Statisticians.org Copyright by Mark Chang
Group Sequential Design for Two Means
Group Sequential Design for Proportions
Group Sequential Design for Survival Medians
Sample-Size Re-estimation Design for Two Means
Sample-Size Re-estimation Design for Proportions
Pick-the-Winner Design for Means (Phase II/III Seamless Design)
Pick-the-Winner Design for Proportions (Phase II/III Seamless Design)
Two-sample z-Test (large sample or
population variance known)
Test for treatment mean difference
with 2x2 crossover design
One way Repeated measures ANOVA for Two groups
Test
two correlation coefficients -- Fisher's arctan transformation
Two-Sample
Hypothesis Test for Proportion Difference
Arcsine Method
(with continuity Correction) (n1 P1>10, n2 P2>10)
Poisson Method
for two unequal sample proportions(np1<10, nP2<10)
Asymptotic
z-method considering variance difference
Person's
Chi-square Test (Improved, Fleiss)- Large sample
Snedecor's Method
for Normal but Heterogeneous Large Samples
Whitehead
proportional odds ratio model with k categories and two treatments (logistic
regression)
Repeated
measures for two group proportions
Chi-square test(normal approximation) for one sample
proportion with k categories
Mantel-Haenszel test for odds ratio with k strata
-Nam-method (with continuity correction))
Kendall's Test of Independence
Pasternack-Gilbert
Method: Exponential survival distribution - No censoring
Test for mean
survival time with exponential distribution (with censoring)
Exponential
survival distribution with uniform patient enrollment rate over time T - with
censoring
LogRank test
for survival analysis
One-Sample
or Paired-sample Hypothesis Test for Mean
Sign Test for
Median Difference - Paired sample
Wilcoxon Sum Rank
Test - One Sample
Chi-square
test(normal approximation) for one sample proportion
Chi-square
test(normal approximation) for one sample proportion - adjusted for finite
population
Chi-square
test(normal approximation) for one sample proportion with k categories
Test for Bloch-Kraemer intraclass Kappa coefficient
(binary outcome)
McNemar's Test for a Paired Sample
Ho: Test
Correlation (regression coefficient) from zero - Fisher's arctan
transformation
Test Ho:
regression coefficient = zero -- Fisher's arctan transformation
Logistic
Regression on x for binary outcome
Logistic
Regression on x for binary outcome with covariates
Linear regression
y=a+bx, test Ho: b=b0, vs. Ha: b<>b0.
Linear regression
y1=a1+b1x, y2=a2+b2x. Test Ho: b1=b2, vs. Ha: b1<>b2.
One Way Contrast
between means
William's Test
For Minimum Effective Dose
Cochran-Armitage
Test for linear/Monotonic Trend (dose response)
1.
Chow,
S.C., Chang, M. (2006). Adaptive Design Methods in Clinical Trials. Chapman
& Hall/CRC.
4.
Chang,
M. (2010). Monte Carlo Simulation for the pharmaceutical industry. Chapman
& Hall/CRC
5.
Chang,
M. (2011). Modern Issues and Methods in Biostatistics, Springer, NY.
7.
Chang,
M. (2012). Paradoxes in Scientific Inference. Taylor & Francis Group, LLC.
8.
Mark
Chang (2014). Principles of Scientific Methods. Mark Chang, Taylor &
Francis Group
FDA-Industry Workshop 2020 Short Course Slides Part 1
FDA-Industry Workshop 2020 Short Course Slides Part 2