How emergency managers (mis?)interpret forecasts

Emergency managers who work on floods and other weather‐related hazards constitute critical frontline responders to disasters. Yet, while these professionals operate in a realm rife with uncertainty related to forecasts and other unknowns, the influence of uncertainty on their decision‐making is poorly understood. Consequently, a national‐level survey of county emergency managers in the United States was administered to examine how they interpret forecast information, using hypothetical climate, flood, and weather scenarios to simulate their responses to uncertain information. The study revealed that even emergency managers with substantial experience take decision shortcuts and make biased choices, just as do members of the general population. Their choices vary depending on such features as the format in which probabilistic forecasts are presented and whether outcomes are represented as gains or losses. In sum, forecast producers who consider these decision processes when developing and communicating forecasts could help to improve flood preparation and potentially reduce disaster losses.     

户外测试检验加速测试

探讨了加速老化测试存在的问题,阐述了户外暴露测试的重要性,指出应该利用户外暴露测试来检验加速老化。使用正确的测试程序,户外暴露测试可以在较短的时间内获得较好的测试结果。采用正确的测试设计、恰当的评估方法、可靠的统计分析及应用参照样品等,可使测试者确定加速程度及加速测试与户外暴露测试的相关系数,以确保加速测试结果的正确性。     

Chorionic villus sampling for prenatal diagnosis in wales using DNA probes—5 years' experience

Chorionic villi were sampled from 125 women who requested prenatal diagnosis, either for genetic disorders or because of advanced maternal age. Of these, 105 samples were obtained by the transcervical route and 20 were obtained by the transabdominal approach. The sampling success rate was 97 per cent (122/125). The mean maternal age of the patients was 31 years (range 17–44) and the mean gestational age at which the chorionic villus sampling was performed was 10 weeks (range 7–13 weeks). Seventy-four of these diagnoses involved the use of DNA markers. The minimal size of the sample used for DNA diagnosis was 5 mg. Maternal contamination was detected in two samples. A diagnosis was provided on all but two samples. The fetal loss rate was high initially but fell to 1·9 per cent in 1988.