Learn R Programming

sjstats (version 0.19.0)

design_effect: Design effects for two-level mixed models

Description

Compute the design effect (also called Variance Inflation Factor) for mixed models with two-level design.

Usage

design_effect(n, icc = 0.05)

Value

The design effect (Variance Inflation Factor) for the two-level model.

Arguments

n

Average number of observations per grouping cluster (i.e. level-2 unit).

icc

Assumed intraclass correlation coefficient for multilevel-model.

Details

The formula for the design effect is simply (1 + (n - 1) * icc).

References

Bland JM. 2000. Sample size in guidelines trials. Fam Pract. (17), 17-20.

Hsieh FY, Lavori PW, Cohen HJ, Feussner JR. 2003. An Overview of Variance Inflation Factors for Sample-Size Calculation. Evaluation and the Health Professions 26: 239-257. tools:::Rd_expr_doi("10.1177/0163278703255230")

Snijders TAB. 2005. Power and Sample Size in Multilevel Linear Models. In: Everitt BS, Howell DC (Hrsg.). Encyclopedia of Statistics in Behavioral Science. Chichester, UK: John Wiley and Sons, Ltd. tools:::Rd_expr_doi("10.1002/0470013192.bsa492")

Thompson DM, Fernald DH, Mold JW. 2012. Intraclass Correlation Coefficients Typical of Cluster-Randomized Studies: Estimates From the Robert Wood Johnson Prescription for Health Projects. The Annals of Family Medicine;10(3):235-40. tools:::Rd_expr_doi("10.1370/afm.1347")

Examples

Run this code
# Design effect for two-level model with 30 observations per
# cluster group (level-2 unit) and an assumed intraclass
# correlation coefficient of 0.05.
design_effect(n = 30)

# Design effect for two-level model with 24 observation per cluster
# group and an assumed intraclass correlation coefficient of 0.2.
design_effect(n = 24, icc = 0.2)

Run the code above in your browser using DataLab