Index

absolute risk reduction see risk difference

abstracts

accuracy, diagnostic

ACP Journal Club

alpha (α)

American Journal of Epidemiology

American Journal of Physiology

American Journal of Public Health

analysis of covariance

analysis of variance

Annals of Internal Medicine

area under receiver operating characteristic (ROC) curve (AUC)

association

attributable risk (AR)

authors

instructions to
statistical checklists
statistical guidelines

baseline differences

benefit per 1000 patients

bias

binary variables

clinical trials
meta-analysis
regression analysis

Binomial distribution

table for confidence intervals

blinding techniques

Bonferroni correction

bootstrap methods

bias corrected and accelerated method
bias corrected method
parametric
percentile method
software

brackets (…)

British Heart Journal

British Medical Journal (BMJ)

misuse of confidence intervals
policies
statistical checklists
clinical trials
comments
general papers
use of confidence intervals

case–control study

matched
with 1. M matching
more than two levels of exposure
unmatched
series

causality

censored observations

checklists, statistical

chi squared (x2) test

CIA software see Confidence Interval Analysis (CIA) software

clinically worthwhile difference, smallest

clinical trials

adjusting for other variables
arguments against confidence intervals
benefit per 1000 patients
cluster
crossover trials see crossover trials guidelines for authors
interpretation
multiple groups
number needed to treat
parallel groups
binary outcome
continuous outcome
time to event outcome
presentation of results
results section
software
statistical checklists
subgroups

Clopper–Pearson method

cluster randomised trials

complex analyses, guidelines for authors

Confidence Interval Analysis (CIA) software

Data Editor window
help
Method window
options
outline of program
Output window
updates and bug fixes

confidence intervals

degree
in multiple comparisons
in non-parametric methods
options in CIA software
dissenting voices
exact
guidelines for authors
link to P values
in medical journals see journals, medical
method of calculation
null values and
presentation of results
rationale
suggested use
use
us P values
us standard errors
width
factors affecting
sample size estimation from

confidence limits

CONSORT statement

checklist
patient flow diagram

continuous variables clinical trials

diagnostic problem
meta-analysis
non-Normal
regression analysis
summary statistics

control group

correlated variables, hypothesis tests

correlation

coefficient
Pearson’s product moment
Spearman’s rank
guidelines for authors
matrix
technical details

cost-effectiveness ratio

Cox regression

in clinical trials
in time to event studies

crossover trials

binary outcome
continuous outcome

data

description
presentation see presentation of results
tied

decimal places, number of

degrees of freedom

design, study

Diabetic Medicine

diagnostic tests

area under ROC curve
classification
into more than two groups
into two groups
clinical cutpoints
interpretation
kappa statistic
likelihood ratios
measurement-based
positive/negative predictive values
sensitivity
specificity

difference, smallest clinically worthwhile

discussion section

distribution-free methods see nonparametric methods

distribution of data

economics, health

editorials, in medical journals

editorial staff, statistical checklists

epidemiological studies

Epidemiology

error bars

errors in journal articles

estimates, sample study

imprecision
presentation

Evidence-Based Medicine

ex

exact confidence intervals

medians
proportions
software

extrapolation

false-negative results

false-positive results

figures

Fisher’s exact test

flow diagram

follow-up studies see also time to event studies

forest plot

Geigy scientific tables

graphical presentation

guidelines, statistical

discussion section
methods section
results section
presentation of results
statistical analysis

hazard

ratio

histograms

hypergeometric distribution

hypothesis tests see also P values

dependent (correlated) data
guidelines for authors
interpretation
medical journal policies
multiple
overemphasis
presentation of results
sample size determination
two-sided
vs confidence intervals

incidence

rates
ratio, standardised see standardised incidence ratio
study
use of term

individuals

data presentation
repeated measurements
as units of analysis

instructions to authors

International Committee of Medical Journal Editors

interquartile range

Journal of the American Medical Association (JAMA)

