Calculation and Conversion of Effect Measures


Effect measures for continuous outcome »

1. Calculate standard deviation

  • S1: Convert based on the range or interquartile range
  • S2 & 3: Calculate standard deviation in paired design
  • S4 & 5: Calculate standard deviation in grouped design
2. Convert Median and others data to Mean and SD

  • Based on information such as median, Q1, Q3, sample size, minimum, and maximum .
  • Tong method, Box-Cox transformation, QE method, and MLN method.
3. Combined Mean and Standard Deviation Between Groups

  • S1: Combining the means and standard deviations of different subgroups.

4. Mean Difference | Standardized Mean Difference in Grouped Designs
4.1 Mean Difference, Standard Error, and Confidence Interval

  • S1: Based on the sample sizes, means, and standard deviations of the two groups:
  • S2: Estimate the standard error and confidence interval from the mean difference and its P value.
4.2 Standardized Mean Difference, Standard Error, and Confidence Interval

  • S1: Based on sample size, means, and standard deviation information
  • S2 & 3: Based on regression coefficients or point-biserial correlation coefficients
  • S4 & 5: Based on t-value or F-value
  • S6: Calculate standard error and confidence interval from P values

5. Mean Difference | Standardized Mean Difference in Paired Designs
5.1 Calculate Within-Group Pre- Post-Mean Difference and Standard Deviation or Standard Error

  • S1: Calculate Pre- Post-Mean Difference and Standard Deviation
  • S2: Calculate Pre- Post-Mean Difference and Standard Error
5.2 Calculate Standardized Mean Difference, Standard Error, and Confidence Interval

  • Calculate standardized mean difference and standard error in paired or single-group designs.
5.3 Calculate Post-Intervention Mean and Standard Deviation

  • In paired or single-group designs, calculate the post-intervention mean and standard deviation based on baseline and change values

6. Correlation Coefficient
6.1 Calculate Correlation Coefficient, Standard Error, and Confidence Interval

  • S1: Calculate the standard error and confidence interval of the correlation coefficient, Convert between the correlation coefficient and Fisher's Z value.
  • S2: Calculate the point-biserial correlation coefficient based on the t-value or p-value from a two independent sample t-test.
  • S3: Calculate the point-biserial correlation coefficient based on standardized mean differences and their standard errors
6.2 Conversion Between Pearson, Spearman, and Kendall Coefficients

  • S1: Conversion between Pearson correlation coefficient and Spearman correlation coefficient.
  • S2: Conversion between Pearson correlation coefficient and Kendall correlation coefficient

7. Ratio of Mean as a Relative Measure
7.1 Calculate Ratio of Mean and its confidence interval.

  • Estimate ratio of mean based on means and standard deviations in two groups, which can be used for meta-analysis with relative measures (OR\RR\HR).