Effect-size indices for dichotomized outcomes in meta-analysis.
read more
Citations
Unemployment impairs mental health: Meta-analyses
Does tailoring matter? Meta-analytic review of tailored print health behavior change interventions.
Are leader stereotypes masculine? A meta-analysis of three research paradigms.
The comprehensive complication index: a novel continuous scale to measure surgical morbidity.
References
Statistical Power Analysis for the Behavioral Sciences
Statistical methods for rates and proportions
Statistical Methods for Meta-Analysis
Statistical Methods for Rates and Proportions.
Related Papers (5)
Frequently Asked Questions (9)
Q2. What was the power transformation used to transform the two standard normal distributions?
Using the Fleishman (1978) power transformation, X a + bZ + cZ2 + dZ3, the authors transformed two standard normal distributions, Z ∼ N(0, 1), to reflect the target distribution shapes.
Q3. What is the way to use the effect size indices?
If the meta-analyst wants to mix all of the studies in the same meta-analysis, the effect-size indices treated here for dichotomized outcomes could be tentatively used on the true dichotomies, with thecaution of including a moderator variable for testing possible differences between the two d metrics.
Q4. What is the significance of the sampling variance of the effect size indices?
In meta-analysis, such sampling variances are important because the meta-analytic statistical models usually weight each effect size by its inverse variance.
Q5. What is the common assumption for the effect size indices?
On the basis of past research, the normal distribution is the most usual assumption for data in the empirical studies; so the authors compared the bias and the sampling variance of the different effect-size indices assuming that the two populations are normally distributed.
Q6. Why did the dCox, dProbit, and dasin indices?
This is because the latter indices correct the artifact of dichotomizing continuous outcomes, introducing a multiplier constant that increases the sampling variance of the estimator, whereas the dp, d , and dasin indices do not correct the negative bias due to dichotomization.
Q7. What is the common meta-analysis method used in the last 20 years?
In the last 20 years, meta-analysis has become a very popular and useful research methodology to integrate the results of a set of empirical studies about a given topic.
Q8. What was the performance of the dHox, dProbit, and das?
the dHH index showed a systematic negative bias of a larger magnitude, and finally, the dp, d , and dasin indices continued showing the worst performance, with a systematic negative bias, the same as under normal distributions.
Q9. What are the two extensions of the standardized mean difference d?
Two extensions of the standardized mean difference d have been applied to studies with such 2 × 2 tables: the standardized proportion differences, dp, obtained from success (or failure) proportions, and the phi coefficient.