Index 377
negatively skewed data, 121
NHANES (National Health and Nutrition
Examination Survey), 82, 86, 93, 148–157
NIH (National Institutes of Health), 72–73
nominal variables, 102
non-code-based methods, 60
nonlinear function, 211
nonlinear least-squares regression, 12
nonlinear regression, 279–286
nonlinear trends, 277
nonparametric regression, 286–290
nonparametric tests, 48–49, 157
non-proportional hazards, 323
non-sampling error, 78
non-steroidal anti-inflammatory drugs (NSAIDS),
159–166
normal distribution, 13, 36, 114, 353
normal Q-Q graph, 222–223
normal-based confidence intervals, 134
normal-based confidence limits, 134
normality assumption, 143
not rule, 31
notches, 128
NPV (negative predicted value), 187
NSAIDs (non-steroidal anti-inflammatory drugs),
159–166
nuisance variables, 145
null hypothesis
definition of, 40
evaluating, 42–43, 150, 152–153, 322, 351–352
example of, 161–165, 319
null model, 228, 242, 259
numerator, 41–42
numerical data, 107–109, 114–123
O
obesity, 177–178, 182–183
observational research, 88–90
observed count, 164
observed versus predicted graph, 245
ODA (SAS OnDemand for Academics), 55–56
odds, 32–33, 181
odds ratio (OR), 94–96, 181–183, 266–267
Office for Human Research Protections (OHRP), 72
one-dimensional arrays, 25
one-group t test, 148
one-sided confidence interval, 133
one-way ANOVA, 144
open-source software, 58–59
OpenStat, 59
OR (odds ratio), 94–96, 181–183, 266–267
or rule, 31
order of operation, 24
ordinal data, 102
ordinary multiple linear regression model. See
multiple regression
ordinary regression, 210
outcome, 208
outcome-related measurements, 64
outliers, 229
overall accuracy, 185
P
p value
definition of, 41, 242
determining, 165–167
evaluating, 42, 144, 147–157, 226–227, 340
from the H-L test, 259
paired t test, 148
paired values, 363
parabolic relationship, 214, 252–253
parallel design, 65
parameters, 33–34, 77, 208, 279
parametric tests, 48–49
parsimonious models, 293
parsimony, 293
participant identification (ID), 240
participant study identifier, 104