Teleological reasoning, not acceptance of evolution, impacts students’ ability to learn natural selection

For decades, biology educators have operated on a working assumption: that students who reject evolution, especially on religious grounds, simply cannot learn it. Acceptance, the thinking goes, is the gate you have to open before understanding can walk through. Change what students believe, and the learning will follow. A study by Barnes, Evans, Hazel, Brownell, and Nesse found the opposite. Over a full semester, students' acceptance of evolution had zero predictive power over whether they actually learned natural selection. The thing that mattered was something else entirely — something cognitive, not cultural — and most teachers aren't measuring it at all. The debate in biology education has long been organized around two competing explanations for why students struggle with evolution. On the cultural and attitudinal side, researchers point to religiosity, parental attitudes, ingroup norms, and prior exposure to negative media about evolution. The logic is that low acceptance of evolution creates resistance, and resistance blocks learning. On the cognitive side, researchers point to intuitive reasoning habits — the kind of mental shortcuts that feel natural but lead students away from mechanistic, scientific explanations. Barnes and colleagues set out to test which of these forces actually predicts who learns natural selection over a semester. To do that, they followed students through a semester-long undergraduate course in evolutionary medicine at a selective public research university. One hundred and ninety students completed a pre-test on the first day of class, and one hundred and twenty-nine provided complete data at both time points. The course was intensive — lectures, small-group discussions with graduate assistants, lab exercises, a field trip, quizzes, a midterm, a final, and a term paper applying evolutionary thinking to a human disease. On the first day, students completed a battery of measures. To assess understanding of natural selection, the team used the Conceptual Inventory of Natural Selection, or CINS — a twenty-item multiple-choice instrument scored out of twenty. To measure acceptance of evolution, they used three Likert items asking whether evolutionary principles explain the origins of viruses, reptiles, and humans. They also collected religiosity scores, parental attitude toward evolution, and a measure of teleological reasoning. That last one is worth pausing on. Teleological reasoning is the cognitive habit of explaining biological traits as if they exist for a purpose — saying "giraffes grew longer necks so they could reach leaves" rather than explaining that variation in neck length existed, and individuals with longer necks survived and reproduced at higher rates. Barnes and colleagues assessed this using three statements tapping goal-directed or need-based thinking about evolution, combined into a single mean score on a one to five scale. Here is what the pre-test data showed. Incoming acceptance of evolution was strongly shaped by cultural variables. When the team ran a regression to predict acceptance, the model explained fifty percent of the variance. Parental attitude toward evolution was the single largest predictor — students whose parents were more accepting of evolution came in with higher acceptance themselves, with a standardized beta of point fifty-two. Student religiosity was a significant negative predictor, with a beta of negative point twenty-seven. That model makes intuitive sense. What you believe about evolution tracks closely with your family and your faith community. But then the team ran the parallel analysis — predicting post-test CINS scores while controlling for pre-test CINS, which isolates actual learning gains over the semester. The cultural variables dropped out entirely. Parental attitude, religiosity, and acceptance of evolution all failed to make significant unique contributions to how much students learned. The model still explained fifty-three percent of the variance in post-test scores, but only two predictors did meaningful work: pre-test natural selection understanding and teleological reasoning at the end of the course. Let that sit for a moment. The factors that powerfully predicted what students believed about evolution had no measurable relationship to how much they learned about it. Teleological reasoning was the cognitive bottleneck. Students who began the course with lower levels of teleological thinking showed the largest learning gains in natural selection by semester's end. And students who still endorsed teleological explanations at the post-test performed worse on the CINS than peers who had reduced those tendencies. The post-test teleological reasoning score carried a standardized beta of negative point twenty-four in the model predicting CINS — a meaningful, significant effect. Barnes and colleagues are careful to distinguish teleological reasoning from religious belief or intentional creationist thinking. Teleology is not the same as believing God designed the giraffe's neck. It is a domain-general cognitive bias that emerges early in childhood and persists into adulthood — a tendency to see purpose and function in natural processes even when you intellectually accept that no purpose is there. That is precisely what makes it so hard to teach around. A student can fully accept that evolution happened and still explain every adaptation in teleological terms. The good news is that the course moved both needles. Natural selection understanding increased significantly over the semester, from a mean pre-test CINS score of about fourteen out of twenty to a mean post-test score of about sixteen and a half. Teleological reasoning dropped significantly too, from a mean of three point eight on the five-point scale to two point four. Those two changes were linked: the students who shed teleological habits most thoroughly were the ones who gained most in understanding. What did acceptance of evolution predict? Incoming acceptance averaged four point one on the five-point scale — so this was not a class full of evolution skeptics. It correlated with parental attitudes at r equals point sixty-eight and with religiosity at r equals negative point forty-six. But in the regression predicting CINS learning gains, acceptance contributed essentially nothing — a standardized beta of negative point oh eight, non-significant. A student who entered the course skeptical of evolution could learn natural selection just as well as a student who fully accepted it, as long as their teleological reasoning was lower. And here is the other reversal the study documents: improving understanding of natural selection did not drive up acceptance of evolution. The two constructs moved somewhat independently. Understanding and acceptance are separable — shaped by different variables and responsive to different kinds of instruction. The implications for teaching are direct. If the goal is students who understand natural selection, the leverage point is teleological reasoning, not prior acceptance. That means explicitly teaching students to notice and interrupt goal-directed explanations — not just presenting the correct mechanistic account, but helping students recognize the cognitive habit that keeps pulling them away from it. The evolutionary medicine framing may help because it grounds natural selection in health contexts that students find immediately relevant, giving the mechanistic logic somewhere concrete to land. Barnes and colleagues note real limitations. There was no comparison group taking a conventional evolution course, so it is impossible to isolate what the evolutionary medicine framing contributed versus what any good evolution instruction would produce. The teleological measure drew on only three of twelve available items. The sample also skewed toward students who already accepted evolution fairly strongly — future work needs to test these findings in classrooms where skepticism runs higher. The study also did not track whether acceptance changed over the semester, so the full relationship between learning and belief remains open. But the core finding is clear enough to act on. Biology education has spent decades pulling on the cultural lever — trying to reduce the perceived conflict between evolution and religious identity, hoping that if students accept the fact, understanding will follow. This study suggests that is the wrong lever, or at least not the most powerful one. The field may have been measuring the obstacle incorrectly. What stands between a student and a genuine grasp of natural selection is not, primarily, what they believe about the origin of species. It is whether they habitually explain the natural world in terms of purpose and need. Address that habit, and learning happens — regardless of what students accept. 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.