| 英文摘要: | The Intergovernmental Panel on Climate Change has issued guidelines for communicating probabilities with words, but readers all over the world think the words mean something different. All forecasts, including scientific ones, are fraught with uncertainty. How should scientists communicate their uncertainty to the public in a consistent and reliable way? One solution, favoured by the IPCC, is to qualify forecasts with verbal descriptors — from 'exceptionally unlikely' to 'virtually certain' — accompanied by numeric probability ranges that correspond to these individual phrases. However, this solution is less helpful than one might think. Budescu et al.1 report in Nature Climate Change that most readers, in 24 countries around the world, do not understand the verbal phrases as intended despite having the IPCC translations available. For instance, half of them think that a 'very likely' climate change is one that has less than a 70% probability of occurring, instead of a probability beyond 90%, as dictated by the translation standards. Psychologists in the area of judgment and decision making have known for a long time that lay people's use of numeric probabilities often differs from normative requirements2. Even experts hesitate to use numeric estimates for events that cannot be precisely calculated. Instead they prefer verbal phrases, taken from natural language, like 'a good chance', 'not unlikely' or 'almost certain'3. But the familiarity of these phrases does not guarantee that they are understood by everyone in the same way. Studies of how people translate words into numerical values indicate that verbal probability phrases are very vague (at least in a probabilistic sense)4, and that people fail to realize the extent of their vagueness5, 6. This partial awareness might create an illusion of communication, where communication partners think they understand each other while drawing different inferences from the same messages.
Y H LIM / ALAMY
- Budescu, D. V. et al. Nature Clim. Change 4, 508–512 (2014).
- Tversky, A. & Kahneman, D. Science 185, 1124–1131 (1974).
- Erev, I. & Cohen, B. L. Organization. Behav. Hum. Decision Proc. 45, 1–18 (1990).
- Budescu, D. V. & Wallsten, T. S. Psychol. Learn. Motiv. 32, 275–318 (1995).
- Brun, W. & Teigen, K. H. Organization. Behav. Hum. Decision Proc. 41, 390–404 (1988).
- Amer, T., Hackenbrack, K. & Nelson, M. Auditing: J. Prac. Theor. 13, 126–136 (1994).
- Budescu, D. V., Broomell, S. B. & Por, H. H. Psychol. Sci. 20, 299–308 (2009).
- Theil, M. J. Risk Res. 5, 177–186 (2012).
- Teigen, K. H. & Brun, W. Organization. Behav. Hum. Decision Proc. 80, 155–190 (1999).
- Juanchich, M., Teigen, K. H. & Villejoubert, G. Acta Psychologica 135, 267–277 (2010).
- Fløttum, K. & Dahl, T. Fachsprache 3–4, 205–219 (2011).
Download references
Affiliations
-
Karl Halvor Teigen is in the Department of Psychology, University of Oslo, PB 1094 Blindern, NO-0317, Oslo, Norway and at the Simula Research Laboratory, Martin Linges vei 17, 1364 Fornebu, Norway
|