Monday, April 16, 2018

Does Activism in Social Science Explain Conservatives’ Distrust of Scientists?

Does Activism in Social Science Explain Conservatives’ Distrust of Scientists? Nathan Cofnas, Noah Carl, Michael A. Woodley of Menie. The American Sociologist, March 2018, Volume 49, Issue 1, pp 135–148.

Abstract: Data from the General Social Survey suggest that conservatives have become less trustful of scientists since the 1970s. Gauchat argues that this is because conservatives increasingly see scientific findings as threatening to their worldview. However, the General Social Survey data concern trust in scientists, not in science. We suggest that conservatives’ diminishing trust in scientists reflects the fact that scientists in certain fields, particularly social science, have increasingly adopted a liberal-activist stance, seeking to influence public policy in a liberal direction.


Gauchat claimed that conservatives had less trust in "science" than liberals. We observed that he found only that they have less trust in scientists, not science, and that there is independent evidence that conservative distrust is directed toward what McCright et al. (2013) call "impact scientists" (e.g., social scientists) rather than "production scientists." We provided evidence that leading social scientists and social science organizations misrepresent research in order to influence public policy in a liberal direction, tolerate censorship of work that challenges liberal beliefs, uncritically accept dubious scientific findings that paint conservatives in an unflattering light, and practice a variety of forms of discrimination against conservative scholars. Conservatives’ recognition of this reality could explain why only 38% of conservatives in 2010, compared with 50% of liberals, said that they had "great deal of confidence" in "the scientific community" (Gauchat 2012).

Losing the trust of conservatives may not ne the only bad consequence of liberal activism in social science. Science itself is harmed. As Weber (2009:146) warned, "whenever the man of science introduces his personal value judgement, a full understanding of the facts ceases." Today, social science is facing a "replication crisis" (Open Science Collaboration 2015): Many findings that were thought to be firmly established are turning out not to be replicable when tested more carefully. It is noteworthy that a significant number of the effects that are falling victim to the replication crisis either supported liberalism or were somehow unflattering to conservatives. "Stereotype threat" is perhaps the most striking example. Since stereotype threat was proposed to explain gaps in the test scores of blacks and Whites more than two decades ago (Steele and Aronson 1995), it has become one of the primary liberal explanations for group differences in performance and has spawned many thousands of follow-up studies. Yet it may turn out that it was all a mistake—a consequence of publication bias and questionable research methods (Ganley et al. 2013; Jussim 2015). Other studies that could not be replicated,while not being explicitly anti-conservative, subtly support liberal ideas or cast conservatives in a bad light. For example, studies that could not be replicated include one where people "increased their endorsement of a current social system after being exposed to money" and another where Americans became more conservative after seeing a U.S. flag (Yong 2013). The former makes money seem to be bad thing, in line with liberal skepticism of capitalism. The latter suggests that conservatism is a primal reaction to tribal symbols. Virtually none of the non-replicable effects were at all favorable to conservatism. This suggests that findings that might favor conservatism are scrutinized much more carefully than those that favor liberalism—if they are not censored or rejected for explicitly moral reasons (e.g., Gardner 2001; Sternberg 2005).

In the past few years, a number of social scientists, led by Jonathan Haidt, have called upon social scientists to diversify the field and make a conscious effort to root out liberal bias (Duarte et al. 2015). We conclude with a prediction: If social scientists begin counteracting liberal activism, the trend of lowering conservative trust in scientists will reverse.

No comments:

Post a Comment