regint: Compute Power for Regression Interaction... In pwr2ppl: Power Analyses for Common Designs (Power to the People)

Description

Compute Power for Regression Interaction (Correlation/Coefficient Approach)

Usage

 1 2 regint(Group1, Group2, sx1 = 1, sx2 = 1, sy1 = 1, sy2 = 1, nlow, nhigh, alpha = 0.05, Prop_n1 = 0.5, by = 2, Estimates = 1)

Arguments

 Group1 Estimates (r or b) for Group 1 Group2 Estimates (r or b) for Group 2 sx1 Standard deviation of predictor, group 1 (defaults to 1) sx2 Standard deviation of predictor, group 2 (defaults to 1) sy1 Standard deviation of outcome, group 1 (defaults to 1) sy2 Standard deviation of outcome, group 2 (defaults to 1) nlow starting sample size nhigh ending sample size alpha Type I error (default is .05) Prop_n1 Proportion of Sample in First Group (defaults to equal sample sizes) by incremental increase in sample (e.g. nlow = 10, nhigh = 24, by = 2, produces estimates of 10, 12, and 14) Estimates 1 for Correlations (default), 2 for coefficients

Value

Power for Regression Interaction (Correlation/Coefficient Approach)

Examples

 1 regint(Group1=-.26,Group2=.25, alpha=.05,Prop_n1=0.5,nlow=110, nhigh=140,by=2,Estimates=1)

pwr2ppl documentation built on June 12, 2019, 5:03 p.m.