Abuse of hypothesis testing statistics in ecological risk assessment

Abstract Statistical hypothesis testing is commonly used inappropriately to analyze data, determine causality, and make decisions about significance in ecological risk assessment. Hypothesis testing is conceptually inappropriate in that it is designed to test scientific hypotheses rather than to estimate risks. It is inappropriate for analysis of field studies because it requires replication and random assignment of treatments. It discourages good toxicity testing and field studies, it provides less protection to ecosystems or their components that are difficult to sample or replicate, and it provides less protection when more treatments or responses are used. It provides a poor basis for decision‐making because it does not generate a conclusion of no effect, it does not indicate the nature or magnitude of effects, it does not address effects at untested exposure levels, and it confounds effects and uncertainty. Attempts to make hypothesis testing less problematical cannot solve these problems. Rather, ri...

[1]  M. Brody,et al.  Managing Ecological Risks at EPA: Issues and Recommendations for Progress , 1995 .

[2]  C. Mann.  Can meta-analysis make policy? , 1994, Science.

[3]  Craig W. Osenberg,et al.  Detection of Environmental Impacts: Natural Variability, Effect Size, and Power Analysis , 1994 .

[4]  A. Underwood.  On Beyond Baci: Sampling Designs That Might Reliably Detect Environmental Disturbances , 1994 .

[5]  Eric P. Smith,et al.  Impact Assessment Using the Before-After-Control-Impact (BACI) Model: Concerns and Comments , 1993 .

[6]  Brian Wynne,et al.  How science fails the environment. , 1993 .

[7]  G. Suter.  Ecological Risk Assessment , 1992 .

[8]  Glenn W. Suter,et al.  Weighing the ecological risk of hazardous waste sites. The Oak Ridge case , 1992 .

[9]  S. Norton,et al.  Framework for ecological risk assessment , 1992 .

[10]  G A Fox,et al.  Practical causal inference for ecoepidemiologists. , 1991, Journal of toxicology and environmental health.

[11]  R. Peterman.  Statistical Power Analysis can Improve Fisheries Research and Management , 1990 .

[12]  David F. Parkhurst,et al.  Statistical Hypothesis Tests and Statistical Power in Pure and Applied Science , 1990 .

[13]  Statistical uncertainty in the no-observed-adverse-effect level , 1989 .

[14]  Glenn W. Suter,et al.  Endpoints for responses of fish to chronic toxic exposures , 1987 .

[15]  D. Cox.  Some General Aspects of the Theory of Statistics , 1986 .

[16]  Allan Stewart-Oaten,et al.  ENVIRONMENTAL IMPACT ASSESSMENT: "PSEUDOREPLICATION" IN TIME?' , 1986 .

[17]  Nigel Smeeton,et al.  Statistics for Toxicologists , 1986 .

[18]  J. Rogers,et al.  Advantages of Using Regression Analysis to Calculate Results of Chronic Toxicity Tests , 1985 .

[19]  F Bro-Rasmussen,et al.  Ecoepidemiology--a casuistic discipline describing ecological disturbances and damages in relation to their specific causes: exemplified by chlorinated phenols and chlorophenoxy acids. , 1984, Regulatory toxicology and pharmacology : RTP.

[20]  S. Hurlbert.  Pseudoreplication and the Design of Ecological Field Experiments , 1984 .

[21]  N. Nelson.  Use of Toxicity Test Data in the Estimation of Risks to Human Health , 1984 .

[22]  Division on Earth.  Risk Assessment in the Federal Government: Managing the Process , 1983 .

[23]  R. Green.  Sampling design and statistical methods for environmental biologists. , 1979 .

[24]  D. Mount.  Present Approaches to Toxicity Testing—A Perspective , 1977 .

[25]  J. Platt.  Strong Inference: Certain systematic methods of scientific thinking may produce much more rapid progress than others. , 1964, Science.

[26]  M. Kendall,et al.  The Logic of Scientific Discovery. , 1959 .