The negative side effects of simplifying science

Our next twitter chat will be Tuesday, December 5th at 7pm PT/8pm MT/9pm CT/10pm ET. Make sure to follow our official account, @scicomm_JC, so you can join in on the fun.

Simplifying science

This month’s chat is hosted by Dr. Mariya Voytyuk @healthy_phd and the topic is about the potential side effects of science communication.

Making science more accessible to the “lay audience” is important- after all, scientific knowledge claims play a central role in our daily lives and how we make decisions. A crucial aspect of increasing this accessibility is making scientific knowledge understandable to the general public. Popularizing science for non-scientists thus involves simplifying the complex nature of scientific information. The paper we will be discussing suggest simplification of science can have an adverse effect on people’s science-related judgments. Specifically, audiences without scientific training can become overconfident in their own epistemic capabilities when making judgments about scientific claims. In other words: people may erroneously decide that they are capable to judge the veracity, relevance, and sufficiency of science-related content without consulting experts in the field. This is termed the “easiness effect of science popularization” and it can lead to misunderstanding and misuse of scientific information.

BACKGROUND: COMMUNICATING SCIENCE TO NON-SCIENTISTS

Scientific knowledge has a high level of complexity, as new insights are often preliminary and there is much uncertainty. Such complexity makes any single individual incapable of understanding science in its entirety, but societies can manage this issue by breaking down knowledge into disciplines (represented by specialized experts). This breakdown of knowledge is called the division of cognitive labor, and it makes each of us a “layperson” with regard to most domains.

Individuals who have not undergone specialized training in any specific domain of science normally lack the deep-level background knowledge and relevant expertise that is necessary for a competent evaluation required of science-related claims. But thanks to the division of cognitive labor, people can cope with such limitations by consulting pertinent others- experts with specialized deep-level knowledge on whose evaluations they can rely.

Here comes the issue: in order for individuals to make use of the division of cognitive labor, they need to be able to realize their own epistemic limitations and recognize the need to defer to others. Unfortunately, we humans have tendencies to overlook gaps in our own understanding, and  thus easily overestimate our own knowledge.

Importantly, the authors stress that there are specific conditions that can make these overestimations more likely. In this paper, they test the effect of one such condition: the simplicity of science-related text information (the “easiness effect”).

TESTING THE EASINESS EFFECT: METHODS

The easiness effect is the simplification of complex information characteristic of popularized science depictions (e.g. online and in magazines). Prior studies have already demonstrated that the ease of processing science-related information affects people’s readiness to rely on their own judgments instead of deferring to experts when appropriate (these studies used artificial texts on fictitious issues).

The present study examined whether the easiness effect is a realistic problem: do we see it when people read real popularized articles they can easily access online?

Articles. The authors used two genres of science depictions: journalistic reports on original research addressed to expert audiences (expert articles), and those published in online tabloid magazines written for people with nonacademic backgrounds (“lay” or popularized articles). To select articles, the authors first searched 16 online journals that focused on causal statements relaed to health (e.g. “eating chilies decreases blood pressure”). They then looked for a corresponding article corroborating the same claim in the other genre. This resulted in a final sample of 16 popularized and 16 expert articles covering 16 different health topics published between 2007 and 2011.

Respondents and Procedures. The study included 73 adult German participants (50 females; Mean age ~ 34; ~40% had a university degree). To ensure “lay” status, the study recruited only adults with no specialized training in medicine, health, or pharmacy. Each participant read 2 expert audience and 2 popularized articles.

All articles were presented to participants in a printed booklet (which retained original layout but removed source information). These booklets embedded each article in a framing scenario: the participant’s fictitious friend who has a topic-related health problem is asking for advice about the correctness of a topic-related claim. Participants were then asked to imagine they had searched online to find an answer to the friend’s question and were presented with the articles in their booklet.

The booklets contained scales tracking a number of dependent measures (e.g. text credibility and comprehensibility, participants’ judgment confidence level, etc).

Main Hypothesis: Respondents will rely on their own judgment capacities more readily after reading the popularized “lay” articles (vs. expert audience articles). This would indicate the presence of the easiness effect for popularized articles.

To assess how much participants trusted in their own judgment (judgment confidence), participants had to rate how far they would follow 3 different judgment strategies both before and after reading each article (each on a 7-pt scale):

  • Strategy 1: based on my present knowledge about (inserted) topic, I am confident in concluding whether (claim inserted) is correct.
  • Strategy 2: I don’t believe my current knowledge suffices to conclude whether (claim inserted) is correct, but I will be confident about answering this question after obtaining further information on the topic.
  • Strategy 3: I don’t believe that my current knowledge suffices to conclude whether it is correct that (claim inserted). I also would not be confident about answering this question after obtaining further information about (topic). I would prefer to consult a trustworthy person who is more knowledgeable about the topic, and rely on their judgment as to whether it is correct that (claim inserted).

To assess the impact of text genre (expert vs. popularized), researchers compared pre- and post-reading measures of strategy preference from both genres separately for each of the 3 judgment strategies. According to their hypothesis, reading the popularized article would produce higher ratings on strategy 1 and 2, but lower on strategy 3.

RESULTS: THE EASINESS EFFECT IS A REAL CONCERN

As expected, respondents were more confident about their claim judgments after reading the popularized depictions: they had higher trust in own judgment based on current knowledge (strategy 1) and a weaker desire for advice from a more knowledgeable source (strategy 3). Thus, authors conclude that people’s overly ready reliance on their own ability to evaluate complex science is a real-life concern.

In addition, respondents rated popularized pieces as more comprehensible (easier to process). This supports the notion that difference in judgment confidence between genres is indeed related to differences in processing ease.

Antidote. One suggested antidote to the “easiness effect” with popularized science communication is to clearly point to the controversial and complex nature of scientific topics. In fact, there is already evidence that denoting an issue as controversial or highly complex can significantly reduce the strength of the easiness effect (though it does not delete it completely).

Interestingly, the authors expected that respondents would perceive the topics covered in popularized articles to be less complex and controversial than in expert articles (as other studies have shown), which was not supported by the results. This lack of difference in perceived complexity between genres was strange- after all, popularized reports make issues appear less complex by presenting uncertainties as facts or omitting details of relevant studies. In explaining this, the authors suspect that the types of articles they chose to represent expert reports offer an explanation: they were already simplified versions of original research, while still targeting expert audiences.

In conclusion, what is the take-home message for all of us SciComm enthusiasts? As we improve and refine our SciComm efforts, it is important to acknowledge the possible side effects of communicating the complexity of science to nonscientists. According to the authors, the goal of improved SciComm efforts would be to explicitly highlight the complexity and controversiality of the the topic while informing the audience in a comprehensible manner, yet keeping the risk of the easiness effect low.

Looks like it’s time for another study to test whether this suggestion can indeed minimize the easiness effect for a range of science topics.

How do you currently simplify your science for a non-expert audience? Will you revise the way you do science communication after reading this summary to guard against the easiness effect?  Come to #scicommjc Twitter chat and lets discuss this topic.

You cannot make it to the live chat?

If you are interested in an updated version of this summary with highlights from the chat, and summaries from our future chats, please subscribe to our blog to receive them in your inbox.

Subscribe to our blog

Leave a Comment