Index 375
interaction, 296–297
interaction terms, 237, 329
intercept, 215, 221, 224–225, 272, 329
intercept row, 224
interim analysis, 76
International Conference on Harmonization
(ICH), 72
International Review Board (IRB), 72
Internet sources. See also G*Power; Microsoft
Excel
ClinCalc, 14
Comprehensive R Archive Network (CRAN), 58
Epi Info, 59
GraphPad, 67
International Business Machines, 57
National Health and Nutrition Examination
Survey (NHANES), 82, 86, 93, 148–157
National Institutes of Health (NIH), 72–73
OpenStat and LazStats, 59
Power and Sample Size Calculation (PS), 60,
171–172, 324–325
SAS OnDemand for Academics (ODA), 55–56
Statista, 34
StatPages, 158, 190, 282
interpolation, 208
inter-quartile range (IQR), 120
inter-rater reliability, 188–190
interval data, 102
intervention-related measurements, 63
interventions, 61–62
intra-rater reliability, 188–190
IQR (inter-quartile range), 120
IRB (International Review Board), 72
iterative models, 246–247
K
Kaplan-Meier method, 313–316, 328, 330
Kaplan-Meier (K-M) survival estimate, 313, 317
kappa (κ), 189–190
kilograms (kg), 218–220, 225–226, 239
Kruskal, William (statistician), 141
Kruskal-Wallis test, 47–49, 144, 157
kurtosis, 121–122
L
Last Observation Carried Forward (LOCF), 75
last-seen date, 303
LazStats, 59
least-squares line, 234
left skewed data, 121
leptokurtic distribution, 122
lethal dose, 262
levels of measurement, 102–103
LFU (lost to follow-up), 303–304, 308–309
life sciences, 1
life-table method, 307, 311–316, 320–321
likert scale, 102, 106
line of best fit, 215–216
linear combination, 329
linear function, 210–211
linear model, 272–273
linear regression, 210
link function, 273–274
liver cancer, 93–97
locally weighted scatterplot smoothing (LOWESS)
curve-fitting, 12, 286–290
LOCF (Last Observation Carried Forward), 75
logarithms, 21–22, 118–119
logistic regression
basics of, 251–254
definition of, 12, 210
disadvantages of, 266–268
evaluating, 257–265
sample size for, 268–269
using, 249–250, 255–257
log-normal distribution, 36, 125, 353–354
log-rank test, 317–325, 328–330
longitudinal research, 90
lost to follow-up (LFU), 303–304, 308–309