Guidelines for Statistical Report

Guidelines for Statistical Report

Any manuscript submitted to the Bangladesh Journal of Pharmacology should have a detailed report of statistical methods. Results are rigorously reported and the appropriate method for study design should be applied. These are:

  >  Software package (name, version and relevant references)

>  Technical details or procedures to reproduce the analysis

>  Identify the research design and independent variables (between- or within-subjects)

>  Pre-processed data:

  • Any analysis carried out to confirm the data (such as linearity, co-linearity, and normality of the distribution).
  • Any data that were transformed with reason to preform and description of the transformation

> How outliers were treated and your analysis (full dataset and outliers removed)

> How missing/excluded data were handled

> Define the threshold for significance (alpha)

> When required, mention sample sizes with a description of how to perform it. Note the inputs for power, the effect on sample size, and alpha. Report the number of independent replications for each experiment

> For analysis of variance (ANOVA), details any post hoc tests

> Details of any correction to account for multiple comparisons. If the correction was not applied, then mention the justification for not doing so

> Mention all options for statistical procedures. If t-tests were carried out, then clarify whether these were performed by either one- or two-tailed. Mention the type of t-test conducted (e.g. one sample, within-/between-subjects)

> In the case of step-wise multiple regression analyses:

  • Report the alpha level
  • Variables for collinearity and interaction
  • Variable selection process (such as forward-stepwise, best subset). 

> Bayesian analysis: Mention the choice of prior trial probabilities and the method of selection. Markov chain Monte Carlo settings should be reported

> Reporting of statistical results: It must be reported appropriately.

  • Units of measurement:Define in all Figures and Tables
  • Distribution properties:Standard deviation, standard error of the mean, confidence intervals and central tendency (mean, median)
  • Regression analyses:Complete results of any regression analysis (as a supplementary file). Mention all calculated regression coefficients, standard error, and confidence intervals and p values
  • Reporting parameters:Test statistics (F/t/r) and associated degrees of freedom are required to report. The effect on sample size and the confidence intervals must be mentioned when necessary. If percentages are mentioned, then the numerator and denominator are to be written
  • P value: Mention the exact p value which is greater than or equal to 0.05. P value less than 0.05 is written as p<0.05