A Meta-Analysis Synthesizing the Effects of Three Uncertainty Types in Science Communication

Abstract

This meta-analysis synthesized 43 empirical articles to summarize the impact of communicating three distinct uncertainty types (i.e., epistemic uncertainty, consensus uncertainty, and technical uncertainty) in science communication. The results revealed divergent effects across these three uncertainty types. Specifically, communicating epistemic uncertainty was found to have a small yet positive effect on credibility perception (d = 0.09), with a nonsignificant impact on attitudes. Conversely, communicating consensus uncertainty (d = −0.12) and technical uncertainty (d = −0.15) demonstrated a slight negative effect on attitudes, without impacting credibility perception. In addition, topic domain, cultural uncertainty avoidance, and text length were important moderators.

Keywords

science communication uncertainty meta-analysis communication effect

Uncertainty constitutes an intrinsic facet of scientific inquiry (Pollack, 2003; Star, 1985). Scientific exploration is inherently characterized by complexities and intricacies, which pose challenges to straightforward explanations (Fischhoff, 2012). As new evidence emerges and methodologies evolve, scientific knowledge remains open to refinement and revision, which introduces an additional layer of uncertainty (Kuhn, 1970). Therefore, the intricate, evolving, and probabilistic nature of scientific endeavors is inevitably characterized by uncertainty (Bromme & Goldman, 2014). In this light, embracing uncertainty in the scientific process is crucial for promoting transparency within the science community (Thorp, 2023), building long-term public trust (Ihlen et al., 2022), and nurturing a sustainable scientific landscape (Kreps & Kriner, 2020).

Nevertheless, effectively conveying uncertainty to the public has long presented considerable challenges. For one thing, the intricate nature of scientific and technical information poses a significant obstacle to comprehensible communication, particularly when addressing a diverse audience (Osman et al., 2018). For another, integrating uncertainties in science communication carries the risk of evoking skepticism and misinterpretation among the general public (Kortenkamp & Basten, 2015). A notable illustration of this challenge manifested during the COVID-19 pandemic, wherein trust in scientists declined from 39% to 23% in response to the uncertainties faced by the scientific community as they progressively gained insights into the nature of the virus (Pew Research Center, 2023). Consequently, journalists often opt to report scientific findings as more certain beyond warranted (Retzbach & Maier, 2015) as a strategy to simplify complex information (Ebeling, 2008) and mitigate possible adverse repercussions (Lehmkuhl & Peters, 2016). All these aspects underscore the persistent dilemma confronted by the science community regarding whether and how to disclose uncertainties in science to the public.

Recently, a growing body of scholarship has explored the impact of communicating scientific uncertainty across diverse contexts, including medicine and health (Hendriks et al., 2023; Hinnant et al., 2023), environmental issues (Hendriks & Jucks, 2020; van der Bles et al., 2020), and cutting-edge technologies (Lin & Kim, 2022; Ratcliff et al., 2021). Despite the burgeoning literature on this front, a notable inconsistency prevails regarding the impact of uncertainty disclosure in science communication. For instance, some investigations suggest that communicating uncertainty yields adverse outcomes, such as increased scientific misperceptions (Cook et al., 2017), reduced trust in scientists (Han et al., 2018; Hinnant et al., 2023), and decreased support intention (Chang, 2015). Other studies, nevertheless, did not identify significant detrimental effects of uncertainty disclosure (Howe et al., 2019; Ratcliff & Wicke, 2023), or even observed a positive impact, such as increased credibility of scientists (Steijaert et al., 2021) and enhanced acceptance of scientific information (Lin & Kim, 2022). It is noteworthy that communicating scientific uncertainty is not simply a unidimensional construct; instead, it incorporates multiple types that permeate the processes of scientific communication (Rice et al., 2018). Against this backdrop, academics have exerted considerable attempts to define the communication of scientific uncertainty and forwarded a constellation of definitions, conceptualizations, and measurements for understanding the characteristics of various types of uncertainty (for reviews, see Lohse, 2023; Ratcliff, 2021). Scholars have argued that the heterogeneous conceptualizations and operationalizations of uncertainty in the literature may give rise to ambiguities and inconsistencies within the findings (Gustafson & Rice, 2020). To this end, a review study has attempted to consolidate the current findings, proposing a taxonomy of distinct types of uncertainty communication (Gustafson & Rice, 2020). While this conceptual review significantly enhanced our understanding of the nature and characteristics of uncertainty in science communication, it fell short of providing statistically meaningful insights into the precise magnitude of the effects of communicating uncertainty. Consequently, information pertaining to the most influential or significant type of uncertainty remains unknown. In light of the limitations, we seek to extend the scholarly discourse by undertaking a quantitative assessment of the effects associated with various types of uncertainty and the potential moderators.

The first goal of our study was to explore whether these effects of communicating are contingent on different types of scientific uncertainty. Following a comprehensive examination of the existing literature, we categorized uncertainty in science communication into three delineated classifications (i.e., epistemic, technical, and consensus uncertainties) (Gustafson & Rice, 2020; Ratcliff, 2021). On top of that, we aim to obtain a nuanced understanding by identifying potential moderators (i.e., topic domain, cultural uncertainty avoidance, and text length) that may influence the effects. This study contributes to a more comprehensive understanding of the impact of uncertainty disclosure in science communication, shedding light on the conditions that amplify, diminish, or nullify these effects. The findings also reveal promising avenues for future research.

The Role and Impact of Uncertainty Communication

Understanding and Communicating Uncertainty

Uncertainty refers to a condition characterized by the lack of definitive knowledge or predictability about the present or future when information is ambiguous, incomplete, or subject to variability (Babrow et al., 1998; Brashers, 2001). It encompasses both objective descriptions of uncertainty and subjective perceptions of uncertainty (Kahneman & Tversky, 1982; Ratcliff et al., 2022). The former involves expressing incomplete or imprecise information, while the latter refers to individuals’ cognitive and emotional responses to uncertain information.

Individuals experience uncertainty in interpersonal communication as a fundamental aspect of human interaction. Uncertainty reduction theory posits that humans have an innate drive to minimize uncertainty in social contexts (Berger & Calabrese, 1975). However, the uncertainty management theory offers a more nuanced perspective, suggesting that uncertainty can, in certain instances, lead to positive outcomes, such as hope and optimism, which may encourage proactive behaviors (Brashers, 2001). In addition, relational dialectics theory introduces the concept of dialectical tension, where individuals simultaneously seek stability (certainty) and novelty (uncertainty) in their relationships, necessitating a balance between these conflicting needs (Baxter & Montgomery, 1996; Baxter & Norwood, 2015). Together, these theories underscore the complex and multifaceted nature of uncertainty in interpersonal communication. Rather than being a solely negative experience, uncertainty can serve adaptive functions (Spiegelhalter & Riesch, 2011), facilitating information-seeking behavior (Bartoszek et al., 2022) and enhancing social bonding (Ersoy et al., 2023).

In science communication, a persistent tension in communicating uncertainty exists between the complex motivations of communicators and the evolving preferences of the audience. Some scientists minimize uncertainty to attract journalists’ interest (Maier et al., 2016) or to avoid negative consequences such as decreased public trust (Post & Maier, 2016). In contrast, others view transparent communication of uncertainty as an ethical imperative (Keohane et al., 2014), using it strategically to highlight further direction (Friedman et al., 1999), and hope that acknowledging uncertainty will foster more critical science engagement (Maier et al., 2016). Similar situations also apply to journalists, as they have to navigate the tension between accurate representation of uncertainty and the creation of engaging, accessible content for their audience (Dunwoody, 1999; J. Schneider, 2010). Moreover, public responses are shaped by individual differences, such as understanding of science (Frewer et al., 2002) and ambiguity tolerance (Ratcliff et al., 2023), which further highlight the complexity of the issue.

The Effects of Uncertainty Communication

Long-standing and intense debates focus on whether and how uncertainty should be conveyed to the public. In one regard, communicating uncertainty may lead to decreased trust in scientific institutions under certain circumstances (Besley & Nisbet, 2013), particularly due to the audience’s lack of knowledge (Maier et al., 2016) or pre-existing distrustful attitudes toward science (Frewer et al., 2002). In contrast, disclosing uncertainty can elevate trust in science (Johnson & Slovic, 1995), as a transparent and honest acknowledgment of uncertainty ultimately enhances perceived credibility (Jensen, 2008).

This intricate landscape has evolved as research efforts have expanded to explore a variety of uncertainty manifestations across different contexts of science communication (e.g., Gustafson & Rice, 2019; van der Bles et al., 2019). For example, Gustafson and Rice (2019) examined the effect of presenting disagreement (e.g., “although some scientists disagree”) as a form of uncertainty on people’s perceptions and attitudes within the environment and technology domains, observing a small negative effect. Another study compared the effects of point estimation (e.g., 116,000) versus range estimation (e.g., range between 17,000 and 215,000) on credibility perception, revealing no discernible difference between the two conditions (van der Bles et al., 2020). Recent scholarship has posited that heterogeneity may exist in both the conceptualization and operationalization of uncertainty communication and suggested that attention to this issue may provide a pathway to reconcile the contradictory findings (e.g., Eveland & Hively, 2009; Nir, 2011). Below, we will present a discussion on the prevalent typologies and propose our expectations regarding their respective impacts.

The Typology of Scientific Uncertainty

The earliest attempt to categorize uncertainty may primarily rely on the media’s reports of controversial topics. For example, Simis (2013) proposed engagement and temporal outcomes as two dimensions to categorize uncertainty and classified uncertainty into four types: personal-immediate, personal-long-term, remote-immediate, and remote-long-term uncertainty. Ruhrmann et al. (2015) studied presentations of uncertainty in molecular medicine reporting on television, delineating uncertainty primarily as conflicts or disputes among evidence and experts.

