10 Tips for Good Science Communication in Beauty.
By The Eco Well SciComm Advisory. Contributors: Jen Novakovich, Annika Lagut PhD, Lyle Burgoon PhD ATS and Todd Squires PhD.
Science Communication is incredibly important. But doing it well is hard! And doing it poorly? That comes w/ very real risks (e.g. unintentionally creating more misinformation, eroding public trust, and sometimes doing real public harm). Here are 10 tips to help you do *good* SciComm. These tips were tailored for the cosmetics industry, but are widely relevant to anyone thinking about doing SciComm!
ICYMI: There is currently a lack of infrastructure to support good science communication, including via accessible training opportunities. This is a really big problem considering a) how important SciComm is, and b) the risks for SciComm done poorly. Our SciComm Mentorship Cohort Program is an effort to address this problem in beauty. These tips were put together for our Cohort training materials. Slightly adjusted, we hope they can help the greater cosmetics world and beyond develop their SciComm skills.
Here are our 10 tips!
1) Show, don’t tell.
People are prone to misinformation for a number of reasons: often fear, uncertainty and our want for simple explanations in a very complicated world play a big role. Good SciComm is about meeting this need, using compelling enough communication to replace the wrong information with truth. Generally, you’re unlikely to be successful in scicomm when you leave it at e.g. “parabens are safe!” You need to demonstrate to them why what they may have previously believed isn’t quite right. And in a way that doesn’t make them immediately tune out - because we as humans can get emotionally invested in some of our opinions. This is hard and takes a lot of practice to do well!
2) Be aware of how people generally form ideas (even scientists!).
This goes hand-in-hand with our first tip. The ideas we form are based on our experiences (Experience -> causal narrative -> prior belief -> selective evidence evaluation (confirmation bias) -> reinforced belief); this pathway is deeply ingrained in human nature. If beliefs are based on e.g. pseudoscience and confirmation bias, and have ended up deeply ingrained in someone’s personhood, you’re probably not going to break through by saying “your experience is wrong.” Instead, you need to break that causal chain, with a combo of validating their experience and showing them an alternative, compelling POV. See the end of this blog entry for an example of this pathway in action.
3) Approach with empathy and curiosity.
If we attack people and their experiences we will likely lose them, and may push them deeper into accepting the misinformation. “Attack ideas” is a nice debate phrase, but people tend to hold their ideas as a part of their identity. The goal of SciComm is to present the research and answer questions, not to force someone to change their mind. It is important that people feel in control and that no one is telling them what to believe
4) Speak carefully and mindfully.
Every topic has different information needs. Too simplified threatens to create unintended misinformation. Too much information threatens to lose your reader’s attention. With the right language, you can create content that is understandable to a lay audience yet interesting to an “expert” audience. You may feel pressure to “sound smart” by using more technical words, but this will translate into limiting your audience to those that are already embedded in a field. Nobody will ever complain if something was too easy to understand or follow. Speaking carefully and mindfully is about trying to convey exactly what you want to convey, but in the simplest and clearest way possible. Make it as easy as possible for your audience to understand your point, to follow your logic, and to understand your conclusions and the rationale that supports them. The role of a SciCommer is to predict this for your specific audience. Each audience will have slightly different information needs. E.g. https://www.instagram.com/reel/DcjXi0vvOUY/
5) Everyone can think like a scientist, but most people (including scientists) don’t.
We need to encourage people to ask questions about their beliefs and the beliefs of others. What sets scientific thinking apart from other ways of thinking is that scientists are supposed to change their beliefs based on the data. Our beliefs are fluid and ever changing, and we grade the evidence. It’s perfectly fine to say, “I believe X based on the current evidence; however, the evidence is not very strong.” This next part is hard: we should tell people what it would take to change our minds. What does that look like? We don’t often do this, but we should make it a habit. It shows intellectual honesty. But if you do that, you best be prepared to change your mind or have a good reason why you didn’t.
6) Know when to hedge your points.
