inclusion of nonoccurrences, 153
independence, 152
independence of observations, 321
independent events, 114
independent variable, 4
inferential statistics, 5
influence, 540
influence of extreme values on the variance and standard
deviation, 43–44
inner fence, 49
interactions, 414, 421–423
intercept, 254, 256
internal validity, 3
interpretation of simple effects, 425
interpretations of regression, 256
interpreting an analysis of covariance, 611–612
interpreting the model, 654
interquartile range, 38
interquartile range and other range statistics, 38–39
interval, 118
interval scale, 7
iterative solution, 566
joint and conditional probabilities, 116
joint probability (5.2), 116–117
kappa (k), measure of agreement, 165–166
Kendall’s coefficient of concordance (W), 309–311
Kendall’s tau coefficient, 304
kernel density plots, 22–24
Kolmogorov-Smirnov Test, 79
Kruskall-Wallis one-way analysis of variance, 683–684
kurtosis, 29
lambda, 638
Laplace, Pierre-Simon, 69
Larzelere and Mulaik test, 381
leading digits, 24
leaves, 25
leptokurtic distribution, 29
less significant digits, 25
leverage, 540
likelihood ratios, 156
likelihood ratio tests, 156–157, 568
linear combination, 371
linear contrasts, 371
linearity of regression, 280
linear regression
of continuous measure, 565
with dichotomous dependent variable, 571
linear regression models, 246
linear transformations, 53
linear transformations, effect on data by, 52–54
nonlinear transformations, 54
reflection as a transformation, 53–54
standardization, 54
listwise deletion, 549
loess, 257
logarithmic transformation, 338
logic of the analysis of variance, 321–323
logistic regression, 561–570, 571
logit, 566
logit transformation, 566
log-linear analysis, 629–654
deriving models, 648–652
model specification, 637–638
odds and odds ratios, 641–642
testing models, 638–641
three-way tables, 643–648
treatment effects, 652–654
treatment effects (lambda), 642–643
two-way contingency tables, 631–637
log-linear models, 630
log odds, 566
looking at one predictor while controlling for another, 521magnitude of the experimental effect, 344, 369
main effect, 416
Mallow’s Cp statistic, 547
Mann-Whitney Ustatistic, 678
Mantel-Haenszel statistic, 157–159
marginal distribution, 265
marginal totals, 146
matched-samples, 194
matched-sample t test, 196
mean, 33, 35
mean, advantages and disadvantages of, 33
mean absolute deviation, 40
mean and variance of a binomial distribution, 130–131
mean square error (MSerror), 322, 327
mean square error ( error), 607
mean squares, 327
mean square treatment (MStreatment), 322
mean square within (MSwithin), 322
measurement data, 4
measurement scales, 6–9
measure of agreement, 165–166
measure of importance, 551
measures of association, 159, 294
measures of association and effect size, 438–443
d-family measures, 440
partial effects, 440
r-family measures, 438
measures of central tendency
mean, 33, 35
median, 32, 34
mode, 32, 34
relative advantages and disadvantages of the mode, the
median, and the mean, 33
trimmed mean, 35–36
measures of central tendency and dispersion
using SPSS, 51–52MS¿762 Index