Gustafson and Rice (2019, 2020) went beyond the exclusive focus on controversy as the source of uncertainty and introduced a more comprehensive typology encompassing four distinct types of uncertainty: scientific, deficient, consensus, and technical uncertainty. This categorization originates from diverse aspects of the entire scientific procedure, including the nature of science, epistemology, conflicting evidence, and technology. Building upon this groundwork, Ratcliff (2021) divided uncertainty into study uncertainty, epistemic uncertainty, and consensus uncertainty, reflecting the methodological, scientific, and evidential origins, respectively. In our investigation, we adopted Ratcliff’s (2021) trichotomy, recognizing that scientific and deficient uncertainties posited by Gustafson and Rice (2019, 2020) often exhibit overlap (Li & Du, 2017, p. 4; van der Bles et al., 2019) and that the scholarly examinations of the uncertain nature of science have been relatively sparse in the literature (Gustafson & Rice, 2020).

Epistemic uncertainty reflects the inherently uncertain nature of science, acknowledging the presence of unknowns and limitations in scientific knowledge. It also includes recognizing known knowledge gaps and highlighting areas where our current understanding is limited (Pollack, 2003; Ratcliff, 2021). Epistemic uncertainty is typically described in qualitative terms, such as “seismic risk assessment is a field with little evidence to guide experts” (e.g., Nakayachi et al., 2018) or “it remains to be seen whether this medication is effective” (e.g., Steijaert et al., 2021). Some studies found epistemic uncertainty yielded nonsignificant main effects (Howe et al., 2019; Jensen et al., 2017). Nevertheless, Ratcliff et al. (2021) revealed that although disclosing epistemic uncertainty had no impact on attitudes toward the science community, it did lead to higher perception of researchers’ ethicality. Another study also indicated that explicitly acknowledging the epistemic uncertainty in seismic risk estimation increased the perceived honesty and openness of the experts (Nakayachi et al., 2018).

Technical uncertainty refers to inherent constraints in the tools, methods, or technologies employed in research procedures, which may result in imprecisions and possible inaccuracies.¹ These constraints encompass factors of equipment imprecision, measurement errors, and complexities in data collection (Gustafson & Rice, 2019; van der Bles et al., 2019). Scientists commonly manage uncertainty by converting it into systematic parameters (Friedman et al., 1999). This is often achieved by employing approaches such as numerical ranges (e.g., 16–24 out of 100) or probability distributions (e.g., 19%). Occasionally, it is communicated in a descriptive form, such as “Around this estimate, it could be somewhat higher or lower.” Some studies found that utilizing numerical ranges resulted in heightened concern and increased credibility perceptions toward communicators (Joslyn & Demnitz, 2021; Joslyn & LeClerc, 2016). Conversely, other studies found no significant effect of technical uncertainty on trust (Kerr et al., 2021; van der Bles et al., 2020).

Consensus uncertainty pertains to inconsistent evidence or disagreements among relevant stakeholders (e.g., scientists; government officials). This uncertainty manifests in several ways. One common form of presentation is through conflicting statistical proportions, which reveal varying levels of agreement among experts (Johnson, 2018). In addition, consensus uncertainty can be framed by presenting a conflicted story (Dixon et al., 2015) or by conveying a two-sided message (Lin & Kim, 2022). Experimental studies have shown that consensus uncertainty can have negative or null effects on attitudinal support and perceived credibility (Cook et al., 2017; Dixon et al., 2015). However, some studies suggest that presenting opposing scientific viewpoints can enhance attitudes or perceptions of credibility (Hendriks et al., 2023; Lin & Kim, 2022).

Based on an extensive literature review, we identify uncertainty perception, credibility perception, attitude, and behavioral intention as the four most relevant and prominent outcome variables of uncertainty in science communication. Uncertainty perception is defined as an individual’s subjective feelings of ambiguity or lack of certainty regarding a specific topic.² Credibility typically refers to the degree of trust in the authenticity and reliability of information or sources. Attitude pertains to a set of (un)favorable evaluations of the target object or behavior. Behavioral intention refers to the expectation of engaging in or planning behavior aligned with the persuasive intent of the information. Hence, the first set of research questions is as follows:

Research Question 1: How does each type of uncertainty affect uncertainty perceptions?

Research Question 2: How does each type of uncertainty affect credibility perceptions?

Research Question 3: How does each type of uncertainty affect attitudes?

Research Question 4: How does each type of uncertainty affect behavioral intentions?

Moderators

The effects of different types of uncertainty disclosure vary depending on the message, individual, and contextual factors (Ratcliff et al., 2022). Therefore, the second goal of this study is to explore potential factors moderating the effects of communicating scientific uncertainty.

First, we propose the topic domain as a potential moderator. Health, environment, and technology stand out as three prominent domains frequently covered by science communication. Previous research has found that uncertainty may have different effects across different domains. For example, Gustafson and Rice (2019) found that consensus uncertainty had a negative impact on environment domains (i.e., climate change), while no such impact was found in technology domains (i.e., Genetically Modified Organisms [GMOs]). Another study by Dixon et al. (2015) found that consensus uncertainty had a negative impact on credibility perceptions in health domains (i.e., the autism-vaccine link). This suggests that the influence of scientific uncertainty may depend on the specific domain.

Second, we deem that the effects of communicating uncertainty may exhibit variability depending upon recipients’ cultural uncertainty avoidance tendencies. Cultural uncertainty avoidance refers to “the extent to which cultural members feel threatened by uncertain or unknown situations” (Hofstede, 2013, p. 161). Specifically, countries with higher cultural uncertainty avoidance have a lower tolerance for ambiguity and are less willing to take risks and face unknown risks, while the opposite is true for countries with lower cultural uncertainty avoidance. For example, Kerr et al.’s (2021) cross-national study found that South Korea, with a high cultural uncertainty avoidance ranking, experienced a significantly greater negative impact of technical uncertainty on credibility perception compared to other countries. Conversely, no significant differences were observed between Japan and France, both of which have similar cultural uncertainty avoidance rankings. All these findings suggest that cultural uncertainty avoidance is a potential moderating variable.

The length of information is considered an important factor influencing the communication effects. According to the elaboration likelihood model (e.g., Petty et al., 1986), message length is considered both central and peripheral routes that allow for judgments of argument quality. Thus, longer messages are generally perceived as more informed because they appear to contain more arguments and additional background information. A meta-analysis on narrative persuasion suggested a facilitating effect of long texts on the effectiveness of narrative messages (Shen et al., 2015). Building on this analysis, we selected message length as the third potential moderator:

Research Question 5: Are there any moderating effects of (a) topic domain, (b) cultural uncertainty avoidance, and (c) text length on the impact of scientific uncertainty on uncertainty perceptions?

Research Question 6: Are there any moderating effects of (a) topic domain, (b) cultural uncertainty avoidance, and (c) text length on the impact of scientific uncertainty on credibility perceptions?

Research Question 7: Are there any moderating effects of (a) topic domain, (b) cultural uncertainty avoidance, and (c) text length on the impact of scientific uncertainty on attitudes?

Research Question 8: Are there any moderating effects of (a) topic domain, (b) cultural uncertainty avoidance, and (c) text length on the impact of scientific uncertainty on behavioral intentions?

Method

Literature Search

We conducted the literature search in major electronic databases.³ The search strategy was designed to balance high-level scientific terms with specific area terminology to ensure comprehensive coverage.⁴ The search was limited to articles published before July 8, 2023. This initial database search generated 1509 results. In addition, we examined the reference list of review articles on this topic (Gustafson & Rice, 2020; Lohse, 2023; Ratcliff, 2021, n = 48). To be thorough, we performed backward and forward reference searches through Google Scholar based on the collected publications to find further eligible studies (n = 28). We conducted another round of searches in July 2024 and identified another four articles.⁵

Study Selection

Following the screening of titles and abstracts, 91 articles were selected, including journal articles, proceedings, and dissertations. Then, the studies selected met the following criteria. First, the article measured and treated (a) scientific uncertainty which we defined in the front (i.e., consensus uncertainty, technical uncertainty, and epistemic uncertainty) as the independent variable and (b) uncertainty perceptions or other outcomes (i.e., credibility perceptions, attitudes, and behavioral intentions) as dependent variables. Accordingly, irrelevant articles were excluded (n = 14). Second, each included study employed an experimental design comparing uncertainty condition(s) with a control group. Those studies lacking eligible uncertainty comparisons (n = 3) or their control group was not exposed to any message (n = 5) were excluded from our analysis. Third, the articles had to focus on the scientific context; studies conducted in socioeconomic contexts (e.g., unemployment estimation, tax) were excluded (n = 1). Fourth, each included study must report sufficient statistical information (e.g., means, standard deviations, t-test values, or correlation coefficients). When no statistics were available in the cases mentioned in the article, we contacted the authors for relevant data. Studies for which we did not receive the information were subsequently excluded (n = 23).⁶ Finally, duplicative studies were excluded (n = 6). The details of exclusion and inclusion criteria can be found in Supplemental Appendix A.

After applying the above criteria, 43 articles were selected as the conclusive sample for the meta-analysis. They yielded 71 empirically independent studies, with a total sample size of 62,895 individuals. The details of the searching and screening procedure are presented in Figure 1.

Figure 1.

Flow Chart of Study Retrieval and Selection Process.

Variable Coding

Uncertainty Typology and Outcome Variable Coding

As defined previously, the uncertainty disclosure was categorized into three types. The definition, examples of operationalization, and inter-coder reliability (ICR) for each type are presented in Table 1.

