Figure 1 - uploaded by Jarret T Crawford
Content may be subject to copyright.
(Hilbig and Moshagen). Moving average (3 periods) of U.S.-party positions on the logit left-right (LLR) scale over time.
Source publication
Psychologists have demonstrated the value of diversity—particularly diversity of viewpoints—for enhancing creativity, discovery, and problem solving. But one key type of viewpoint diversity is lacking in academic psychology in general and social psychology in particular: political diversity. This article reviews the available evidence and finds sup...
Citations
... Research has shown that in the U.S., the U.K., and many other Western nations, there are more left-leaning than right-leaning journalists [19], [20], [21]. Similarly, academics also tend to lean, on average, left-of-center [22], [23], [24]. If the political preferences of individuals within the news media and academia influence the content they produce-especially content with political implications-and this content is subsequently used to train LLMs, then prevailing perspectives within these institutions could percolate into the models trained on that content. ...
Political biases in Large Language Model (LLM)-based artificial intelligence (AI) systems, such as OpenAI's ChatGPT or Google's Gemini, have been previously reported. While several prior studies have attempted to quantify these biases using political orientation tests, such approaches are limited by potential tests' calibration biases and constrained response formats that do not reflect real-world human-AI interactions. This study employs a multi-method approach to assess political bias in leading AI systems, integrating four complementary methodologies: (1) linguistic comparison of AI-generated text with the language used by Republican and Democratic U.S. Congress members, (2) analysis of political viewpoints embedded in AI-generated policy recommendations, (3) sentiment analysis of AI-generated text toward politically affiliated public figures, and (4) standardized political orientation testing. Results indicate a consistent left-leaning bias across most contemporary AI systems, with arguably varying degrees of intensity. However, this bias is not an inherent feature of LLMs; prior research demonstrates that fine-tuning with politically skewed data can realign these models across the ideological spectrum. The presence of systematic political bias in AI systems poses risks, including reduced viewpoint diversity, increased societal polarization, and the potential for public mistrust in AI technologies. To mitigate these risks, AI systems should be designed to prioritize factual accuracy while maintaining neutrality on most lawful normative issues. Furthermore, independent monitoring platforms are necessary to ensure transparency, accountability, and responsible AI development.
... First, we found that by far most of the authors of the publications we reviewed are affiliated with universities in the Global North. While overrepresentation of knowledge production institutions located in the global north might not be unique for fundamentalism studies, we still think this finding is worth reporting as empirical (e.g., Duarte et al., 2015) and theoretical studies (e.g., Anderson, 1995;Elgin, 2017) have shown that diversity along many dimensions benefits nuanced scholarship and has a bigger chance of counteracting bias. Therefore, while biasing effects of the accumulation of epistemic power in the global north most likely is not a problem unique for fundamentalism studies, these effects nevertheless merit attention. ...
This scoping review of conceptualizations of fundamentalism scrutinizes the concept's domain of application, defining characteristics, and liability to bias. We find fundamentalism in four domains of application: Christianity, other Abrahamic religions, non-Abrahamic religions, and non-religious phenomena. The defining characteristics which we identify are organized into five categories: belief, behavior, emotion, goal, and structure. We find that different kinds of fundamentalisms are defined by different characteristics, with violent and oppressive behaviors, and political beliefs and goals being emphasized for non-Christian fundamentalisms. Additionally, we find that the locus of fundamentalism studies is the Global North. Based on these findings, we conclude that the concept is prone to bias. When conceptualizing fundamentalism, three considerations deserve attention: the mutual dependency between the domain of application and the specification of defining characteristics; the question of usefulness of scientific concepts; and the connection between conceptual ambiguity and the risk of bias in the study of fundamentalism.
... Furthermore, politics can shape individuals´ perception of the expertise, trustworthiness, or objectivity of scientists, resulting in the dismissal of their messages due to a perceived lack of credibility (Philipp-Muller et al., 2022). Scientists may be perceived as imbuing their research and methods with liberal values, thereby producing conclusions that mischaracterize conservatives (Duarte et al., 2015) or exhibit bias against Christians (Barnes et al., 2020). ...
... This approach reflects Slovak challenges with left-right self-identification, common in post-communist contexts (Otjes & Rekker, 2021). Political conservatism's association with unfavorable attitudes toward science in Western studies (Mann & Schleifer, 2020) may extend to Slovakia, where some conservatives may perceive scientists as aligned with liberal values (Duarte et al., 2015). However, the varied influence of religiosity and conservatism as predictors underscores the need for context-sensitive methodologies when studying populist attitudes. ...
