - not “it’ll go well,” but “we’ll get ten email sign-ups” or “three people will say they’re interested in coming back.” Then check afterwards.

2. Find Something to Count. Pick something you’ve been doing regularly and look for something you can count to assess success, like email sign ups or people coming back. A notes app on your phone is all you need to get started.

3. Try a New Approach. When you attend or host an event, change one thing - a new venue, a different opening line, a shorter ask - and see how it affects the results you’ve started tracking. Running it a few times will tell you more than a single attempt.

Thanks to Beta Readers: Kim Evans, Vishnu Amrit, and Jacob Waters

1. There have been numerous indigenous or traditional medicinal practices that were surprisingly effective for the time period and used local plants and animals. But those often evolved over thousands of years of trial and error.

2. While not an exact measure of who’s doing science, about 2% of the American population hold PhDs. Or in other words, there’s about as many people actually doing new science out there as there are vegans. That is to say, not a lot.

3. Of course, we still learn through trial and error but that is a slow and error-prone process.

4. Formally, X would be the independent variable (the thing we change) and Y would be the dependent variable (the thing we expect to change with X).

5. Technically, we don’t prove or disprove anything, we just fail to find evidence that what’s happening is the result of anything other than random chance.

6. You can also ask people for help with proper statistical analysis if you get the data - there’s plenty of activists out there with enough science background who would love nothing more than to get their hands on some data (me included).

7. I decided not to put in the spreadsheet data I’m drawing these conclusions from, but would be happy to do a more detailed breakdown of the data if enough folks are interested.