2 · Which of these does your manuscript actually print (in the text or tables)? Select all that apply — this suggests which checks may have usable inputs.
2×2 tables or event counts by group Subgroup Ns that should add up to the total Means ± SD Two groups' means ± SD, each group's n, and the p-value comparing them Percentages with their denominators (n / N) Odds / hazard / risk ratios (or differences) with a 95% CI and the p-value Test statistics with degrees of freedom (t, F, χ², z, r) and the p-value each gave Table 1 categories with percentages (and counts) that cover every patient Medians with IQR, or ranges / scale limits alongside a mean or median A mean ± SD of a measurement (not a change score) for a whole group and again for some of the same patients (those with follow-up, say), with a sentence that prints both sizes and says the smaller group is drawn from the larger one, in one of the forms the checks page lists (“104 of the 120 patients”) A sweeping statement in the text (“all”, “almost every”, “most”, “none”) about a group your tables also count
And in your methods section:
Two subgroups of one comparison, each with its own estimate and confidence interval, where one is significant and the other is not A non-inferiority margin, the interval it is judged against, and the conclusion you draw (equivalence and bioequivalence margins are not recomputed) A minimal important difference you state, and the confidence interval for the outcome you call unchanged A logistic or Cox model with its event count and the predictors entered The number of patients lost to follow-up, alongside a 2×2 table for a statistically significant outcome A participant flow diagram (assessed, excluded, randomized, analyzed) Meta-analysis heterogeneity: Cochran's Q with its degrees of freedom — both printed — and the I² you report A sample-size calculation for a two-arm parallel comparison, stating the target N (and whether it is a total or per group) plus the effect size and variability — or the event rates — you assumed (paired, crossover, cluster and multi-arm designs are not recomputed)
Pick a study type to continue.