This checklist for design and analysis provides guidance for transparent reporting and scientific rigour of preclinical research as set out in Experimental design and analysis and their reporting II: updated and simplified guidance for authors and peer reviewers (http://onlinelibrary.wiley.com/doi/10.1111/bph.14153/full). This checklist is intended as a guide for submission to the British Journal of Pharmacology.
| Criteria | Number | Issue | Where to place information |
|---|---|---|---|
| Group sizes | 1a | The exact group sizes (n) are provided, not a range. | Figure/Table Legends/Methods |
| 1b | Group size refers to biological samples and not technical replicates. | Methods | |
| 1c | Inferential statistics (comparisons between groups) are undertaken only if n = 5/group or more. A valid explanation is provided for data with n of less than 5. | Methods/Results | |
| Experimental design | 2a | The Methods declare whether randomization was undertaken and, if it was not, a valid scientific justification is provided. | Methods |
| 2b | The Methods declare whether blinding was undertaken and, if it was not, a valid scientific justification is provided. | Methods | |
| 2c | The Methods declare that group sizes were designed to be equal, and any loss of samples resulting in inequalities is explained. | Methods | |
| 2d | Details of any prior sample size estimation (e.g. power calculations). | Methods | |
| Statistical plan | 3a | A data and statistical analysis section is provided giving details of all summary and inferential statistical tests used. | Methods |
| 3b | Details of any statistical package or program employed are provided and details of which tests were used in the Data and Statistical Analysis section. | Methods | |
| 3c | If ANOVA is used a statement is provided indicating that post‐hoc tests were conducted only if F was significant and there was no variance inhomogeneity. | Methods | |
| Data and statistical analysis | 4a | Any data normalisation (e.g., expression of values as ‘fold mean control’), is explained with a valid scientific justification (i.e., to control for unwanted sources of variation). | Results |
| 4b | If normalisation generates values with no variance (i.e., control SEM = 0) the data should not be subjected to parametric statistical analysis. | Results | |
| 4c | A valid explanation is provided for any data transformation (such as log transformation to generate a Gaussian‐distributed dataset). | Results | |
| Level of probability | 5 | The threshold P value deemed to constitute statistical significance should be defined in the Methods and this value only should be used to denote statistical significance in the Results. | Methods/Results |
| Outliers/ exclusion criteria | 6 | Inclusion and/or exclusion criteria are clearly defined in the Methods. | Methods |