Those with the strongest opinions on complex societal issues are highly selective in their consideration of relevant evidence

From Non-Scientific Beliefs Among Undergraduate Students by Chris Impey, Sanlyn Buxner, and Jessie Antonellis. There are some definitional bones to pick but this research has the advantage of a large population base (N=11,000) and is longitudinal (22 years).

In discussions among scientists and educators, it is often assumed that pseudoscience belief stems from a lack of critical thinking skills and that pseudoscience belief is negatively correlated with science knowledge and an understanding of how science works. For the analysis described in this paper, those conjectures were hypotheses to be tested. In addition, we were interested in how pseudoscience beliefs change during progress through the undergraduate experience, and whether or not susceptibility to pseudoscience is related to religious belief.

Although we see small differences in the performance of students who held strong faith-based beliefs or strong beliefs in unscientific phenomena, these corresponded to about a one point difference, out of 15, in science literacy. This difference is less than the difference we see between science and education majors in the sample, which corresponds to an average of two points out of 15 on the science literacy scale. We also see relatively small reductions in pseudoscience belief over the course of an undergraduate career that includes at least three science courses. Confused signs of the nature of any relationship between formal education and pseudoscientific thinking have also been noted elsewhere (Goode 2002).

These generally null results present a challenge to educators. If nonscientific ways of thinking are widespread and resistant to standard modes of instruction, do educators need to be concerned and what do belief systems have to do with efforts to create a citizenry that is informed enough to vote on issues that involve science and technology? Some commentators have contended that pseudoscience and the paranormal are often skirted in the classroom, or treated as taboo subjects, meaning that nonscientific belief systems are not confronted (Martin 1994). Another point of view holds that pseudoscience beliefs cannot be addressed without classroom content that is relevant to societal issues (Hobson 2000, 2008). But psychologists Lindeman and Aarnio (2007) find support for a conceptual framework where a variety of superstitious, magical, and pseudoscientific beliefs accrue from the same kind of ontological confusion, and analytic thinking coexists with nonscientific forms of intuitive thinking. Moreover, research shows that people tend to motive their reasoning process based on prior beliefs and biases (Kunda 1990), and those with the strongest opinions on complex societal issues are highly selective in their consideration of relevant evidence (Lord, Ross, and Lepper 1979). Both of these effects mitigate against a completely rational world view. Religious beliefs exist in a distinct realm since they are based on faith. They may be similarly decoupled from aspects of worldview that are associated with evidence and causal connection, explaining the weak relationship seen here between religious belief and science knowledge. Improving science literacy will require a more nuanced approach that takes account of this psychological landscape. Since we also have found that science knowledge and attitudes are not substantially altered by the college experience, we have embarked on a separate study to understand where young adults get their information about science and technology.

Sometimes you have to work to unclutter cryptic verbiage to get to the really interesting conclusions. Based on this evidence:

Religious believers hold only half as many unscientific beliefs as education majors.

A college education which includes at least three science courses has no significant impact on extent of pseudoscientific beliefs.

People reach the conclusions they wish based on their prior beliefs rather than on the evidence.

Those with the strongest opinions on complex societal issues are highly selective in their consideration of relevant evidence.

I love the concluding sentence - "Since we also have found that science knowledge and attitudes are not substantially altered by the college experience, we have embarked on a separate study to understand where young adults get their information about science and technology." Well, yes, that might be useful.

This is consistent with the more general finding a couple of years ago by Richard Arum and Josipa Roksa in their book Academically Adrift, where they found 45% of undergraduates “demonstrated no significant gains in critical thinking, analytical reasoning, and written communications during the first two years of college.” The most recent data indicates that there is some improvement over the course of their four year education. But by the time they are seniors, 36% still show no improvement in critical thinking, analytical reasoning and written communications.

One might posit that this is a distinctly American issue but the National Science Board's Science and Engineering Indicators, 2014 reports that "Levels of factual knowledge in the United States are comparable to those in Europe and are generally higher than levels in countries in other parts of the world."

Some observations.

"Students have a complex web of prior belief systems and understandings about nature based on experience, upbringing, popular culture, and social interactions." What I have referred to elsewhere as KESVB - Knowledge, Experience, Skills, Values and Behaviors.

Knowledge and educational skill gaps are material (up to three years) by the time children begin Kindergarten.

Those gaps widen over the duration of K-12.

University has little impact on knowledge and critical thinking in terms of measured results.

Given that Knowledge and Skills are the things one would expect to be most positively affected by education and yet are not, this seems to suggest that Values and Behaviors (often referred to as culture) has a disproportionate impact on outcomes.

Given that culture is established via the home environment, all this layering of evidence seems to support that the biggest single source of impact on life outcomes is parental values and behaviors. A belief not uncommon among the hoi polloi but not widely shared in academia. We spend a lot of time and money trying to make teachers and professors more effective and what the evidence suggests is that we ought to be spending that time and money figuring out how to make parents more effective; and indeed, figuring out what are the critical values and behaviors fostered by parents which make the biggest difference in life outcomes.

I'd suggest we actually know the answers to those questions but academia is reluctant to acknowledge those answers.