Statistical Methods for Continuous Outcome Measures

Methodological Approaches for Continuous Outcomes »

1. Standard Deviation Estimation Methods
  • Method 1: Estimation based on range or interquartile range
  • Method 2 & 3: Standard deviation calculation in paired designs
  • Method 4 & 5: Standard deviation calculation in parallel-group designs
2. Transformation of Median and Other Statistics to Mean and Standard Deviation
  • Estimation based on median, quartiles, sample size, and range statistics
  • Implementation of Tong method, Box-Cox transformation, QE method, and MLN method
3. Pooling of Means and Standard Deviations Across Subgroups
  • Method 1: Combining means and standard deviations from multiple subgroups
4. Effect Measures in Parallel-Group Designs: Mean Difference and Standardized Mean Difference
4.1 Calculation of Mean Difference, Standard Error, and Confidence Intervals
  • Method 1: Based on sample sizes, means, and standard deviations from both groups
  • Method 2: Estimation of standard error and confidence interval from mean difference and P-value
4.2 Calculation of Standardized Mean Difference, Standard Error, and Confidence Interval
  • Method 1: Based on sample sizes, means, and standard deviations
  • Method 2 & 3: Based on regression coefficients or point-biserial correlation coefficients
  • Method 4 & 5: Based on t-statistics or F-statistics
  • Method 6: Estimation of standard error and confidence intervals from P-values
5. Effect Measures in Paired Designs: Mean Difference and Standardized Mean Difference
5.1 Calculation of Pre-Post Mean Difference and Standard Deviation or Standard Error
  • Method 1: Calculation of pre-post mean difference and standard deviation
  • Method 2: Calculation of pre-post mean difference and standard error
5.2 Calculation of Standardized Mean Difference, Standard Error, and Confidence Interval
  • Statistical methods for calculating standardized mean difference and standard error in paired or single-group designs
5.3 Calculation of Post-Intervention Mean and Standard Deviation
  • Estimation of post-intervention mean and standard deviation based on baseline and change values in paired or single-group designs
6. Methods for Correlation Coefficient
6.1 Calculation of Correlation Coefficients, Standard Errors, and Confidence Intervals
  • Method 1: Calculation of standard error and confidence intervals for correlation coefficients, Transformation between correlation coefficients and Fisher's Z-scores
  • Method 2: Calculation of point-biserial correlation coefficients from t-statistics or P-values in two-group comparisons
  • Method 3: Calculation of point-biserial correlation coefficients from standardized mean differences and their standard errors
6.2 Transformation Between Different Correlation Coefficients
  • Method 1: Transformation between Pearson and Spearman correlation coefficients
  • Method 2: Transformation between Pearson and Kendall correlation coefficients
7. Relative Effect Measures: Mean Ratio Analysis
7.1 Calculation of Mean Ratio and Confidence Intervals
  • Estimation of mean ratio based on means and standard deviations from two groups, suitable for meta-analysis using relative effect measures (OR/RR/HR)