Journal of Obstetrics and Gynaecology

journals, medical

editorials and expository articles
instructions to authors
misuse of confidence intervals
policies
secondary publication
statistical checklists
statistical guidelines
surveys of use

Kaplan–Meier survival curves

kappa statistic

The Lancet

life tables

likelihood ratios (LR)

log method
score method

logarithmic transformation

likelihood ratio
means/differences
in meta-analysis
relative risk

loge x

logistic regression

in clinical trials
multiple

logit method

logistic regression
pooled estimate of odds ratio

logrank test

McNemar’s test

Mann–Whitney U test

Mantel–Haenszel method

in meta-analysis

masking see blinding techniques

mean

difference (two samples)
non-Normal data
paired data
presentation
sampling distribution
standardised
unpaired data
weighted
geometric
ratio
guidelines for authors
notation
sampling distribution
single sample
non-Normal data
Normal data
two-group comparison, sample size
variability, sample size estimation

median

definition
difference (two samples)
paired data
tables for confidence intervals
unpaired data
exact method
guidelines for authors
sampling distribution
single sample
survival time
difference between
single sample
two samples
table for confidence intervals
technical note

Medical Journal of Australia

meta-analysis

analysis methods
binary outcome
continuous outcome
fixed effects approach
interpretation
presentation of results
random effects approach
time to event outcome

methods section, guidelines for authors

mid-P confidence interval

MINITAB

misuse, of confidence intervals

mortality

ratio, standardised see standardised mortality ratio
studies

multiple comparisons

multiple regression

in clinical trials
guidelines for authors

multivariate techniques

Newcombe’s method

paired proportions
unpaired proportions

non-Normal data

means/differences
medians/differences

non-parametric methods

presentation of results
terminology

non-significant results

Normal distribution

area under ROC curve (AUC)
assumption, guidelines
correlation analysis
logit method
means/differences
notation
substitution method and
table for confidence intervals

notation

null value

confidence intervals including
sample size estimation and

number needed to harm (NNH)

number needed to treat (NNT)

for benefit (NNTB)
for harm (NNTH)
meta-analysis

observational studies

odds ratio (OR)

clinical trials
homogeneity test
logistic regression
logit method
logit pooled estimate
Mantel–Haenszel method
matched case–control studies
meta-analysis
more than two levels of exposure
relative odds reduction
stratified analysis
unmatched case–control studies

one-sided test

outliers

p

P see P values

paired data

in clinical trials
means
medians
presentation
proportions

parameter

parametric bootstrap

parametric methods

Pearson’s product moment correlation coefficient

percentages

presentation
sample size estimation and

percentile

method, bootstrapping
range

person-years at risk

πx

planning, study

± sign

Poisson distribution

table for confidence intervals

power, study

sample size estimation

precision

presented results
sample study estimates

prediction

diagnostic tests and
intervals, in regression analysis

predictive values

negative
positive
prevalence-adjusted

presentation of results clinical trials

confidence intervals
guidelines for authors
limitations of P values
meta-analysis
summary statistics
in tables
textual

prevalence

predictive values and
studies

probability

posterior
prior
values see P values

proportional hazards regression analysis see Cox regression

proportions

difference see also proportions, two samples
sampling distribution
notation
single sample
recommended (Wilson’s) method
traditional method
software
survival
technical note
two-group comparison, sample size
two samples, paired data
recommended (Newcombe’s) method
traditional method
two samples, unpaired data
recommended (Newcombe’s) method
traditional method
very low/high
zero (no observed events)