Science-related populism has recently emerged alongside political populism. However, its occurrence, connections with political populist attitudes, and underlying factors have not been explored in the Central-European context. This study seeks to achieve three key objectives: (1) examine the relationships among science-related populist attitudes, their subdimensions, and political populist attitudes; (2) assess the prevalence of populist attitudes; and (3) analyze a set of explanatory and exploratory factors contributing to political and science-related populist attitudes. Drawing from a national sample of 643 Slovaks, our study highlights the importance of addressing science-related populism in parallel with political populism. It elucidates the intricate interplay between these attitudes, with political populism being more prevalent in the Slovak population, albeit intertwined with science-related populism. Several contributing factors, such as education, political orientation, religiosity, trust in science, relative deprivation, faith in intuition, and susceptibility to pseudoscientific beliefs, were identified. These factors offer valuable insights into the emergence of populist attitudes in Slovakia. The findings from this study hold significance for further research aimed at comprehending and addressing the phenomenon of science-related populism within the Central-European context.
... The possibility that these "cries" may be well-founded does not occur to them. That psychology (in particular) already has very serious problems with ideological (and, in particular, left-wing and Marxist) bias is virtually inarguable Clark et al., 2024;Duarte et al., 2015;Honeycutt & Jussim, 2023). Consistent with this is the belief held by many psychologists (both in private and in public) that the APA has become little more than a (leftist) partisan advocacy organization (Fergusson, 2023;Silander & Tarescavage, 2023). ...
... However, acknowledging the "moral ordinariness" of scientists offers a more grounded and realistic perspective. Research in social psychology shows that academics, like individuals in any other sector, are susceptible to the same coercive social pressures and groupthink prevalent throughout wider society (Duarte et al. 2015). This vulnerability often manifests as confirmation bias, where researchers are inclined to interpret data in ways that reinforce their existing beliefs, rather than seeking evidence that might contradict them. ...
This article offers an appreciative but critical response to Donovan Schaefer’s book Wild Experiment, focusing on the connection between emotion and reason, particularly in the context of conspiracy theories and scientific inquiry. While acknowledging the valuable insights provided by Schaefer’s cogency theory and its emphasis on the role of emotions in shaping beliefs and reasoning, I argue for a more nuanced understanding of the factors that contribute to the popularity of conspiracy theories and the success of science. In particular, I challenge Schaefer’s characterization of scientists as primarily driven by cold emotions and a fear of making mistakes. Instead, I emphasize the social structure of science and institutional practices that enable collective intellectual vigilance and the advancement of reliable knowledge.
... "[representants of woke culture] in their attempt to separate Black discursive accountability praxes--calling out, reading, and cancelling--from their origins in the creative spaces occupied by the oppressed, and reposition them as a threat to their real and aspirational peers, elite public figures fall victim to their own worst fears: a realisation that the social capital they've worked so hard for is hyperinflated currency in the attention economy" (Clark, 2020, p. 4 From the academic perspective, it is useful to explore whether "cancel culture" is a necessary tool for social justice or rather a harmful trend that undermines free expression in academia. Critics argue that the fear of being cancelled stifles open debate and discussion, leading to self-censorship (e.g., Williams, 2016) and a lack of diverse perspectives in academic and public discourse (e.g., Duarte et al., 2015;Clark, 2020;Norris, 2021;Traversa et al., 2023;etc.).13 ...
... Furthermore, the examination will explore the intersectionality of power dynamics and the voices that often bear the brunt of deplatforming efforts, shedding light on the potential consequences of suppressing these perspectives. By engaging in this analysis, we hope to uncover insights that can inform strategies for promoting open discourse while advancing diversity, equity, and inclusion in academic settings (Duarte et al., 2014;Norris, 2021;Sousa & Clark, 2018). In conclusion, a comprehensive analysis of deplatforming and its implications for free speech and DEI initiatives is necessary in order to understand the complexities and potential consequences of silencing certain voices and to find a balanced approach that promotes both open discourse and inclusivity in higher education (Norris, 2020). ...
Introduction: In recent years, the number of deplatforming attempts has increased at universities. Deplatforming is an attempt to block a person from speaking when certain groups find their content to be objectionable. An institution’s culture of free speech may affect the number of deplatforming incidents.