In science, there is generally a lot of uncertainty for topics. When you come on too strong about a point, you may end up misrepresenting the reality. E.g. “sunscreen does not harm coral” - this simple point can very easily have holes poked into it. Studies (that are questionable) have shown sunscreen can harm coral in high exposure scenarios (not relevant to nature) (e.g. NASEM, 2022). Furthermore, science evolves, and research is underway looking at this topic. A more accurate statement would therefore be something like “sunscreen has not been shown to harm coral reefs as they appear in nature.” That is a very different statement than where we started. Treat uncertainty as a benefit, not a hindrance, it tells us what still needs to be studied.
7) Transparently disclose where you’re getting your information from.
Whether it's from research papers, books or other science communicators you’ve learned from, if you’re adding points based on info you’ve learned from them. For the latter, this is a really good practice to get into so as to support the general scicomm ecosystem, rather than creating drain on your fellow scicommers - plagiarism is a big and underappreciated issue in this space. If you’re citing studies, make sure you’re citing the right studies, and make sure you’re accurately representing them. Avoid just going out to selectively support a view point; this kind of cherry picking may point your content more towards pseudoscience, which you don’t want!
8) Don’t use an AI chatbot to do the thinking for you.
Chatbots tend to tell you what you want to hear, exploiting confirmation bias, often translating to confidently wrong POVs, especially for topics like science. Chatbot info sources tend to be what’s more widely accessible; there’s a lot of misinformation out there, & sometimes the better science is less available due to things like paywalls - big and complicated problems. Furthermore, because of the way chatbots work, they tend to be plagiarism machines, trained from e.g. the content from your fellow scicommers. See last point for more on the problem with that. Finally, chatbot content tends to be unnecessarily wordy, often doesn’t get it right - even when they’re creating summaries of transcripts. Often they end up making it take way longer than it needs to take to create good scicomm content. All of these reasons are why the use of LLMs to create content is not allowed in our SciComm cohorts.
9) Most people are too busy to dive deeply into the science. That’s why SciCommers exist! But as communicators we need to be honest brokers of information.
We are not advocates. We shouldn’t have a vested interest in what the data says, all we should care about is the quality of the research and accurate communication of results. We need to be careful not to hold a belief just because this is what the data used to say. If there are a bunch of studies that show X is toxic, and others showing X is not toxic, then we need to buckle down and figure out why there’s a difference here. Typically, this is a situation where the studies are of varying quality, or they were performed in different species, in different cells or tissues, or maybe different routes of exposure. This is where finding a relevant scientist (highly trained toxicologist in this case) can help you understand what’s going on. Noone knows everything, and this stuff is complicated! Intellectual humility is a super important trait for good SciComm.
10) Get to your points and move on, and organize your points in a way that makes logical sense for building an idea or story.
Unnecessary repetition will very quickly lose the attention of your audience. Not giving the right groundwork for a topic with your points may make it harder for your audience to take away the points you want them to take away. This takes strategy!
Example of belief formation from point 2!
In wilderness survival training, we teach warfighters how to identify poisonous plants and fruits. But we also teach them how to test for poisons. When people experience the burning lips sensation for themselves, that experience is a far better teacher than me holding a branch of berries and saying, “Don’t eat these, they’re poisonous.”
Now that they have that experience, they build the causal narrative: “those tiny, red berries burn really bad!” Now, in many of these warfighters (especially if they lack a botany background), they immediately build a prior belief that all tiny, red berries burn bad and are poisonous. Some of them will decide to experiment again a few days later. If the berries burn again, they’ve reinforced their belief. Those don’t have the confirmation bias experience, they just have a confirming experience.
Now, if some of those warfighters searched “tiny, red berries poisonous North Dakota” on the internet, they might find websites that just talk about tiny, red poisonous berries from North Dakota. That could become the confirmation bias that then reinforces their beliefs. And now if they go out in the wilderness and someone else says, “just avoid red berries, they’re poisonous out here” that’s further reinforcing that belief.
If that belief is based on pseudoscience (i.e., there are tiny, red, non-poisonous wild berries out there) and I want to correct that, I can’t just say, “Your experience is wrong”. They know what they felt, and me attacking their experience is attacking them. Instead, I need to break the causal chain. I need to help them go from “All tiny, red berries are poisonous” to “Some tiny, red berries are poisonous, and others are not.”
he way I do that is by validating their experience, “Yes, you’re right, there are many tiny, red poisonous berries out here.” And then I need to gently break the causal chain, “Let me show you some tiny, red berries out here that are not poisonous.”