P values see also hypothesis tests; mid-P values

Bonferroni correction
cut-off points
exact
interpretation
limitations
link to confidence intervals
medical journal policies
presentation
suggested use
vs confidence intervals

quality of life studies

quantiles see also median

QUOROM statement

random allocation

randomised controlled trials see also clinical trials

statistical checklists

receiver operating characteristic (ROC) curve, area under (AUC)

referees, statistical checklists

reference interval

regression

Cox see Cox regression guidelines for authors
interpretation
line(s)
common slope of two
difference between slopes of two
intercept (A)
mean value of y for any given value of x
slope (B)
vertical distance between two
linear
logistic see logistic regression more than two samples
multiple see multiple regression
prediction intervals
residuals
single sample
technical details
two samples

relative risk (risk ratio)

clinical trials
incidence study
matched case–control study
meta-analysis
reduction
unmatched case–control study

repeated measurements

replicate readings

Resampling Stats

residuals, regression analysis

response rates

results, presentation of see presentation of results

results section

presentation of results
graphical
hypothesis tests
for individuals
numerical precision
summary statistics
tables
use of technical terms
statistical analysis
complex analyses
confidence intervals
correlation
data transformation
descriptive information
hypothesis tests
outliers
paired observations
regression
repeated measurements
survival data
underlying assumptions
units of analysis

risk difference (absolute risk reduction, ARR)

meta-analysis
number needed to treat and

risk ratio see relative risk

risk reduction

absolute (ARR) see risk difference relative

rounding up/down, numbers

sample size

confidence interval approach
formulae
limitations
notation
recommendations
confidence interval width and
effective
hypothesis test-based approach
hypothesis test interpretation and
notation
sampling distribution and
small
statistical checklists

sample study estimates see estimates, sample study

sampling distribution

sampling variation

scatter plots

screening tests

SD see standard deviation

SE see standard error sensitivity

serial measurements

σ

significance

clinical
levels see also P values sample size estimation
statistical
confidence intervals and
vs clinical importance
tests see hypothesis tests

skewed data see non-Normal data

smallest clinically worthwhile difference

small samples

software, computer

Spearman’s rank correlation coefficient

specificity

SPlus

standard deviation (SD)

assumptions involving
confidence interval width and
notation
± sign notation
pooled estimate
pooled residual, two regression lines
prediction interval for individuals
residual, regression line

standard error (SE)

area under ROC curve (AUC)
confidence interval calculation
difference between independent estimates
kappa statistic

mean

difference between
median survival times
difference between
notation

odds ratio (log)

± sign notation
presentation
proportions
difference between
regression line, predicted values
relative risk (log)
standardised rate
survival proportions
difference between
vs confidence intervals

standardisation

direct method
indirect method

standardised incidence ratio (SIR)

ratio of two

standardised mortality ratio (SMR)

ratio of two
sampling distribution

standardised rates

standardised ratios

ratio of two

Stata

statistical analysis

checklists
methods section
results section

statistical checklists

general papers
other types of study
randomised controlled trials

statistical guidelines see guidelines, statistical statisticians

advice from
assessment of medical papers

views on confidence intervals

StatXact software

stratified analysis

bootstrap method
odds ratio

subgroups

in clinical trials
in correlation analysis
in regression analysis

substitution method

summary statistics

presentation
standard error

surveys

response rates

survival

analysis see also time to event studies

guidelines for authors
curves, Kaplan–Meier
proportions
time
median see median, survival time regression
tables

t distribution

table for confidence intervals

t test

test statistics, suggested use

tied data

time to event studies

in clinical trials
Cox regression
hazard ratio
meta-analysis
number needed to treat
single sample
two samples

“to”, in range of confidence intervals

transformation, data, see also logarithmic transformation

guidelines for authors

uncertainty, standard error as measure

units of analysis

Vancouver guidelines

variability

confidence interval width and
in sample size estimation
standard deviation as measure

weaknesses, in discussion section

Wilcoxon matched pairs signed rank sum distribution

Wilcoxon one sample test

Wilcoxon test, paired

Wilcoxon two sample rank sum distribution

Wilson’s method, single proportions

img_240_001.gif

Z1–α/2 value

Zα/2 value