Background: Nonlinear regression, like linear regression, assumes that the scatter of data around the ideal curve follows a Gaussian or normal distribution. This assumption leads to the familiar goal of regression: to minimize the sum of the squares of the vertical or Y-value distances between the points and the curve. Outliers can dominate the sum-of-the-squares calculation, and lead to misleading results. However, we know of no practical method for routinely identifying outliers when fitting curves with nonlinear regression.Results: We describe a new method for identifying outliers when fitting data with nonlinear regression. We first fit the data using a robust form of nonlinear regression, based on the assumption that scatter follows a Lorentzian distribution. We devised a new adaptive method that gradually becomes more robust as the method proceeds. To define outliers, we adapted the false discovery rate approach to handling multiple comparisons. We then remove the outliers, and analyze the data using ordinary least-squares regression. Because the method combines robust regression and outlier removal, we call it the ROUT method.When analyzing simulated data, where all scatter is Gaussian, our method detects (falsely) one or more outlier in only about 1–3% of experiments. When analyzing data contaminated with one or several outliers, the ROUT method performs well at outlier identification, with an average False Discovery Rate less than 1%.Conclusion: Our method, which combines a new method of robust nonlinear regression with a new method of outlier identification, identifies outliers from nonlinear curve fits with reasonable power and few false positives.

The American Society for Pharmacology and Experimental Therapeutics has revised the Instructions to Authors for Drug Metabolism and Disposition, Journal of Pharmacology and Experimental Therapeutics, and Molecular Pharmacology. These revisions relate to data analysis (including statistical analysis) and reporting but do not tell investigators how to design and perform their experiments. Their overall focus is on greater granularity in the description of what has been done and found. Key recommendations include the need to differentiate between preplanned, hypothesis-testing, and exploratory experiments or studies; explanations of whether key elements of study design, such as sample size and choice of specific statistical tests, had been specified before any data were obtained or adapted thereafter; and explanations of whether any outliers (data points or entire experiments) were eliminated and when the rules for doing so had been defined. Variability should be described by S.D. or interquartile range, and precision should be described by confidence intervals; S.E. should not be used. P values should be used sparingly; in most cases, reporting differences or ratios (effect sizes) with their confidence intervals will be preferred. Depiction of data in figures should provide as much granularity as possible, e.g., by replacing bar graphs with scatter plots wherever feasible and violin or box-and-whisker plots when not. This editorial explains the revisions and the underlying scientific rationale. We believe that these revised guidelines will lead to a less biased and more transparent reporting of research findings.

The American Society for Pharmacology and Experimental Therapeutics has revised the Instructions to Authors for Drug Metabolism and Disposition, Journal of Pharmacology and Experimental Therapeutics, and Molecular Pharmacology. These revisions relate to data analysis (including statistical analysis) and reporting but do not tell investigators how to design and perform their experiments. Their overall focus is on greater granularity in the description of what has been done and found. Key recommendations include the need to differentiate between preplanned, hypothesis-testing, and exploratory experiments or studies; explanations of whether key elements of study design, such as sample size and choice of specific statistical tests, had been specified before any data were obtained or adapted thereafter; and explanation of whether any outliers (data points or entire experiments) were eliminated and when the rules for doing so had been defined. Variability should be described by S.D. or interquartile range, and precision should be described by confidence intervals; S.E. should not be used. P values should be used sparingly; in most cases, reporting differences or ratios (effect sizes) with their confidence intervals will be preferred. Depiction of data in figures should provide as much granularity as possible, e.g., by replacing bar graphs with scatter plots wherever feasible and violin or box-and-whisker plots when not. This editorial explains the revisions and the underlying scientific rationale. We believe that these revised guidelines will lead to a less biased and more transparent reporting of research findings.

Ideally, any experienced investigator with the right tools should be able to reproduce a finding published in a peer-reviewed biomedical science journal. In fact, however, the reproducibility of a large percentage of published findings has been questioned. Undoubtedly, there are many reasons for this, but one reason may be that investigators fool themselves due to a poor understanding of statistical concepts. In particular, investigators often make these mistakes: 1) P-hacking, which is when you reanalyze a data set in many different ways, or perhaps reanalyze with additional replicates, until you get the result you want; 2) overemphasis on P values rather than on the actual size of the observed effect; 3) overuse of statistical hypothesis testing, and being seduced by the word "significant"; and 4) over-reliance on standard errors, which are often misunderstood.

Linked Article: This article is part of a Virtual Themed Issue on Best Practice in Statistical Reporting. To view the other articles in this section visit http://onlinelibrary.wiley.com/journal/10.1111/(ISSN)1476‐5381/homepage/statistical_reporting.htm. The latest BJP Instructions to Authors on statistics can be found in the editorial Curtis et al. (2014). DOI:10.1111/bph.12856

Ideally, any experienced investigator with the right tools should be able to reproduce a finding published in a peer-reviewed biomedical science journal. In fact, the reproducibility of a large percentage of published findings has been questioned. Undoubtedly, there are many reasons for this, but one reason maybe that investigators fool themselves due to a poor understanding of statistical concepts. In particular, investigators often make these mistakes: 1. P-Hacking. This is when you reanalyze a data set in many different ways, or perhaps reanalyze with additional replicates, until you get the result you want. 2. Overemphasis on P values rather than on the actual size of the observed effect. 3. Overuse of statistical hypothesis testing, and being seduced by the word “significant”. 4. Overreliance on standard errors, which are often misunderstood.