Methods: This study utilized data from a new database of deplatforming incidents on campus. The data were tested for correlations in relation to an institution’s free speech ranking and its level of DEI (Diversity, Equity, and Inclusion) staffing.
Results: These initial results show a negative relationship between free speech ranking and deplatforming incidents, and a positive relationship between DEI staffing and deplatforming incidents.
Discussion: Institutions face the challenge of balancing an inclusive and equitable environment with the preservation of robust intellectual diversity and academic freedom. Finding this balance requires a thoughtful and deliberate approach that considers the nuances of promoting diversity, equity, and inclusion while safeguarding the principles of free expression and open dialogue.
Limitations: These results were only for Power 5 universities in the United States. A more diverse sample may show different results.
Conclusions: The culture of an institution is related to the number of deplatforming incidents that take place. This culture can be examined via its commitment to free speech and to its DEI staffing levels. Initial results show that these two factors have a relationship with the number of deplatforming incidents.
... Furthermore, decades of psychological research suggest that when groups contain diverse viewpoints, they perform better in a wide range of creative, problem-solving, predictive, and discovery-based tasks than do individuals and homogenous groups. (See, for example, Moscovini and Personnaz 1980;Hastie 1986;Laughlin and Ellis 1986;Nemeth 1986;Nemeth 1995;Gigone and Hastie 1997;Kuhn et al. 1997;Mosheman and Geil 1998;Jehn et al. 1999;Brodbeck et al. 2002;Greitemeyer et al. 2006;Schulz-Hardt et al. 2006;van Knippenberg and Schippers 2007;Crisp and Turner 2011;Krause et al. 2011;Trouche et al. 2014;Duarte et al. 2015;and Woolley et al. 2015.) 8 Since groups with robust viewpoint diversity typically perform better in a wide range of intellectual tasks than do individuals or homogenous groups, it would not be surprising if scientific inquiry attempted to harness this fact. And indeed, this is just what philosophers of science have postulated. ...
p>In recent years, epistemologists have devoted enormous attention to this question: what should happen when two epistemic peers disagree about the truth-value of some proposition? Some have argued that that in all such cases, both parties are rationally required to revise their position in some way. Others have maintained that, in at least some cases, neither party is rationally required to revise her position. In this paper, I examine a provocative and under-appreciated argument for the latter view due to Catherine Z. Elgin (2010, 2012, 2017, and 2018). I defend it against a series of objections, and I then identify some fruitful ways in which her view could be developed further.</p
... In the postpositivism tradition, science requires an open and pluralistic environment in which often incompatible views are expressed and conflicting aims pursued in a rational manner (Popper, 1959). The interplay of diverse viewpoints-over time-produces progressively more accurate understandings of reality (Duarte et al., 2015). Scientists interacting with each other from different social locations yield an increasingly better understanding of intervention effectiveness. ...
Quantitative research plays an instrumental role in facilitating health and wellness. Effective interventions are developed, assessed, and refined through the application of logic and empirical evidence in a scientific milieu characterized by diverse views. However, the increasing colonization of academic discourse by postmodern/critical theory (PCT) may undermine foundational concepts upon which quantitative research rests. The postpositivist tradition in which quantitative methods are embedded emphasizes probabilistic truth, logic, empirical evidence, and the interplay of diverse, frequently conflicting perspectives. Conversely, some currents of thought within PCT emphasize multiple realities, intuition, personal stories of lived experience, and the exclusion of views that challenge PCT tenets. The increasing adoption of these concepts in academia may limit or even delegitimize quantitative research as currently practiced in social work. We conclude by arguing that social work clients are best served by a scientific discourse that values multiple theoretical frameworks, research methodologies, and scientists operating from different social locations.
... Selecting a diverse collection of reviewers for a article is especially important when the editor holds a strong position on the article topic. Most psychologists are liberal in their political orientation (Duarte et al., 2015). Yet the three reviewers of Hommel (2022), chosen by Fiedler and praised by him for their contributions, were seen by Roberts (2022) to be nonliberal, not balanced, and not diverse. ...
In this open science era, psychology demands researchers be transparent in their research practices. In turn, researchers might ask if journal editors are being equally transparent in their editorial practices. Editor bias is when editors fail to be fair and impartial in their handling of articles. Editor bias can arise because of identity—who authors are—or because of content—what authors write. Proposed solutions to editor bias include masking author identity and increasing editor diversity. What is needed is greater transparency. By being more transparent, editors would be in a better position to encourage others to embrace open science.