Table 1.

Coding of Uncertainty Types.

Types	Definition	Operationalization	ICR
Consensus uncertainty	Inconsistent evidence or disagreements among relevant stakeholders (e.g., scientists, government officials)	• Compare high and low proportion rates of expert consensus. Examples: “An estimated 97% [60%] of medical scientists agree that . . .” (Kerr & van der Linden, 2022); “About 98% [60%] of all scientists believe that the results are credible.” (Johnson, 2018) • Present either one-sided or two-sided messages. Examples: “. . . Because there is a disagreement between experts” (C. R. Schneider et al., 2022); “The article presents two opposing scientific viewpoints regarding . . .” (Dunwoody & Kohl, 2017).	0.98
Epistemic uncertainty	The nature of science inherently involves recognizing the presence of unknowns, limitations in scientific knowledge, and areas where our current understanding is limited.	• Add information emphasizing the inherent uncertainty of the science process. Examples: “. . . More studies are needed to confirm whether the vaccine is effective at preventing severe cases of COVID-19.” (Kelp et al., 2022) “But the limits of biomedical science should be acknowledged . . . There is no guarantee of breakthrough scientific discoveries.” (Ratcliff et al., 2021)	0.97
Technical uncertainty	Inherent constraints in the tools, methods, or technologies employed in scientific research procedures, which result in imprecisions and possible inaccuracies.	• Portray uncertainty as quantitative form (e.g., probability, interval) and compare a point estimation with range estimation. Examples: “Average temperature increase of 3°F [2–4°F]” (Joslyn & LeClerc, 2016) “About 17% [between 10% and 34%] of those who catch it need hospital care.” (Kerr et al., 2021)	1.00

Note: Example stimuli only show the key manipulated parts; the rest remains consistent between the two conditions.

In sum, four outcomes were examined: uncertainty perception, credibility perceptions, attitudes,⁷ and behavioral intentions. All the experiments assessed multiple effects. For example, Chang (2015) assessed the effect of consensus uncertainty on credibility perceptions, attitudes, and behavioral intentions. Consistent with Schmidt and Hunter’s (1999) suggestion, each effect in the current meta-analysis was treated as a separate unit of analysis. In total, the 71 empirical studies yielded 128 effects.

Moderator Coding

Topic Domain

The included studies were categorized into three major domains: health (e.g., nutrition, vaccines), environment (e.g., climate change, global warming), and technology (e.g., nanotechnology, gene editing).⁸

Cultural Uncertainty Avoidance

Referring to Hofstedes’ index (0–100) (Hofstede, 1983, 2013), the level of cultural uncertainty avoidance for countries involved in the study was encoded as low or high. Following previous research practices (Wang et al., 2024; Watts et al., 2020), the 50th percentile was used as the cutoff point to divide countries into high (e.g., Japan, South Korea) and low (e.g., the United States, Germany) groups.⁹

Text Length

Following previous research approaches (Larrimore et al., 2011), text length was coded as a continuous variable by retrieving the stimuli of each study.¹⁰ Text length ranged from 23 to 1,227 words (Mdn = 123).

The coding process began with a discussion among the three authors to create the codebook framework. The first author then developed the detailed codebook, training the second author. To ensure reliability, two authors independently coded a random subsample of half the total sample. The inter-coder agreement was assessed using Krippendorff’s alpha, which ranged from .92 to 1,¹¹ indicating a high level of reliability. Discrepancies were resolved through discussions among all three authors. After validation, the first author coded the remaining articles. The complete coding books are provided in Supplemental Appendix A.

Effect Size Computation and Analysis

The standardized mean difference, Cohen’s d, was calculated to represent the effect of scientific uncertainty on outcomes (Cohen, 1992). A positive d value indicated that scientific uncertainties enhanced outcomes. Contrary to the fixed effects model, the random-effects model does not assume homogeneity in the study population and allows for generalization beyond the scope of the meta-analysis (Borenstein et al., 2010). Therefore, we used the random-effects model to estimate for all analyses. The Q statistic and I² were used to assess the heterogeneity of the effect size. Finally, we examined publication bias using two approaches. We employed a funnel plot to visually assess the potential absence of studies with small effect sizes. Subsequently, we conducted Egger’s test, which offered statistical confirmation of publication bias (Egger et al., 1997). Statistical analyses were conducted using the Comprehensive Meta-Analysis Software (CMA) 3.0 program (Borenstein et al., 2013).

Results

Among these 71 studies, sample sizes ranged from 57 to 2,000 individuals (Mdn = 380), and 30 of these studies were published after 2020. The gender ratios varied from 31% to 100% for female participants (Mdn = 53%). In addition, 39% of the studies utilized student samples. Table 2 illustrates a summary of the studies included in our analysis. Table 3 shows the resluts of moderator analysis. The forest plots for main effects and moderator analysis can be found in Supplemental Appendices B and C.

Table 2.

Overview of Studies Included in Meta-Analysis.

Author names	N	Uncertainty type	Outcome variables	Topic domain	Uncertainty avoidance	Length
Adams et al. (2024)	1,186	EU	IU, CP	Health	LUA	158
Aklin and Urpelainen (2014)—Study 1	418	CU	A	Environment	LUA	53
Aklin and Urpelainen (2014)—Study 2	800	CU	A	Environment	LUA	63
Chang (2015)—Study 3	120	CU	IU, CP, A, BI	Health	HUA	N/A
Chinn et al. (2018)	377	CU	IU, BI	Health	LUA	128
Clarke et al. (2015)	128	CU	IU	Health	LUA	417
Dixon and Clarke (2013)	164	CU	IU	Health	LUA	N/A
Dixon et al. (2015)	124	CU	IU	Health	LUA	N/A
Dunwoody and Kohl (2017)	505	CU	IU	Environment	LUA	200
Flemming et al. (2020)—Study 1	300	EU	IU, CP	Technology	LUA	909
Flemming et al. (2020)—Study 2	480	EU	IU, CP	Technology	LUA	919
Flemming et al. (2020)—Study 3	68	EU	IU, CP	Health	LUA	1,227
Flemming et al. (2020)—Study 4	68	EU	IU, CP	Environment	LUA	1,172
Haigh and Birch (2021)—Study 2	400	CU	A, CP	Health	LUA	26
Han et al. (2018)	1,790	EU	A, CP, BI	Health	HUA	202
Hendriks and Jucks (2020)—Study 2	57	CU	CP	Environment	LUA	702
Hendriks et al. (2023)	603	CU	CP	Health	LUA	414
Hinnant et al. (2023)	172	CU	IU, A, CP	Health	LUA	764
Howe et al. (2019)	405	TU	CP	Environment	LUA	183
Jensen and Hurley (2012)	242	CU	IU, CP	Environment	LUA	N/A
Jensen (2008)	601	EU	IU, CP	Health	LUA	613
Jensen et al. (2011)	1,016	EU	A	Health	LUA	613
Johnson (2018)—Topic 1	164	CU	IU, BI	Health	LUA	127
Johnson (2018)—Topic 2	185	CU	IU, BI	Health	LUA	129
Johnson (2018)—Topic 3	181	CU	IU, BI	Technology	LUA	173
Johnson (2018)—Topic 4	175	CU	IU, BI	Environment	LUA	146
Johnson (2018)—Topic 5	165	CU	IU, BI	Health	LUA	132
Johnson (2018)—Topic 6	180	CU	IU, BI	Health	LUA	110
Johnson (2018)—Topic 7	172	CU	IU, BI	Technology	LUA	139
Johnson (2018)—Topic 8	188	CU	IU, BI	Technology	LUA	118
Joslyn and LeClerc (2016)—Study 1	780	TU	CP, A	Environment	LUA	33
Joslyn and Demnitz (2019)—Study 1	416	TU	CP	Environment	LUA	71
Joslyn and Demnitz (2019)—Study 2	382	TU	CP	Environment	LUA	71
Joslyn and Demnitz (2021)	1,331	TU	CP, A	Environment	LUA	67
Jurhill (2022)	195	TU	CP	Health	LUA	37
Kalny and Walter (2024)—Study 2	344	EU	A	Technology	LUA	401
Kelp et al. (2022)	117	EU	CP, A, BI	Health	LUA	62
Kerr and van der Linden (2022)—Study 2	1,156	CU	A, BI	Health	LUA	23
Kerr et al. (2021)—Study 1 Aus	441	TU	IU, CP	Health	LUA	36
Kerr et al. (2021)—Study 1 Chi	468	TU	IU, CP	Health	HUA	36
Kerr et al. (2021)—Study 1 Fra	2,000	TU	IU, CP	Health	HUA	36
Kerr et al. (2021)—Study 1 Ger	462	TU	IU, CP	Health	LUA	36
Kerr et al. (2021)— Study 1 Ita	391	TU	IU, CP	Health	HUA	36
Kerr et al. (2021)— Study 1 Jap	464	TU	IU, CP	Health	HUA	36
Kerr et al. (2021)— Study 1 Kor	467	TU	IU, CP	Health	HUA	36
Kerr et al. (2021)— Study 1 Mex	438	TU	IU, CP	Health	HUA	36
Kerr et al. (2021)— Study 1 Spa	464	TU	IU, CP	Health	HUA	36
Kerr et al. (2021)— Study 1 Swe	453	TU	IU, CP	Health	LUA	36
Kerr et al. (2021)—Study 1 U.K.	453	TU	IU, CP	Health	LUA	36
Kerr et al. (2021)—Study 1 U.S.	482	TU	IU, CP	Health	LUA	36
Kerr et al. (2021)—Study 2	1,538	TU	IU, CP	Health	LUA	45
Kerr et al. (2021)—Study 3	1,007	TU	IU, CP	Environment	LUA	38
Kimmerle et al. (2015)	179	EU	IU	Technology	LUA	709
Kortenkamp and Basten (2015)—Topic 1	164	CU	CP	Environment	LUA	175
Kortenkamp and Basten (2015)—Topic 2	164	CU	CP	Health	LUA	169
Kortenkamp and Basten (2015)—Topic 3	164	CU	CP	Environment	LUA	210
Lin and Kim (2022)	142	CU	A	Technology	HUA	1,046
Longman et al. (2012)	120	TU	CP	Health	LUA	188
Morton et al. (2011)—Study 1	88	TU	BI	Environment	LUA	27
Morton et al. (2011)—Study 2	120	TU	BI	Environment	LUA	28
Nagler et al. (2019)	738	CU	IU, A	Technology	LUA	246
Nakayachi et al. (2018)	750	EU	CP, A, BI	Environment	LUA	54
Ratcliff and Wicke (2023)	502	EU	CP, BI	Health	LUA	451
Ratcliff et al. (2021)	329	EU	IU, A, CP, BI	Technology	LUA	354
Retzbach and Maier (2015)	496	CU	A, CP	Technology	LUA	715
C. R. Schneider et al. (2022)—Study 1	1,294	CU	A, CP	Health	LUA	39
Shi et al. (2023)	734	CU	CP	Health	LUA	246
Sladakovic et al. (2016)	71	TU	A, BI	Health	LUA	120
Steijaert et al. (2021)	88	EU	IU, CP, BI	Health	LUA	192
van der Bles et al. (2020)—Study 1	248	TU	IU, CP	Environment	LUA	37
Wiedemann et al. (2021)—Study 3	120	EU	CP	Technology	LUA	78

