Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Videos

Adjusting for clustering in survey research.

J A Ferguson¹, P N Corey

¹Department of Community Medicine and General Practice, Green College, University of Oxford, England.

DICP : the Annals of Pharmacotherapy

|March 1, 1990

Summary

Cluster sampling in surveys can distort statistical power and sample size. This study introduces correction factors to adjust for clustered sampling effects, ensuring more accurate analysis of patient data in large surveys.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

The correlation between occupation type and fibromyalgia severity.

Occupational medicine (Oxford, England)·2023

Same author

A systematic surveillance programme for infectious salmon anaemia virus supports its absence in the Pacific Northwest of the United States.

Journal of fish diseases·2017

Same author

Impact of Lay-Administered CPR on Survival Rates.

Canadian family physician Medecin de famille canadien·2011

Same author

Suspected Pesticide Poisoning: Evaluating calls to a poison control center.

Canadian family physician Medecin de famille canadien·2011

Same author

Algorithm for inverting lidar returns.

Applied optics·2010

Same author

ECHINOCOCCUS CYST OF THE OMENTUM.

Canadian Medical Association journal·2010

Area of Science:

Epidemiology
Biostatistics
Health Services Research

Background:

Cluster sampling is widely used in surveys but raises concerns about the unit of study.
The choice of study unit can significantly impact statistical power, sample size, and individual data representation.

Purpose of the Study:

To illustrate methodologies for deflating inflated statistical results caused by clustered sampling.
To provide correction factors for accurate analysis of clustered survey data, using the National Ambulatory Medical Care Survey as an example.

Main Methods:

Utilized the 1985 National Ambulatory Medical Care Survey data.
Calculated correction factors using the individual patient as the unit of study.
Applied univariate analysis of variance to develop design effect correction factors.

Main Results:

Correction factors ranged from 1.99 to 35.40 for deflating t-tests, chi-square, and F values.
Clustering effects were more pronounced for continuous outcome variables and those related to physician practice style.
Correction factors were relevant for physician-specific, but not patient-specific, predictors when the physician was the cluster unit.

Conclusions:

Correction factors are essential for accurate analysis of clustered survey data, particularly in pharmacoepidemiology.
Adjusting for clustering ensures the integrity of statistical findings and avoids spurious conclusions.
The proposed methods enable accurate analysis of large, complex survey databases with clustered samples.

Related Experiment Videos