Line 205-206: p-values are calculated using values that asymptotically approximate population values, as per the central limit theorem. Most readers would not know the difference. Line 192: Please define what you mean by prediction intervals vs confidence intervals. Line 184-186: You may want to mention Ioannidis’s ‘Proteus phenomenon’ here. Some statisticians have provided counterarguments to the notion of uniform distribution, even though it is broadly true. Line 173: I would be more careful with mentioning uniform distributions of p-values. Line 165: I think that this is an excellent time to mention Ioannidis’s paper on small study effects. odds ratios) takes relative risk = 1 as the null (unless you consider those risks in the exponential scale). For example, hypothesis testing for measures of relative risk (e.g. Line 89: Your reference is excellent – please point out that this is a news feature, not a peer-reviewed research article. I see that you cited this paper in 532, excellent. You may want to cite Fisher’s work (which is also beautifully described in Steve Goodman’s paper) on p-values, his objection to using them as a mechanistic threshold and his concern that Neyman-Pearson’s approach would lead to that. Line 71: I am not aware of Boring (1919), but modern hypothesis testing emerged in the 1920’s after work by Fisher and Neyman-Pearson. Validity of the findingsĬonclusions more or less follow from the discussion. I think that the abstract of this work would derive much benefit from including the following within the abstract: aims and objectives of their review, its perceived significance, innovativeness and its conclusions. The citations within the article are excellent and cover most classical papers in the field. data dredging, p-hacking, publication bias, etc. I would also urge the authors to provide definitions to terms with which not everyone is familiar, e.g. The authors have an engaging style of writing, but I would rather at times this was slightly more formal and less wordy. This article is written in good, but not excellent, English.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |