Research on science communication has expanded and diversified. Correspondingly, an increasing number of literature reviews, handbooks, and bibliometric studies have taken stock of the field’s accumulated knowledge. However, it is still difficult to identify robust, consolidated answers to key questions on science communication—answers that may help scholars, practitioners, and decision-makers. Meta-analyses—the statistical aggregation of empirical results across multiple studies to determine average effect sizes—can provide some of those answers, complementing existing knowledge syntheses and addressing their shortcomings. Yet, meta-analyses in research on science communication are rare, whereas other disciplines are much closer to establishing a culture of meta-analyses and demonstrate the usefulness of meta-analyses for specific aspects of science communication research within and beyond our discipline. We therefore propose a meta-analytical research agenda of science communication scholarship and argue that it will increase the impact of our field in the social sciences and society.
This study examines public attitudes toward gene-edited foods in Japan, focusing on how underlying worldviews shape these responses. Drawing on a national survey and a latent class segmentation model based on 23 worldview variables, we identify four distinct segments: Communitarian, Environmentalist, Governmentalist, and Individualist. The findings indicate modest rather than sharply polarized differences in attitudes, with the Communitarian and Environmentalist segments expressing greater caution than the Governmentalist and Individualist segments. Across all four segments, gene-edited foods are broadly perceived as unnatural, yet support for commercialization remains comparatively high, suggesting that perceptions of unnaturalness do not necessarily translate into strong opposition to commercialization. This pattern suggests that evaluations of gene-edited foods are shaped by multiple normative considerations, in which moral or symbolic concerns coexist with pragmatic judgments about societal benefits and technological governance. These findings suggest that similar levels of support or skepticism can emerge from differing normative orientations.
Amid the politicization of science and the globalization of technological applications, does the national-origin label of artificial intelligence authors affect their perceived credibility? Drawing on the tension between artificial intelligence as an apolitical referee (machine heuristic) and artificial intelligence as a geopolitical actor like humans (social identity theory and computers-are-social-actors paradigm), two online experiments were conducted in China to examine the role of author type (human, artificial intelligence, and human-artificial intelligence hybrid) and national-origin labels in scientific content with varying degrees of politicization. In a highly politicized context, domestic sources were preferred across all author types, with nationalism amplifying this preference. The effect of artificial intelligence's national-origin label was significant but weaker than that for human authors. In low-politicized contexts, the national-origin label no longer significantly affected artificial intelligence credibility, while foreign humans were perceived as more credible. These findings highlight the intersecting dynamics of scientific politicization and the public politicized understanding of artificial intelligence technology.
Prevailing approaches to measuring public trust in scientists rely on facet averages that overlook within-person variability, a dimension traditional views equate with ambivalence or attitudinal weakness. This study introduces the Structural Contingency Model of Trust (SCMT), proposing that the functional meaning of this variability is contingent on an individual's overall trust level. In a Bayesian analysis of survey data from 68 countries (N = 69,534), higher Trust Level (mean across 12 facets) strongly predicted greater general confidence in scientists. Critically, a credible interaction effect supported the SCMT: among highly trusting individuals, greater Trust Variability (variance across facets) was associated with the highest probability of reporting a great deal of confidence in scientists. This Critical Support profile suggests that variability can signify robust, differentiated engagement rather than evaluative weakness, offering a nuanced framework for understanding the complex, multidimensional nature of public trust in scientists.
Public understanding of science is often framed as either factual knowledge or generalized trust in experts. This article advances an alternative: an institutional model, grounded in cultural awareness of the norms, practices, and boundaries through which science produces credible knowledge. Existing approaches struggle to distinguish scientific controversy from broader social conflict, limiting their ability to explain why science's legitimacy falters in politicized contexts. An institutional perspective instead shifts the focus from individual deficits to science's role as a social system-embedded in tensions between expertise, politics, industry, and democratic accountability. We outline the implications for science communication and public engagement, arguing that meaningful participation requires not only understanding science's institutional role in civic life but also the capacity to evaluate scientific actors against cultural expectations and normative commitments. The article concludes by proposing a research agenda centered on the cultural meaning of science and how its credibility depends on public evaluation.
