Adaptive Tests of Significance Using Permutations of Residuals with R and SAS

INDEX

Adaptive confidence intervals, see Confidence intervals

Adaptive estimates, see Estimates

Adaptive tests

analysis of covariance, 140

completely randomized designs, 111

cross-over trials, 169

F test statistic, 49

general method, 87

justification of weighting, 70, 71

maximum test, 248

multicenter trials, 169

multifactorial designs, 141

multiple comparison procedures, 118

multiple linear regression, 96

multivariate tests, 191

p-value, 90

paired data, 149

R code for p-value, 59

R code for test statistics, 50, 56, 58

randomized completely block designs, 120

repeated measures, 207

simple linear regression, 91

test statistics, 48, 89

two-way designs, 127

variances not equal, 134

weighted least squares approach, 88

Wilks' lambda test statistic, 48

Analysis of covariance

model, 140

R code, 141

SAS macro, 141

Bandwidth, see Normalizing transformations, bandwidth

Completely randomized designs

example, 114

model, 111

power, 112, 113

R code, 115

SAS macro, 117

significance level, 112

test, 112

Confidence intervals

adjusted y-values, 255

comparison of average widths, 261, 262

convergence criteria for updating method, 258

coverage probabilities, 260

difference between means, 259

example, 259

example of updating method, 256, 257

Garthwaite's method, 254

general updating formula, 264, 265

initial values in updating method, 258

iterative search, 254

lower limit, 256

R code, 266, 267

recommendations, 279

relationship with tests, 253

SAS macro, 268

slope in a linear regression, 263

comparison ...