Trust in science and scientists and the acceptance of evolution

If you do not trust the people doing science, you discount what they find. When you discount their findings, you reject their conclusions — even when those conclusions are among the most well-supported in all of biology. This is not a philosophical argument. It's a measurable chain. Nadelson and Hardy went out and measured it. Their question was straightforward: among ordinary undergraduates — not biology majors and not people who attended a science lecture because they already cared — what actually predicts whether someone accepts biological evolution? The answer, it turns out, has three moving parts: trust in science and scientists, religious commitment, and political orientation. How these three factors combine tells us something important not just about evolution, but about how scientific knowledge travels, or does not travel, from researchers to the public. The sample consisted of one hundred fifty-nine students enrolled in an introductory psychology course at a large university in the western United States. The average age was just under twenty. Only about seven percent were biology majors, and students reported taking fewer than two college science courses on average. This group represents the general public, more or less. A large portion of that public, as the authors note bluntly, does not currently accept biological evolution. To measure trust, Nadelson and Hardy used the Trust in Science and Scientists Inventory — the TSIS — a twenty-one item scale that captures whether someone believes scientists are competent, honest, and that the scientific process produces reliable knowledge. Items range from "I trust scientists can find solutions to our major technological problems" to reverse-phrased statements like "We cannot trust scientists because they are biased." Internal consistency was high, with a Cronbach's alpha of 0.85, and the sample's average trust score was 3.35 on a five-point scale — slightly above the midpoint, but not by much. Acceptance of evolution was measured with the Inventory of Student Evolution Acceptance, or I-SEA — a twenty-four item scale developed by Nadelson and Southerland. What makes this instrument methodologically important is that it does not treat evolution as one monolithic question. It breaks acceptance into three separate subscales: microevolution, which covers small-scale changes within a species; macroevolution, which covers large-scale changes that produce new species; and human evolution specifically. Each subscale has eight items. The overall alpha was 0.95. That three-way split turns out to be enormously significant. Microevolution acceptance averaged 3.92 out of five. Macroevolution dropped to 3.57. Human evolution came in at 3.33. Even within this single sample, acceptance was not uniform — it declined as the stakes, from a worldview perspective, got higher. Religious commitment and political orientation were each captured on single ten-point scales, and the sample landed near the middle of both — slightly more liberal than conservative, with moderate average religious commitment. The researchers then ran correlations, regressions, and a structural equation model to determine not just whether these factors move together, but how they flow into each other. The core findings are consistent and clear. Trust in science was positively associated with evolution acceptance across all three contexts. In a single predictor regression, trust explained thirty-three percent of the variance in overall acceptance, with a standardized coefficient of 0.57. This held up across subscales: twenty-eight percent of variance for microevolution, twenty-six percent for macroevolution, and twenty-five percent for human evolution. Religious commitment ran in the opposite direction, and its effect grew as the context became more personal. In single predictor models, religiosity had a coefficient of negative 0.42 for microevolution, negative 0.55 for macroevolution, and negative 0.60 for human evolution — explaining eighteen, thirty-one, and thirty-five percent of variance respectively. The more the question touched on human origins, the more tightly religious commitment was tied to rejection. Political conservatism also predicted lower acceptance, but more weakly. Single predictor coefficients ranged from negative 0.28 to negative 0.33 across the three contexts, explaining roughly eight to eleven percent of variance — real, but smaller than either trust or religiosity. What happens when you put all three together is where the study gets genuinely interesting. In the full structural equation model, known as SEM, a technique that allows researchers to test whether one variable influences another through a third rather than just showing that two things correlate, the combined picture explained forty-one percent of the variance in evolution acceptance. Trust correlated at 0.43, religiosity at negative 0.45, and political conservatism at a smaller negative 0.12. However, political orientation's direct path to evolution acceptance was no longer statistically significant once trust and religiosity were included in the model. Its influence, such as it is, runs largely through those other two variables rather than acting as an independent driver. The authors acknowledge the fit indices for the SEM were mixed — the chi-square to degrees-of-freedom ratio was 14.87, which they flag as problematic, likely a function of sample size — so the path picture is suggestive rather than definitive. Nevertheless, the pattern is consistent with the regression results. The two-predictor models help clarify the shape of this. When trust in science and religious commitment were entered together for human evolution, the combined model explained forty-four percent of variance — with religiosity at negative 0.48 and trust at positive 0.31, both statistically significant. For microevolution, the same combination explained thirty-three percent, with trust actually being the stronger predictor at 0.43 and religiosity at negative 0.26. Thus, the relative weight of religiosity versus trust flips depending on which kind of evolution you are asking about. For microevolution, trust is the bigger lever, while for human evolution, religious commitment takes over. This context-dependence is one of the study's most important contributions. "Do you accept evolution?" is not one question; it is at least three, each with different psychological profiles. Human evolution is the least accepted, the most contested, and the context most tightly bound to religious identity. Microevolution is the most accepted and the context where trust in science does the most explanatory work. Collapsing these into a single question, as much prior research has done, masks variation that matters both scientifically and practically. There are real limits here that Nadelson and Hardy are careful to acknowledge. This is a correlational study conducted at a single institution, with a sample of one hundred fifty-nine undergraduates who were mostly European American and mostly in their first year. Causal claims are not warranted. The single item measures of religious commitment and political orientation, while validated in prior work, are rougher than multi-item scales. Multicollinearity among the predictors — the fact that religiosity, political conservatism, and lower trust in science tend to cluster in the same individuals — makes it difficult to cleanly separate their independent contributions. However, the practical implication they draw is worth pondering. Increasing students' trust in science may be a more tractable intervention point than changing religious belief or political identity. You are unlikely to change what someone believes about God or how they vote by teaching them about evolution. But trust in science is a learned orientation — meaning it might be teachable. If trust mediates the relationship between worldview and scientific acceptance, then strategies that build trust — that demonstrate how science works, who scientists are, and why the process is reliable — could shift acceptance without requiring anyone to abandon their identity. The authors also suggest that the same dynamics visible here extend far beyond evolution. The mechanisms linking low trust to rejection of scientific consensus — the tendency to discount evidence, to favor alternative explanations, and to treat scientific expertise as just one opinion among many — are not specific to Darwin. They appear in debates about climate change, vaccines, and genetically modified foods. While evolution is the test case this study used, the trust problem it reveals is much broader. What Nadelson and Hardy have built is a quantitative picture of a chain that many people have described qualitatively. Trust shapes what evidence you are willing to accept. The evidence you accept shapes the conclusions you reach. And the context in which you are asked — whether regarding small changes in bacteria or the origin of your own species — shapes how hard that chain is to break. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.