Index 765Tukey’s test, 391
unequal sample sizes and heterogeneity of variance, 392
post hoc power, 239
power, 225–241, 348–354
an alternative way to think of power, 351–353
effect size, 229–231
example, 349
factors affecting the power of a test, 227–229
G*Power to simplify calculations, 238–239, 353–354
power calculations, 350–351
power calculations for differences between two
independent means, 236
power calculations for matched-sample t, 236–237
power calculations for the one-sample t, 231–236
power calculations in more complex designs, 238
retrospective power, 239–241
writing up the results of a power analysis, 241
power analysis for factorial experiments, 429–430
power as a function of alpha, 227–228
power as a function of H 1 , 228
power calculation for Pearson’s r, 283–284
power calculations, 350–351
power calculations for differences between two independent
means, 236
equal sample sizes, 233
unequal sample sizes, 234
power calculations for matched-sample t, 236–237
power calculations for the one-sample t, 231–236
estimating required sample size, 230–232
noncentrality parameters, 232–233
power calculations in more complex designs, 238
power of a test, 99
prediction, 246
predictor, 248
PRESS (Predicted RESidual Sum of Squares), 545
probability, 112–113
generic formula for, 126
probability, basic concepts of, 111–134
about, 111–112
basic terminology and rules, 114–117
Bayes’ theorem, 123–127
binomial distribution, 127–131
binomial distribution to test hypotheses, 131–133
discrete versus continuous variables, 118
multinomial distribution, 133–134
permutations and combinations, 120–123
probability, 112–113
probability distributions for continuous variables, 119–120
probability distributions for discrete variables, 118–119
probability distributions for continuous variables, 119–120
probability distributions for discrete variables, 118–119
probable limits on an observation, 75–76
Q-Q plots, 77
proportional improvement in prediction (PIP), 263
proportional reduction in error (PRE), 263
prospective study, 159
Q-Q (quantile-quantile) plots, 77–79
quadratic functions, 402
qualitative data, 4
quantiles, 52
quantitative data, 4
quartile location, 49
Quetelet, Adolph, 70r^2 and the standard error of estimate, 260
r^2 as a measure of predictable variability, 261
and effect size r^2 , 298
random assignment, 2, 3, 612
random designs, 431
random factor, 431
randomization tests, 661
randomized clinical trial, 160
random-model analysis of variance, 341
random sample, 2
random variable, 246
range, 38
range restrictions, 281
ranking, 304
ranking data, 304
rank-randomization tests, 677
ratio scale, 7–8
real lower limit, 19
real upper limit, 19
reciprocal transformation, 340–341
recombining the effect size and n, 230–231
reflection, 54
reflection as a transformation, 53–54
regression, 246
regression coefficients, 523
regression diagnostics, 539–546
comparing models, 544–546
diagnostic plots, 543–544
regression equation, 516–519
regression equation construction, 546–550
all subsets regression, 547–548
cross-validation, 549–550
missing observations, 550–551
selection methods, 546–547
stepwise regression, 549
regression lines, 248
correlation and beta, 257
intercept, 256
interpretations of regression, 256
a note of caution, 257
slope, 256
standardized regression coefficients, 257
regression surface, 534
REGWQ (Ryan Procedure), 394
rejection level, 96
rejection region, 96
related samples, 194
relationship between fand x^2 , 300rpb^2