Note: Aus = Australia, Chi = China, Fra = France, Ger = Germany, Ita = Italy, Jap = Japan, Kor = Korea, Mex = Mexico, Spa = Spain; N = sample size, CU = consensus uncertainty, TU = technical uncertainty, EU = epistemic uncertainty, IU = internal uncertainty, A = attitudes, CP = credibility perception, BI = behavioral intention, LUA = low uncertainty avoidance, HUA = high uncertainty avoidance, N/A = not available.

Table 3

Moderator (Subgroup) Analysis.

Moderator	Moderator statistics			95% CI	Between-study heterogeneity
Moderator	Outcome	Subgroup	ES(k)	95% CI	Q-test	p
Topic domain	Uncertainty	Environment	0.17(6)	[-0.09, 0.42]	0.87	.65
Topic domain		Health	0.29(29)	[0.18, 0.40]
		Technology	0.23(6)	[0.01, 0.45]
	Attitudes	Environment	−0.13(5)	[−0.26, −0.00]	4.9	.09
		Health	−0.17(10)	[−0.28, −0.06]
		Technology	0.12(4)	[−0.11, 0.34]
	Credibility	Environment	0.06(14)	[−0.01, 0.14]	8.83*	.01
		Health	−0.08(29)	[−0.14, −0.01]
		Technology	0.05(5)	[−0.07, 0.17]
	Intention	Environment	0.00(4)	[−0.14, 0.14]	2.57	.28
		Health	−0.12(13)	[−0.22, 0.02]
		Technology	−0.01(3)	[−0.16, 0.14]
Cultural uncertainty avoidance	Uncertainty	HUA	0.20(8)	[0.03, 0.36]	0.72	.40
		LUA	0.28(33)	[0.18, 0.38]	0.72	.40
	Attitudes	HUA	−0.09(3)	[−0.65, 0.46]	<0.01	.98
		LUA	−0.10(16)	[−0.17, −0.03]	<0.01	.98
	Credibility	HUA	−0.15(9)	[−0.24, −0.06]	9.29**	< .01
		LUA	0.01(39)	[−0.04, 0.07]	9.29**	< .01
	Intention	HUA	−0.53(2)	[−1.29, 0.23]	1.66	.20
		LUA	−0.03(18)	[−0.08, 0.03]	1.66	.20

Note. ES = estimated effect size, k = number of studies.

p < .05. **p < .01. ***p < .001.

Weighted Effect Sizes by Comparison

RQ1 explored the overall effect size of different uncertainty types on perception. The random-effects model showed that consensus uncertainty had a medium summary effect (d = 0.41, k = 17, 95% confidence interval [CI] = [0.26, 0.56], p < .01); epistemic uncertainty had a modest overall effect (d = 0.29, k = 9, 95% CI = [0.00, 0.59], p = .05), and technical uncertainty had a weak overall effect (d = 0.13, k = 15, 95% CI = [0.05, 0.21], p < .01). RQ2 found that epistemic uncertainty had a small but significant summary effect on credibility perception, with a standardized mean difference of d = 0.09 (k = 13, 95% CI = [0.00, 0.17], p = .05). In contrast, the overall effect of consensus uncertainty (d = −0.08, k = 12, 95% CI = [−0.19, 0.02], p = .12) and technical uncertainty (d = −0.06, k = 23, 95% CI = [−0.13, 0.02], p = .12) were not significant.

RQ3 assessed the overall effects of different types of uncertainty on attitudes. The results showed that both consensus uncertainty (d = −0.12, k = 10, 95% CI = [−0.25, −0.00], p = .05) and technical uncertainty (d = −0.15, k = 3, 95% CI = [−0.24, −0.07], p < .01) had a marginal but significant overall effect on attitudes. On the contrary, the overall effect of epistemic uncertainty on attitudes was nonsignificant (d = −0.05, k = 6, 95% CI = [−0.21, 0.10], p = .50). Interestingly, RQ4 revealed that there was no significant overall impact of any uncertainty type on behavioral intentions: consensus uncertainty (d = −0.12, k = 11, 95% CI = [−0.24, 0.00], p = .06), epistemic uncertainty (d = −0.07, k = 6, 95% CI = [−0.18, 0.05], p = .24), and technical uncertainty (d = 0.04, k = 3, 95% CI = [−0.19, 0.28], p = .73).

Potential Moderators

To answer the RQ5, the results suggested that only text length significantly moderated the effect (Q_between = 4.39, p = .04), with longer texts reducing uncertainty perception (b = −0.0003, p = .04). Topic domain (Q_between = 0.87, p = .65) and cultural uncertainty avoidance (Q_between = 0.72, p = .40) did not moderate the effect. RQ6 found that topic domain significantly moderated the impact of scientific uncertainty on credibility perception (Q_between = 8.83, p = .01). Specifically, scientific uncertainty was perceived as less credible in the health domain (d = −0.08, p = .02), but not for technology domain (d = 0.05, p = .39) or environment domain (d = 0.06, p = .11). Cultural uncertainty avoidance was also indicated as a significant moderator for credibility perceptions (Q_between = 9.29, p < .01), with a negative effect in high-avoidance countries (d = −0.15, p < .01), and no effect in low avoidance countries (d = 0.01, p = .70). However, text length had no significant moderating effect (Q_between = 1.82, p = .18).

RQ7 explored moderators of the effect of scientific uncertainty on attitudes. Text length significantly moderated this effect (Q_between = 7.63, p < .01). The longer text yielded a marginally positive impact on attitudes (b = 0.0004, p < .01). Neither topic domain (Q_between = 4.9, p = .09) nor cultural uncertainty avoidance (Q_between < 0.01, p = .98) had significant moderating effects. RQ8 investigated potential moderators of the effect of scientific uncertainty on behavioral intention. The analysis showed that topic domain (Q_between = 2.57, p = .28), cultural uncertainty avoidance (Q_between = 1.66, p = .20), and text length (Q_between = 2.60, p = .11) did not significantly moderate this effect.

Publication Bias

We employed two methods to test for publication bias (Du et al., 2017). As shown in the funnel plot (Figure 2), most studies were symmetrically distributed around the center of the summary effect size, indicating the absence of publication bias. Since the subjectivity of interpreting funnel plots, we conducted Egger’s regression tests to assess publication bias (Egger et al., 1997). The Egger’s test was nonsignificant (t = 0.11, p = .91). Overall, we did not find any evidence of publication bias.

Figure 2

Funnel Plot of Effect Sizes to Check Publication Bias.

Discussion

Disclosing uncertainty is an integral part of scientific communication with the public. However, there remains a lack of consensus regarding its communicative impact. The current study employed a meta-analytic approach to assess how three types of scientific uncertainty affect communication and to identify factors influencing these effects. Our findings suggest a nuanced and complex pattern underlying the communicative process, indicating that further efforts should be made to investigate the underlying mechanisms and contextual factors.

Effects on Uncertainty

Our study found that all three types of uncertainty had modest yet significant effects on individuals’ perceptions. It is notable that the effects on perceived uncertainty were small. From a methodological perspective, two primary factors warrant consideration. First, when embedded within complex scientific contexts, experimental manipulations of uncertainty may prove insufficiently salient. The intricate nature of scientific topics can hinder recipients’ ability to fully process uncertain information, potentially overshadowing the intended effects of uncertainty manipulation. For another, we propose that measurement-related issues may contribute to the observed modest effect size. Specifically, the psychometric instruments employed to assess perceived uncertainty may lack the sensitivity to detect subtle fluctuations in perception. Alternatively, these measures may reflect individuals’ long-term dispositions rather than their immediate responses to uncertainty (e.g., “The effect of overprescribing antibiotics is uncertain”). Therefore, future studies can make efforts in the following directions: for the manipulation messages, it is suggested to reduce the complexity of the scientific information to allow uncertainty messages to be more salient and impactful; for the measures, items should explicitly measure people’s uncertainty perception toward the topic described in the message, rather than toward the subject or science. In addition, future studies may consider implementing a pre-and post-test design to examine the genuine effects triggered by uncertainty manipulations (Bolger et al., 2019). In our observation, only a limited number of studies employed this design (e.g., Retzbach & Maier, 2015).

