Log-Logistic Analysis of Herbicide Dose-Response Relationships
Steven S. Seefeldt, Jens‐Erik Beck Jensen, E. Patrick Fuerst
Abstract
Steven S. Seefeldt, Jens‐Erik Beck Jensen, E. Patrick Fuerst
Abstract
Dose-response studies are an important tool in weed science. The use of such studies has become especially prevalent following the widespread development of herbicide resistant weeds. In the past, analyses of dose-response studies have utilized various types of transformations and equations which can be validated with several statistical techniques. Most dose-response analysis methods 1) do not accurately describe data at the extremes of doses and 2) do not provide a proper statistical test for the difference(s) between two or more dose-response curves. Consequently, results of dose-response studies are analyzed and reported in a great variety of ways, and comparison of results among various researchers is not possible. The objective of this paper is to review the principles involved in dose-response research and explain the log-logistic analysis of herbicide dose-response relationships. In this paper the log-logistic model is illustrated using a nonlinear computer analysis of experimental data. The log-logistic model is an appropriate method for analyzing most dose-response studies. This model has been used widely and successfully in weed science for many years in Europe. The log-logistic model possesses several clear advantages over other analysis methods and the authors suggest that it should be widely adopted as a standard herbicide dose-response analysis method.
OpenAlex reports 1367 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Dose-response studies are an important tool in weed science. The use of such studies has become especially prevalent following the widespread development of herbicide resistant weeds. In the past, analyses of dose-response studies have utilized various types of transformations and equations which can be validated with several statistical techniques. Most dose-response analysis methods 1) do not accurately describe data at the extremes of doses and 2) do not provide a proper statistical test for the difference(s) between two or more dose-response curves. Consequently, results of dose-response studies are analyzed and reported in a great variety of ways, and comparison of results among various researchers is not possible. The objective of this paper is to review the principles involved in dose-response research and explain the log-logistic analysis of herbicide dose-response relationships. In this paper the log-logistic model is illustrated using a nonlinear computer analysis of experimental data. The log-logistic model is an appropriate method for analyzing most dose-response studies. This model has been used widely and successfully in weed science for many years in Europe. The log-logistic model possesses several clear advantages over other analysis methods and the authors suggest that it should be widely adopted as a standard herbicide dose-response analysis method.
Key concepts: Logistic regression, Weed science, Logistic function, Response analysis, Statistical analysis, Response time, Statistics, Variety (cybernetics)