This analysis included calculation of proportions of patients with ≥50%, ≥75%, or 100% reduction in monthly migraine days (MMD) from baseline to last 4 weeks of a 12-week double-blind phase.
The calculation of proportions or percentages allows comparison across non-homogeneous datasets, and was executed as follows.
Unless otherwise stated the calculation of proportions will based on the sample size of the population of interest.
Unless otherwise stated, the calculation of proportions will be based on the sample size of the population of interest.
The pitfall relates to the calculation of proportions (including percent changes), which are used in common outcome criteria like response.
These included calculation of means with standard error (SE) of means or medians for continuous variables and calculation of proportions for dichotomous variables.
Descriptive statistics included calculation of proportions of those who received each immunization antigen and those who had completed the immunization schedule.
We repeated the calculation of proportions of regions containing a gene, and bootstrapped n = 10,000 to obtain empirical "expected values" and p values that a random set of regions would, more often than observed, contain genes.
The descriptive analysis included calculation of proportions over each of our independent variables (all categorical) for persons with and without CBD using a chi-squared test suitable for complex survey data to test whether each variable was distributed differently between those with and without CBD.