Theoretically, the nuanced effect of uncertainty manipulations on perceived uncertainty can be understood through two possible explanations. First, this may indicate that the primary impact of such communications does not necessarily manifest as heightened uncertainty perceptions per se. Instead, it is plausible that disclosing uncertainty in scientific messages primarily increases perceptions of transparency and honesty, ultimately fostering increased trust in the scientific process and its communicators. That is, by openly acknowledging the inherent uncertainties, communicators may be perceived as more credible and trustworthy (Bromme & Goldman, 2014). Following this line of reasoning, perceived uncertainty may not serve as an optimal manipulation check but rather as a theoretically relevant construct (O’Keefe, 2003). Indeed, several scholars have attempted to establish relationships between uncertainty manipulation and various psychological variables, including perceived objectivity, critical reflection, and perceived reactance (e.g., Adams et al., 2024; Kalny & Walter, 2024; Ratcliff & Wicke, 2023). Furthermore, this may suggest potential differences in the cognitive mechanisms perceived by individuals across different types of uncertainty, indicating the need for a deeper understanding. We will elaborate on this in the theoretical implications section.

Effects on Credibility

Regarding credibility perception, our findings indicate that communicating both consensus and technical uncertainty have non-significant impact on credibility perception, which appears to differ from the conclusions drawn by Gustafson and Rice (2020). Those discrepancies can be attributed to the distinct corpus of studies in our meta-analysis, with our analysis adopting more stringent inclusion criteria due to methodological considerations. Moreover, our analysis incorporated recent studies that capture emerging trends in the field, many of which are contextualized within the COVID-19 pandemic (e.g., Hendriks et al., 2023; Kerr et al., 2021). This context may explain the observed differences in results, as these studies address the unique challenges posed by the global health crisis. Another possible explanation is that we used a different methodology to examine the existing literature. While narrative reviews aim to outline general trends based on notable studies, meta-analyses provide a quantitative synthesis that may enhance the ability to detect overall effects. By employing a random-effects model, our meta-analysis considers both sample size and variability between studies, which could contribute to findings with greater external validity.

For epistemic uncertainty, Gustafson and Rice (2020) noted that drawing conclusions is challenging due to the limited number of studies and inconclusive results. In our study, we found communicating epistemic uncertainty had a relatively small positive effect on credibility perception. This intriguing finding of epistemic uncertainty can probably be understood via the lens of attribution theory, which posits that individuals interpret behaviors based on perceived causes and subsequently adjust their responses accordingly (Crowley & Hoyer, 1994). In this light, epistemic uncertainty, rooted in uncontrollable external factors such as complexity in the natural world, may lead individuals to attribute uncertainty to external circumstances rather than scientists themselves. Consequently, individuals may easily acknowledge uncertainty as an inherent and unalterable limitation (Maxim & Mansier, 2014; Weiner, 1985). Furthermore, they may appreciate scientists’ willingness to engage in open conversations with the public, thereby attributing credit to their transparency.

Effects on Attitude

Compared with credibility perception, the effects on attitudes showed a relatively consistent pattern: while communicating epistemic uncertainty yielded a null effect, consensus and technical uncertainty had a small but negative effect. Lay epistemology theory may shed light on explaining this finding. This theory posits that certainty, consistency with prior attitudes, and accuracy serve as crucial cognitive factors guiding individuals’ information processing (Kruglanski, 1989; Landau et al., 2014). In contrast, consensus uncertainty, characterized by conflicting opinions, inherently contradicts the principle of consistency. Similarly, technical uncertainty, often presented as range estimations, contravenes lay expectations of precision. Consequently, both types of uncertainty may present cognitive challenges, potentially diminishing individuals’ motivation for deeper comprehension. This lack of clarity may engender frustration and disengagement, thereby reinforcing negative attitudes toward the presented information. Therefore, future studies can explore strategies to mitigate these negative effects. One potential way is to employ multimodal methods to facilitate comprehension and decrease backlash attitudes, particularly in the context of technical uncertainty (Padilla et al., 2021).

Effects on Behavioral Intention

It is noteworthy that despite the various impacts on credibility perception and attitude, these impacts did not further translate into behavioral intentions. One explanation is that behavior is influenced by a myriad of factors, among which uncertainty stands out as a significant one, although not always the decisive factor (Sheeran, 2002). A typical mind-set that would affect people’s decisions is the “better safe than sorry” principle (Aven, 2023; Pielke, 2002). For instance, despite individuals appreciating scientists’ transparent communication about the uncertainties of GMO foods—potentially enhancing their attitudes—they may still opt for what they perceive as the safer choice. This reflects their preference for perceived safety in the face of disclosed uncertainties. In this light, we encourage scholars to integrate insights from behavioral science and theories in specific subjects (e.g., Theory of Planned Behavior, Ajzen, 1991; Transtheoretical Model of Behavior Change, Prochaska & DiClemente, 1983). By integrating these behavioral theories with uncertainty communication research, we can develop more effective approaches to science communication that not only enhance credibility but also facilitate desired behavioral outcomes.

Moderators From Contextual, Cultural, and Message Levels

Our study identified the topic domain as a moderator such that uncertainty negatively affects credibility perception within the health topic domain but not for the environment and technology topic domain. This moderation effect may be explained by the construal level theory. According to this theory, people’s thinking and decision-making processes are influenced by their perceived proximity or distance to the subject matter (Trope & Liberman, 2010). Specifically, when considering the health domain, individuals tend to engage in detailed and sensory-oriented thinking due to the personal and self-relevant nature of such issues (Kim, 2019). Consequently, they may rely heavily on personal experiences while adopting a skeptical stance toward scientific research and expert opinions. In contrast, many individuals perceive environment or technology topic issues as psychologically distant (Tang & Chooi, 2023), prioritizing long-term and overarching goals. This “high-level construal” may lead individuals to place more trust in information provided by professionals.

In addition, cultural uncertainty avoidance emerges as another moderator shaping the communicating impact of scientific uncertainty on credibility perception. Specifically, compared to low uncertainty avoidance cultures, those characterized by high cultural uncertainty avoidance exhibit lower trust in uncertainty communication. This finding echoed a recent study, which highlighted the negative relationship between trust and uncertainty avoidance (Küçükkömürler & Özkan, 2022). However, this result should be interpreted with caution, since the effect size is relatively small, and most of the samples tested under high uncertainty avoidance culture come from a multicountry study (Kerr et al., 2021). Hence, future studies can examine the effect of uncertainty communication in more diverse cultures.

As we predicted, text length plays a moderating role in both uncertainty perception and attitudes. Specifically, text length showed a marginally negative relationship with uncertainty perception and positively related to attitudes. This result validates that verbosity can serve as an uncertainty reduction strategy, and individuals could be informed better about scientific findings (Berger & Calabrese, 1975; Flanagin, 2007). Nevertheless, the conclusion should be understood with care, due to the negligible effect size and potential for confounding with other factors. Therefore, we hope future research could further explore this effect by the following approach: first, experimental studies could examine the interaction with individual characteristics (e.g., perceived information overload, Tandoc et al., 2024) and second, future meta-analyses could conduct a more detailed operationalization, such as calculating the proportion of uncertainty text within the full text or identifying the specific positions where uncertainty text appears (e.g., Kalny & Walter, 2024).

Theoretical and Practical Implications

Building upon our findings, we developed a comprehensive theoretical framework to explain the observed patterns in the impact of various types of uncertainty in science communication. While we fully acknowledge the validity and utility of the current uncertainty typology predicated on its sources, we argue for a more profound investigation into the public’s interpretation of these sources. To bridge this gap, we propose to apply attribution theory (Weiner, 1972), with its focus on locus of control, stability, and controllability dimensions, to gain insights into the essential factors that shape the public’s interpretations of scientific uncertainty. Specifically, attribution theory postulates that individuals attempt to understand and explain events or behaviors by attributing causes to them along three dimensions: (a) locus of control (whether the cause is perceived as within the person [a.k.a., internal] or in the environment [a.k.a., external]); (b) stability (whether the cause is seen as constant or variable over time); and (c) controllability (the degree to which the cause is perceived to be under an individual’s control).

Our meta-analysis revealed a discernible pattern wherein epistemic uncertainty tends to positively affect credibility judgment, while consensus and technical uncertainty appear to attenuate individuals’ attitudes. We posit that this differential impact may be rooted in divergent attributions of these uncertainty sources. Our preliminary hypothesis suggests that epistemic uncertainty might be predominantly attributed to external, unstable, and uncontrollable factors (e.g., the inherent complexity and dynamic nature of the universe). Conversely, technical uncertainty might be construed as stemming from internal, stable, and potentially controllable factors (e.g., current methodological limitations that could be surmounted through technological advancements). Similarly, consensus uncertainty might also receive such characteristics (e.g., disagreements among experts that could potentially be resolved through further research or debate). Thus, these disparate attributional patterns may engender contrasting perspectives and expectations toward these two categories of scientific uncertainty. Figure 3 presents our tentative conceptualization of the attributional characteristics for each type of uncertainty along the three dimensions. It is important to note that these propositions represent our very preliminary interpretations and require rigorous empirical validation through methodologies such as in-depth interviews and large-scale surveys. Should this conceptual speculation withstand empirical examination, we believe that this theoretical framework provides a more profound foundation and a coherent structure for understanding the cognitive processes underlying public interpretations of scientific uncertainty, potentially stimulating new lines of inquiry and theoretical development in scientific uncertainty.

