Effects of communicating uncertainty descriptions in hazard identification, risk characterization, and risk protection

2021 · Mixed · Moderate evidence

← Research stream

MixedModerate evidencePaper 1105

Study snapshot

Publicationmanual · 2021

Study typerandomized_trial

Reported directionMixed

Evidence ratingModerate

Abstract

Effects of communicating uncertainty descriptions in hazard identification, risk characterization, and risk protection Peter Wiedemann, Franziska U Boerner, Frederik Freudenstein. Effects of communicating uncertainty descriptions in hazard identification, risk characterization, and risk protection. PLoS One. 2021 Jul 13;16(7):e0253762. doi: 10.1371/journal.pone.0253762. Abstract Uncertainty is a crucial issue for any risk assessment. Consequently, it also poses crucial challenges for risk communications. Many guidebooks advise reporting uncertainties in risk assessments, expecting that the audience will appreciate this disclosure. However, the empirical evidence about the effects of uncertainty reporting is sparse and inconclusive. Therefore, based on examples of potential health risks of electromagnetic fields (EMF), three experiments were conducted analysing the effects of communicating uncertainties separately for hazard identification, risk characterisation and risk protection. The setups aimed to explore how reporting and how explaining of uncertainty affects dependent variables such as risk perception, perceived competence of the risk assessors, and trust in risk management. Each of the three experiments used a 2x2 design with a first factor presenting uncertainty descriptions (as used in public controversies on EMF related health effects) or describing a certainty conditions; and a second factor explaining the causes of uncertainties (by pointing at knowledge gaps) or not explaining them. The study results indicate that qualitative uncertainty descriptions regarding hazard identification reduce the confidence in the professional competencies of the assessors. In contrast, a quantitative uncertainty description in risk characterisation-regarding the magnitude of the risk-does not affect any of the dependent variables. Concerning risk protection, trust in exposure limit values is not affected by qualitative uncertainty information. However, the qualitative description of uncertainty regarding the adequacy of protection amplifies fears. Furthermore, explaining this uncertainty results in lower text understandability. Conclusions Our experiments provide several implications for reporting uncertainties. They show that informing about uncertainties is not always a means for improving trust and credibility. Science skeptics and activists with a particular political agenda might instrumentalize uncertainty for spreading distrust in science, primarily when uncertainty refers to hazard identification. Risk communicators should be aware that admitting uncertainty is a double-edged sword. Laypersons might attribute it to the lack of professional expertise. Further research should focus on the issues of cognitive resources and prior beliefs. It should focus on how the recipients interpret the specific social context in which the information is given and how the recipients’ cognitive resources, motivations to process information, and their prior beliefs influence uncertainty information interpretation. These issues could be tackled by dual-process theories developed in judgment and decision- making [53]. Recipients might understand the same information differently depending on whether the information is processed by elaborative or heuristic cognition. Furthermore, prior beliefs shape the interpretation of uncertainties in risk assessment. This issue can be tackled by theories that explain how motivated reasoning influences judgments and opinions about uncertainty information [54]. Finally, we would like to stress the importance of the performative goal of informing about uncertainty. It makes a massive difference whether uncertainty is used to raise fears and distrust in science or honestly report the limits of evidence-based knowledge. Further research should pay more attention to this subject. New studies could learn from the analysis of the politicization of science [16] that has focused on how uncertainty is exploited for undermining trust in science. Open access paper: journals.plos.org

Machine Enhanced Logic

AI evidence extraction

Structured fields below were generated from the paper’s abstract and metadata to make the evidence easier to inspect. Automated extraction may be incomplete or incorrect; verify important details against the original paper.

Study typerandomized_trial

Effect directionmixed

PopulationNot supplied

Sample sizeNot supplied

ExposureNot supplied

Evidence strengthModerateConfidence: 74% · Peer reviewed: yes

Main findings

Using examples of potential health risks of EMF, three 2x2 experiments tested effects of uncertainty descriptions (vs certainty) and explaining causes of uncertainty (vs not) across hazard identification, risk characterization, and risk protection. Qualitative uncertainty descriptions for hazard identification reduced confidence in assessors' professional competence; quantitative uncertainty in risk characterization did not affect measured outcomes; for risk protection, qualitative uncertainty about adequacy of protection amplified fears and explaining uncertainty lowered text understandability, while trust in exposure limit values was not affected by qualitative uncertainty information.

Outcomes measured

risk perception
perceived competence of risk assessors
trust in risk management
trust in exposure limit values
fears
text understandability

Limitations

Empirical evidence described as sparse and inconclusive (contextual statement)
Population characteristics not reported in provided abstract
Sample size not reported in provided abstract
EMF used as an example context; no exposure measurement or health outcomes assessed

View raw extracted JSON

{
  "study_type": "randomized_trial",
  "exposure": {
    "band": null,
    "source": null,
    "frequency_mhz": null,
    "sar_wkg": null,
    "duration": null
  },
  "population": null,
  "sample_size": null,
  "outcomes": [
    "risk perception",
    "perceived competence of risk assessors",
    "trust in risk management",
    "trust in exposure limit values",
    "fears",
    "text understandability"
  ],
  "main_findings": "Using examples of potential health risks of EMF, three 2x2 experiments tested effects of uncertainty descriptions (vs certainty) and explaining causes of uncertainty (vs not) across hazard identification, risk characterization, and risk protection. Qualitative uncertainty descriptions for hazard identification reduced confidence in assessors' professional competence; quantitative uncertainty in risk characterization did not affect measured outcomes; for risk protection, qualitative uncertainty about adequacy of protection amplified fears and explaining uncertainty lowered text understandability, while trust in exposure limit values was not affected by qualitative uncertainty information.",
  "effect_direction": "mixed",
  "limitations": [
    "Empirical evidence described as sparse and inconclusive (contextual statement)",
    "Population characteristics not reported in provided abstract",
    "Sample size not reported in provided abstract",
    "EMF used as an example context; no exposure measurement or health outcomes assessed"
  ],
  "evidence_strength": "moderate",
  "confidence": 0.74,
  "peer_reviewed_likely": "yes",
  "keywords": [
    "uncertainty communication",
    "risk communication",
    "hazard identification",
    "risk characterization",
    "risk protection",
    "electromagnetic fields",
    "trust",
    "risk perception",
    "exposure limits"
  ],
  "suggested_hubs": [
    {
      "slug": "who-icnirp",
      "weight": 0.38,
      "reason": "Discusses trust in exposure limit values and risk assessment/communication around EMF, which relates to guideline-setting context."
    }
  ]
}

AI can be wrong. Always verify against the paper.

AI-extracted fields are generated from the abstract and metadata and may be incomplete or incorrect. This material is for informational purposes and is not medical advice.

Identifiers and source

DOI: 10.1371/journal.pone.0253762
PubMed: Not supplied

Read the original paper →

Related papers

← Back to the Research Library