The Influence of Information Delivery on Risk Ranking by Lay People

Abstract

An experiment was conducted in a real environment to test how information delivery affects risk ranking. Another aim was to propose the best format for delivering information. Different people received different types of information about risks in a risk ranking exercise: Group 1 received a descriptive paragraph about the hazards (Format 1); Group 2 added a table with specific information on risk attributes (Format 2); Group 3 added information on the steps taken locally to mitigate the risks (Format 3), and Group 4 received a data table without identifying the hazard (Format 4). Agreement among subjects' rankings within a group and from group to group was used to measure the potential impact of information delivery. Average pair‐wise Spearman correlation was used to compare the level of agreement within each group. Results showed greater consensus in the group using Format 4 than in Formats 1, 2, and 3, with the only significant difference between Format 4 and each one of the others. The results show that the amount of information, and the way it is delivered, may affect how lay people rank risks, but the differences are not statistically significant.

Keywords:

• Risk ranking
• risk communication
• developing country

Acknowledgments

The authors would like to acknowledge the financial support of the Chilean Commission on Science and Technology, who funded this work through Fondecyt Project number 1020501. We thank the referees for their comments, and the participants of a session at SRA 2004 and SRE‐E 2005, where previous versions of this work were presented. We specially thank Dr. Rómulo Chumacero for his advice on statistical methods. Finally we thank the residents of Pudahuel, who kindly participated in the exercise. Any remaining errors are our sole responsibility.

Notes

1. NG are one of the basic organizations where people from the same neighborhood get together. Pudahuel has in total 35 NGs.

2. The main tenet of the bootstrap method is that the frequency distribution of those computed from the resampling is an estimation of the sample distribution of .

3. 41 being the number of responses for format 1.