Figure 3.

Preliminary Attribution Analysis of Each Uncertainty Type.

Our meta-analysis also offers some practical insights: First, educators and science communicators should prioritize teaching the nature of science, emphasizing its tentative, evolving, and adaptive nature within an iterative process of theory and empirical evidence. This approach can deepen our understanding and foster acceptance of scientific uncertainties. In addition, to address misconceptions about scientific endeavors, collaboration among scientists, journalists, and policymakers is essential. These scientific communicators should articulate the magnitude and categorization of uncertainty in a manner that is both accessible and contextually appropriate.

Limitations and Future Directions

Our study has several limitations that indicate directions for future research. First, due to the limited sample size, our study combined variables “attitudes” and “beliefs.” Future studies with a larger sample could explore them separately. Similarly, we also did not differentiate credibility types; future research should examine if scientific uncertainty affects source credibility and information credibility differently. Second, we recognize the challenge in categorizing topics that straddle health and technology (e.g., health technology). Future studies should develop clearer criteria to differentiate these overlapping contexts. Third, using country-level measures to gauge cultural uncertainty avoidance may oversimplify individual views. As Hofstede (2001) pointed out, cultural dimensions can vary widely within a country, and individuals may not reflect national averages. Our study does not account for these variations or cultural shifts over time (Beugelsdijk et al., 2015). With few studies (n = 3) measuring this at the individual level, we hope future research could directly assess uncertainty avoidance to better gauge its influence. Fourth, our meta-analysis shows significant small effects of uncertainty communication on public trust and engagement. However, these effects may not be practically significant in real-world contexts. Future studies should identify conditions that allow these small effects to accumulate or interact. More research on potential moderators related to audience characteristics (e.g., ideology, numeracy, trust in science) or message traits (e.g., modality) is needed. Fifthly, our moderator analysis combined three types of uncertainty into one predictor for statistical power, which might overlook nuances between uncertainty types. Future meta-analyses should analyze each type separately to capture these distinctions. Finally, despite our thorough search, some relevant studies might have been missed. Future meta-analyses should perform a more interdisciplinary and comprehensive literature review.

Conclusion

To conclude, this meta-analysis represented an initial synthesis of the existing research on the effects of three different types of scientific uncertainty (i.e., epistemic uncertainty, technical uncertainty, and consensus uncertainty). This study contributed to science communication in several ways. First, it refined the typology of scientific uncertainty by offering clear definitions and detailed operationalizations. Second, it presented a quantitative synthesis of the effects of these different types of uncertainty and identified moderators at the contextual, individual, and message levels. Finally, we proposed a comprehensive theoretical framework based on the observed patterns, which necessitates further validation in future research.

Supplemental Material

sj-docx-1-scx-10.1177_10755470251314129 – Supplemental material for A Meta-Analysis Synthesizing the Effects of Three Uncertainty Types in Science Communication

Supplemental material, sj-docx-1-scx-10.1177_10755470251314129 for A Meta-Analysis Synthesizing the Effects of Three Uncertainty Types in Science Communication by Zhuo Guo, Ke Liu and Meng Chen in Science Communication

Footnotes

Acknowledgements

The authors sincerely thank the two anonymous reviewers and Editor, Dr. Kahlor, for their invaluable guidance and support. The authors also extend special gratitude to scholars who generously agreed to provide their data.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Science and Technology Commission of Shanghai Municipality (No. 24DZ2301500).

ORCID iDs

Zhuo Guo

Ke Liu

Meng Chen

Supplemental Material

Supplemental material for this article is available online at

Notes

Author Biographies

Zhuo Guo is a master’s student in the School of Media and Communication at Shanghai Jiao Tong University, China. Her research interests focus on the communication of uncertain scientific findings and the correction of scientific misinformation.

Ke Liu is a master’s student in the School of Media and Communication at Shanghai Jiao Tong University, China. Her research interests are in the areas of information behaviors, health/risk communication, and misinformation correction in the context of social media.

Meng Chen (PhD, University of California, Davis) is an associate professor in the School of Media and Communication at Shanghai Jiao Tong University, China. Her research interests are environmental communication and language effects.

References

*Adams

D. R.

Ratcliff

C. L.

Pokharel

Jensen

J. D.

Liao

(2024). Communicating scientific uncertainty in the early stages of the COVID-19 pandemic: A message experiment. Risk Analysis, 44(7), 1700–1715.

Ajzen

(1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179–211.

*Aklin

Urpelainen

(2014). Perceptions of scientific dissent undermine public support for environmental policy. Environmental Science & Policy, 38, 173–177.

Aven

(2023). A risk and safety science perspective on the precautionary principle. Safety Science, 165, 106211.

Babrow

A. S.

Kasch

C. R.

Ford

L. A.

(1998). The many meanings of uncertainty in illness: Toward a systematic accounting. Health Communication, 10(1), 1–23.

Bartoszek

Ranney

R. M.

Curanovic

Costello

S. J.

Behar

(2022). Intolerance of uncertainty and information-seeking behavior: Experimental manipulation of threat relevance. Behaviour Research and Therapy, 154, 104125.

Baxter

L. A.

Montgomery

B. M.

(1996). Relating: Dialogues and dialectics. Guilford Press.

Baxter

L. A.

Norwood

K. M.

(2015). Relational dialectics theory: Navigating meaning from competing discourses. Engaging Theories in Interpersonal Communication: Multiple Perspectives, 2, 279–292.

Berger

C. R.

Calabrese

R. J.

(1975). Some explorations in initial interactions and beyond: Toward a developmental theory of interpersonal communication. Human Communication Research, 1(2), 99–112. https://doi.org/10.1111/j.1468-2958.1975.tb00258.x

10.

Besley

J. C.

Nisbet

(2013). How scientists view the public, the media and the political process. Public Understanding of Science, 22(6), 644–659.

11.

Beugelsdijk

Maseland

Van Hoorn

(2015). Are scores on Hofstede’s dimensions of national culture stable over time? A cohort analysis. Global Strategy Journal, 5(3), 223–240.

12.

Bolger

Zee

K. S.

Rossignac-Milon

Hassin

R. R.

(2019). Causal processes in psychology are heterogeneous. Journal of Experimental Psychology: General, 148(4), 601–618.

13.

Borenstein

Hedges

L. V.

Higgins

J. P. T.

Rothstein

H. R.

(2010). A basic introduction to fixed-effect and random-effects models for meta-analysis. Research Synthesis Methods, 1(2), 97–111. https://doi.org/10.1002/jrsm.12

14.

Borenstein

Hedges

L. V.

Higgins

J. P. T.

Rothstein

H. R.

(2013). Comprehensive meta-analysis version 3 [Computer software]. Biostat.

15.

Brashers

D. E.

(2001). Communication and uncertainty management. Journal of Communication, 51(3), 477–497.

16.

Bromme

Goldman

S. R.

(2014). The public’s bounded understanding of science. Educational Psychologist, 49(2), 59–69. https://doi.org/10.1080/00461520.2014.921572

17.

*Chang

(2015). Motivated processing: How people perceive news covering novel or contradictory health research findings. Science Communication, 37(5), 602–634.

18.

Chen

Dong

Wang

(2024). A meta-analysis examining the role of character-recipient similarity in narrative persuasion. Communication Research, 51(1), 56–82. https://doi.org/10.1177/00936502231204834

19.

*Chinn

Lane

D. S.

Hart

P. S.

(2018). In consensus we trust? Persuasive effects of scientific consensus communication. Public Understanding of Science, 27(7), 807–823.

20.

*Clarke

C. E.

Dixon

G. N.

Holton

McKeever

B. W.

(2015). Including “evidentiary balance” in news media coverage of vaccine risk. Health Communication, 30(5), 461–472.

21.

Cohen

(1992). Quantitative methods in psychology. Psychological Bulletin, 112, 155–159.

22.

Cook

Lewandowsky

Ecker

U. K.

(2017). Neutralizing misinformation through inoculation: Exposing misleading argumentation techniques reduces their influence. PLOS ONE, 12(5), Article e0175799. https://doi.org/10.1371/journal.pone.0175799

23.

Crowley

A. E.

Hoyer

W. D.

(1994). An integrative framework for understanding two-sided persuasion. Journal of Consumer Research, 20(4), 561–574.

24.

*Dixon

G. N.

Clarke

C. E.

(2013). Heightening uncertainty around certain science: Media coverage, false balance, and the autism-vaccine controversy. Science Communication, 35(3), 358–382.

25.

*Dixon

G. N.

McKeever

B. W.

Holton

A. E.

Clarke

Eosco

(2015). The power of a picture: Overcoming scientific misinformation by communicating weight-of-evidence information with visual exemplars. Journal of Communication, 65(4), 639–659. https://doi.org/10.1111/jcom.12159

26.

Cheng

Zhang

Mei

(2017). The impact of medication adherence on clinical outcomes of coronary artery disease: A meta-analysis. European Journal of Preventive Cardiology, 24(9), 962–970.

27.

*Dunwoody

(1999). Scientists, journalists, and the meaning of uncertainty. In Friedman

Dunwoody

Rogers

(Eds.), Communicating uncertainty: Media coverage of new and controversial science (pp. 59–80). Lawrence Erlbaum.

28.

Dunwoody

Kohl

P. A.

(2017). Using weight-of-experts messaging to communicate accurately about contested science. Science Communication, 39(3), 338–357.

29.

Ebeling

M. F.

(2008). Mediating uncertainty: Communicating the financial risks of nanotechnologies. Science Communication, 29(3), 335–361. https://doi.org/10.1177/1075547007312068

30.

