Seven common errors in statistical analysis by environmental scientists all stem from an outdated understanding of statistics. I'll define the seven 'perilous errors' and how each can be avoided. They revolve around old ideas about hypothesis tests, p-values, using logarithms of data, evaluating what is a good regression equation, evaluating outliers and dealing with nondetects. Understanding why each error is perilous can save the scientist from publishing incorrect statements, using inefficient analysis methods, and wasting scarce financial resources. These errors have persisted through the years -- break the cycle and step into the 21st Century.
Dennis Helsel (Ph.D. Environmental Science and Engineering) is a consultant and trainer on statistical methods for environmental and natural resource scientists through his firm, Practical Stats. He has authored three textbooks including Statistics for Censored Environmental Data using Minitab and R (2012), which presents methods for handling data below detection/reporting limits. He regularly conducts webinars, seminars and courses on topics such as Urban Legends in Environmental Statistics and Statistics for Contaminated Sites. He worked as a hydrologist, geochemist, and statistician for 30 years at the US Geological Survey before starting Practical Stats. For his training courses in applied statistics within and outside North America he received the Distinguished Achievement Award in 2003 from the American Statistical Association’s Section on Statistics and the Environment.