What question did this study set out to answer?

To evaluate statistical power determination and error rates when applying generalized estimating equations (GEE) to geographically clustered binary outcomes.

September 30, 1996Statistics in Medicine

Power Determination for Geographically Clustered Data Using Generalized Estimating Equations

View Full Paper

Ask AI

Bookmark

To evaluate statistical power determination and error rates when applying generalized estimating equations (GEE) to geographically clustered binary outcomes.
Conducted a simulation study modeling a cluster-randomized trial of an intervention to prevent lower-back injuries among nursing home personnel.
Modeled baseline nursing home injury risk via a beta distribution and individual worker risk via a Bernoulli logit model with fixed covariates and random intercepts.
Failing to control for intracluster correlation produced overestimates of statistical power and inflated type I error rates by as much as 20 percent.
Generalized estimating equations effectively accounted for intracluster correlation when it existed, but correlation estimates were negatively biased and type I error remained inflated when no cluster correlation was present.

Discussion

No takes yet. Share an insight, caveat, or question.

Implication

Simulation study demonstrates power overestimation and type I error inflation in cluster designs, highlighting critical limitations of generalized estimating equations.

Cite This Study

Hendricks et al. (1996) studied this question.

synapsesocial.com/papers/6a74a48a1078cf7de2b1cea1 https://doi.org/10.1002/(sici)1097-0258(19960930)15:18<1951::aid-sim407>3.0.co;2-p

View Full Paper

Ask AI

Bookmark

Implication

Simulation study demonstrates power overestimation and type I error inflation in cluster designs, highlighting critical limitations of generalized estimating equations.

Power Determination for Geographically Clustered Data Using Generalized Estimating Equations

Key Points

Discussion

Implication

Cite This Study

Power Determination for Geographically Clustered Data Using Generalized Estimating Equations

Key Points

Discussion

Implication

Cite This Study

Also Consider

Also Consider