Egger

Smith

G. D.

Schneider

Minder

(1997). Papers bias in meta-analysis detected by a simple, graphical test. British Medical Journal, 315, 629–631.

31.

Ersoy

Mahmood

Sharif

Ersoy

Ehtiyar

(2023). Exploring the associations between social support, perceived uncertainty, job stress, and emotional exhaustion during the COVID-19 crisis. Sustainability, 15(3), 2150.

32.

Eveland

W. P.

Jr. Hively

M. H.

(2009). Political discussion frequency, network size, and “heterogeneity” of discussion as predictors of political knowledge and participation. Journal of Communication, 59(2), 205–224.

33.

Fischhoff

(2012). Communicating uncertainty fulfilling the duty to inform. Issues in Science and Technology, 28(4), 63–70. https://www.jstor.org/stable/43315647

34.

Flanagin

A. J.

(2007). Commercial markets as communication markets: Uncertainty reduction through mediated information exchange in online auctions. New Media & Society, 9, 401–423.

35.

*Flemming

Kimmerle

Cress

Sinatra

G. M.

(2020). Research is tentative, but that’s okay: Overcoming misconceptions about scientific tentativeness through refutation texts. Discourse Processes, 57(1), 17–35.

36.

Frewer

L. J.

Miles

Brennan

Kuznesof

Ness

Ritson

(2002). Public preferences for informed choice under conditions of risk uncertainty. Public Understanding of Science, 11(4), 363–372.

37.

Friedman

S. M.

Dunwoody

Rogers

C. L.

(Eds.). (1999). Communicating uncertainty: Media coverage of new and controversial science. Lawrence Erlbaum Associates.

38.

Gustafson

Rice

R. E.

(2019). The effects of uncertainty frames in three science communication topics. Science Communication, 41(6), 679–706. https://doi.org/10.1177/1075547019870811

39.

Gustafson

Rice

R. E.

(2020). A review of the effects of uncertainty in public science communication. Public Understanding of Science, 29(6), 614–633. https://doi.org/10.1177/0963662520942122

40.

*Haigh

Birch

H. A.

(2021). When “scientists say” coffee is good for you one day and bad for you the next: Do generic attributions to “scientists” and “experts” amplify perceived conflict? Collabra: Psychology, 7(1), 23447.

41.

*Han

P. K.

Zikmund-Fisher

B. J.

Duarte

C. W.

Knaus

Black

Scherer

A. M.

Fagerlin

(2018). Communication of scientific uncertainty about a novel pandemic health threat: Ambiguity aversion and its mechanisms. Journal of Health Communication, 23(5), 435–444.

42.

*Hendriks

Janssen

Jucks

(2023). Balance as credibility? How presenting one-vs. two-sided messages affects ratings of scientists’ and politicians’ trustworthiness. Health Communication, 38(12), 2757–2764.

43.

*Hinnant

Hong

Young

(2023). Contested certainty and credibility: The effect of personal stories and scientific evidence in user comments on news story evaluation and relevance. Science Communication, 45(1), 65–94.

44.

Hofstede

(1983). National cultures in four dimensions: A research-based theory of cultural differences among nations. International Studies of Management & Organization, 13(1–2), 46–74.

45.

Hofstede

(2001). Culture’s consequences: Comparing values, behaviors, institutions and organizations across nations. Sage.

46.

Hofstede

(2013). Culture’s consequences: Comparing values, behaviors, institutions, and organizations across nations (2nd ed.). Sage.

47.

*Howe

L. C.

MacInnis

Krosnick

J. A.

Markowitz

E. M.

Socolow

(2019). Acknowledging uncertainty impacts public acceptance of climate scientists’ predictions. Nature Climate Change, 9(11), 863–867. https://doi.org/10.1038/s41558-019-0587-5

48.

*Hendriks

Jucks

(2020). Does scientific uncertainty in news articles affect readers’ trust and decision-making? Media and Communication, 8(2), 401–412.

49.

Ihlen

Ø.

Just

S. N.

Kjeldsen

J. E.

Mølster

Offerdal

T. S.

Rasmussen

Skogerbø

. (2022). Transparency beyond information disclosure: Strategies of the Scandinavian public health authorities during the COVID-19 pandemic. Journal of Risk Research, 25(10), 1176–1189. https://doi.org/10.1080/13669877.2022.2077416

50.

*Jensen

J. D.

(2008). Scientific uncertainty in news coverage of cancer research: Effects of hedging on scientists’ and journalists’ credibility. Human Communication Research, 34(3), 347–369.

51.

*Jensen

J. D.

Carcioppolo

King

A. J.

Bernat

J. K.

Davis

Yale

Smith

(2011). Including limitations in news coverage of cancer research: Effects of news hedging on fatalism, medical skepticism, patient trust, and backlash. Journal of Health Communication, 16(5), 486–503.

52.

*Jensen

J. D.

Hurley

R. J.

(2012). Conflicting stories about public scientific controversies: Effects of news convergence and divergence on scientists’ credibility. Public Understanding of Science, 21(6), 689–704.

53.

Jensen

J. D.

Pokharel

Scherr

C. L.

King

A. J.

Brown

Jones

(2017). Communicating uncertain science to the public: How amount and source of uncertainty impact fatalism, backlash, and overload. Risk Analysis, 37(1), 40–51.

54.

Johnson

B. B.

(2018). “Counting votes” in public responses to scientific disputes. Public Understanding of Science, 27(5), 594–610.

55.

Johnson

B. B.

Slovic

(1995). Presenting uncertainty in health risk assessment: Initial studies of its effects on risk perception and trust. Risk Analysis, 15(4), 485–494.

56.

*Joslyn

Demnitz

(2019). Communicating climate change: Probabilistic expressions and concrete events. Weather, Climate, and Society, 11(3), 651–664.

57.

*Joslyn

Demnitz

(2021). Explaining how long CO₂ stays in the atmosphere: Does it change attitudes toward climate change? Journal of Experimental Psychology: Applied, 27(3), 473–484.

58.

*Joslyn

S. L.

LeClerc

J. E.

(2016). Climate projections and uncertainty communication. Topics in Cognitive Science, 8(1), 222–241.

59.

*Jurhill

B. A.

(2022). The effects of personalized health risks and communication of epistemic uncertainty on risk perceptions and trustworthiness in healthcare [Master Thesis]. Tilburg University.

60.

Kahneman

Tversky

(1982). Variants of uncertainty. Cognition, 11(2), 143–157.

61.

*Kalny

Walter

(2024). Tipping the scales of psychological reactance: A closer look at imperative language and the role of epistemic certainty. Science Communication, 46(3), 366–393.

62.

*Kelp

N. C.

Witt

J. K.

Sivakumar

(2022). To vaccinate or not? The role played by uncertainty communication on public understanding and behavior regarding COVID-19. Science Communication, 44(2), 223–239.

63.

Keohane

R. O.

Lane

Oppenheimer

(2014). The ethics of scientific communication under uncertainty. Politics, Philosophy & Economics, 13(4), 343–368.

64.

*Kerr

J. R.

van der Bles

A. M.

Schneider

Dryhurst

Chopurian

Freeman

A. L.

van der Linden

(2021). The effects of communicating uncertainty around statistics on public trust: An international study. medRxiv. https://doi.org/10.1101/2021.09.27.21264202

65.

*Kerr

J. R.

van der Linden

(2022). Communicating expert consensus increases personal support for COVID-19 mitigation policies. Journal of Applied Social Psychology, 52(1), 15–29.

66.

Kim

D. H.

(2019). “How do you feel about a disease?” The effect of psychological distance towards a disease on health communication. International Journal of Advertising, 38(1), 139–153.

67.

*Kimmerle

Flemming

Feinkohl

Cress

(2015). How laypeople understand the tentativeness of medical research news in the media: An experimental study on the perception of information about deep brain stimulation. Science Communication, 37(2), 173–189.

68.

*Kortenkamp

K. V.

Basten

(2015). Environmental science in the media: Effects of opposing viewpoints on risk and uncertainty perceptions. Science Communication, 37(3), 287–313.

69.

Kreps

S. E.

Kriner

D. L.

(2020). Model uncertainty, political contestation, and public trust in science: Evidence from the COVID-19 pandemic. Science Advances, 6(43), Article eabd4563.

70.

Kruglanski

A. W.

(1989). Lay epistemics and human knowledge: Cognitive and motivational bases. Springer.

71.

Küçükkömürler

Özkan

(2022). Political interest across cultures: The role of uncertainty avoidance and trust. International Journal of Intercultural Relations, 91, 88–96.

72.

Kuhn

T. S.

(1970). The structure of scientific revolutions (2nd ed.). University of Chicago Press.

73.

Landau

M. J.

Keefer

L. A.

Rothschild

Z. K.

(2014). Epistemic motives moderate the effect of metaphoric framing on attitudes. Journal of Experimental Social Psychology, 53, 125–138.

74.

Larrimore

Jiang

Larrimore

Markowitz

Gorski

(2011). Peer to peer lending: The relationship between language features, trustworthiness, and persuasion success. Journal of Applied Communication Research, 39(1), 19–37.

75.

Lehmkuhl

Peters

H. P.

(2016). Constructing (un-)certainty: An exploration of journalistic decision-making in the reporting of neuroscience. Public Understanding of Science, 25(8), 909–926. https://doi.org/10.1177/0963662516646047

76.

(2017). Artificial intelligence with uncertainty. CRC Press.

77.

*Lin

Kim

(2022). The effects of gain-loss framing and message sidedness on GMO acceptance: The mediating role of psychological reactance. Asian Communication Research, 19(2), 70–91.

