Table 1:

Assumptions and imputations used to calculate data if missing from published report

Data neededData availableAssumptions/imputations
SD of the differenceSE of the differenceMultiply SE by square root of sample size
Confidence interval around the differenceFor n ≥ 100, assume standard normal distribution
For n < 100, assume t distribution
SE of the differencep value for mean differenceConvert p value to t value at same degrees of freedom; divide mean difference by t value
Final scoreBaseline and change scoresAdd or subtract the change score from baseline value
SD of final scoresSD of baseline and change scoresSum baseline and change variances
SDRangeNo appropriate conversion possible
  • Note: SD = standard deviation, SE = standard error.