78.

Lohse

(2023). Mapping uncertainty in precision medicine: A systematic scoping review. Journal of Evaluation in Clinical Practice, 29(3), 554–564. https://doi.org/10.1111/jep.13789

79.

*Longman

Turner

R. M.

King

McCaffery

K. J.

(2012). The effects of communicating uncertainty in quantitative health risk estimates. Patient Education and Counseling, 89(2), 252–259.

80.

Maier

Milde

Post

Günther

Ruhrmann

Barkela

(2016). Communicating scientific evidence: Scientists,’ journalists’ and audiences’ expectations and evaluations regarding the representation of scientific uncertainty. Communications, 41(3), 239–264.

81.

Maxim

Mansier

(2014). How is scientific credibility affected by communicating uncertainty? The case of endocrine disrupter effects on male fertility. Human and Ecological Risk Assessment: An International Journal, 20(1), 201–223.

82.

*Morton

T. A.

Rabinovich

Marshall

Bretschneider

(2011). The future that may (or may not) come: How framing changes responses to uncertainty in climate change communications. Global Environmental Change, 21(1), 103–109.

83.

*Nagler

R. H.

Yzer

M. C.

Rothman

A. J.

(2019). Effects of media exposure to conflicting information about mammography: Results from a population-based survey experiment. Annals of Behavioral Medicine, 53(10), 896–908.

84.

*Nakayachi

Johnson

B. B.

Koketsu

(2018). Effects of acknowledging uncertainty about earthquake risk estimates on San Francisco bay area residents’ beliefs, attitudes, and intentions. Risk Analysis, 38(4), 666–679.

85.

Nir

(2011). Motivated reasoning and public opinion perception. Public Opinion Quarterly, 75(3), 504–532.

86.

O’Keefe

D. J.

(2003). Message properties, mediating states, and manipulation checks: Claims, evidence, and data analysis in experimental persuasive message effects research. Communication Theory, 13(3), 251–274.

87.

Osman

Heath

A. J.

Löfstedt

(2018). The problems of increasing transparency on uncertainty. Public Understanding of Science, 27(2), 131–138. https://doi.org/10.1177/0963662517711058

88.

Padilla

L. M.

Powell

Kay

Hullman

(2021). Uncertain about uncertainty: How qualitative expressions of forecaster confidence impact decision-making with uncertainty visualizations. Frontiers in Psychology, 11, Article 579267.

89.

Petty

R. E.

Cacioppo

J. T.

Petty

R. E.

Cacioppo

J. T.

(1986). The elaboration likelihood model of persuasion (pp. 1–24). Springer.

90.

Pew Research Center. (2023). Americans’ trust in scientists, positive views of science continue to decline. https://www.pewresearch.org/science/2023/11/14/americans-trust-in-scientists-positive-views-of-science-continue-to-decline/

91.

Pielke

(2002). Better safe than sorry. Nature, 419, 433–434. https://doi.org/10.1038/419433a

92.

Pollack

H. N.

(2003). Uncertain science. . . Uncertain world. Cambridge University Press.

93.

Post

Maier

(2016). Stakeholders’ rationales for representing uncertainties of biotechnological research. Public Understanding of Science, 25(8), 944–960.

94.

Prochaska

J. O.

DiClemente

C. C.

(1983). Stages and processes of self-change of smoking: Toward an integrative model of change. Journal of Consulting and Clinical Psychology, 51(3), 390–395.

95.

Ratcliff

C. L.

(2021). Communicating scientific uncertainty across the dissemination trajectory: A precision medicine case study. Science Communication, 43(5), 597–623.

96.

Ratcliff

C. L.

Harvill

Wicke

(2023). Understanding public preferences for learning about uncertain science: measurement and individual difference correlates. Frontiers in Communication, 8, 1245786.

97.

Ratcliff

C. L.

Wicke

Harvill

(2022). Communicating uncertainty to the public during the COVID-19 pandemic: A scoping review of the literature. Annals of the International Communication Association, 46(4), 260–289. https://doi.org/10.1080/23808985.2022.2085136

98.

Ratcliff

C. L.

Sun

(2020). Overcoming resistance through narratives: Findings from a meta-analytic review. Human Communication Research, 46(4), 412–443.

99.

*Ratcliff

C. L.

Wicke

(2023). How the public evaluates media representations of uncertain science: An integrated explanatory framework. Public Understanding of Science, 32(4), 410–427.

100.

*Ratcliff

C. L.

Wong

Jensen

J. D.

Kaphingst

K. A.

(2021). The impact of communicating uncertainty on public responses to precision medicine research. Annals of Behavioral Medicine, 55(11), 1048–1061.

101.

*Retzbach

Maier

(2015). Communicating scientific uncertainty: Media effects on public engagement with science. Communication Research, 42(3), 429–456. https://doi.org/10.1177/0093650214534967

102.

Rice

R. E.

Gustafson

Hoffman

(2018). Frequent but accurate: A closer look at uncertainty and opinion divergence in climate change print news. Environmental Communication, 12(3), 301–321.

103.

Ruhrmann

Guenther

Kessler

S. H.

Milde

(2015). Frames of scientific evidence: How journalists represent the (un)certainty of molecular medicine in science television programs. Public Understanding of Science, 24(6), 681–696.

104.

Schmidt

F. L.

Hunter

J. E.

(1999). Comparison of three meta-analysis methods revisited: An analysis of Johnson, Mullen, and Salas (1995). Journal of Applied Psychology, 84, 144–148. https://doi.org/10.1037/0021-9010.84.1.144

105.

*Schneider

C. R.

Freeman

A. L.

Spiegelhalter

van der Linden

(2022). The effects of communicating scientific uncertainty on trust and decision making in a public health context. Judgment and Decision Making, 17(4), 849–882.

106.

Schneider

(2010). Making space for the “nuances of truth”: Communication and uncertainty at an environmental journalists’ workshop. Science Communication, 32(2), 171–201.

107.

Sheeran

(2002). Intention—Behavior relations: A conceptual and empirical review. European Review of Social Psychology, 12(1), 1–36.

108.

Shen

Sheer

V. C.

(2015). Impact of narratives on persuasion in health communication: A meta-analysis. Journal of Advertising, 44(2), 105–113.

109.

*Shi

Rothman

A. J.

Yzer

M. C.

Nagler

R. H.

(2023). Effects of exposure to conflicting information about mammography on cancer information overload, perceived scientists’ credibility, and perceived journalists’ credibility. Health Communication, 38(11), 2481–2490.

110.

Simis

(2013). Framing uncertainty: A case for purposefully using frames in science communication. In Goodwin

Dahlstrom

M. F.

Priest

(Eds.), Ethical issues in science communication: A theory-based approach proceedings of the Iowa State University summer symposia on science communication, book 3 (p. 265). University of Chicago Press.

111.

*Sladakovic

Jansen

Hersch

Turner

McCaffery

(2016). The differential effects of presenting uncertainty around benefits and harms on treatment decision making. Patient Education and Counseling, 99(6), 974–980.

112.

Spiegelhalter

D. J.

Riesch

(2011). Don’t know, can’t know: Embracing deeper uncertainties when analysing risks. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 369(1956), 4730–4750.

113.

Star

S. L.

(1985). Scientific work and uncertainty. Social Studies of Science, 15(3), 391–427. https://doi.org/10.1177/030631285015003001

114.

*Steijaert

M. J.

Schaap

Riet

J. V. T.

(2021). Two-sided science: Communicating scientific uncertainty increases trust in scientists and donation intention by decreasing attribution of communicator bias. Communications, 46(2), 297–316.

115.

Tandoc

E. C.

Jr. Lee

J. C. B.

Lee

Quek

P. J.

(2024). Does length matter? The impact of fact-check length in reducing COVID-19 vaccine misinformation. Mass Communication and Society, 27, 679–709.

116.

Tang

Y. T.

Chooi

W. T.

(2023). From concern to action: The role of psychological distance in attitude towards environmental issues. Current Psychology, 42(30), 26570–26586.

117.

Thorp

H. H.

(2023). Correction is courageous. Science, 382(6672), 743–743.

118.

Trope

Liberman

(2010). Construal-level theory of psychological distance. Psychological Review, 117(2), 440–463.

119.

*van der Bles

A. M.

van der Linden

Freeman

A. L.

Mitchell

Galvao

A. B.

Zaval

Spiegelhalter

D. J

. (2019). Communicating uncertainty about facts, numbers and science. Royal Society Open Science, 6(5), 181870.

120.

van der Bles

A. M.

van der Linden

Freeman

A. L.

Spiegelhalter

D. J

. (2020). The effects of communicating uncertainty on public trust in facts and numbers. Proceedings of the National Academy of Sciences, 117(14), 7672–7683. https://doi.org/10.1073/pnas.1913678117

121.

Wang

Thier

Lee

Nan

(2024). Persuasive effects of temporal framing in health messaging: A meta-analysis. Health Communication, 39, 563–576.

122.

Watts

L. L.

Steele

L. M.

Den Hartog

D. N.

(2020). Uncertainty avoidance moderates the relationship between transformational leadership and innovation: A meta-analysis. Journal of International Business Studies, 51(1), 138–145.

123.

Weiner

(1972). Attribution theory, achievement motivation, and the educational process. Review of Educational Research, 42(2), 203–215.

124.

Weiner

(1985). An attributional theory of achievement motivation and emotion. Psychological Review, 92(4), 548–573.

125.

*Wiedemann

Boerner

F. U.

Freudenstein

(2021). Effects of communicating uncertainty descriptions in hazard identification, risk characterization, and risk protection. PLOS ONE, 16(7), Article e0253762.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.32 MB