Statistical Software
Partial Testing Design
(Written by Stuart G. Baker)
New Approach (with ROC curves): See Baker SG, Pinsky P. A proposed design and analysis for comparing digital and analog mammography: special ROC methods for cancer screening. JASA. June 2001.
Original Approach: These files are based on the manuscript by Baker SG, Connor RJ, Kessler LG. The partial testing design: a less costly way to test equivalence for sensitivity and specificity. Statistics in Medicine 1998;17:2219-2232.
It runs in Mathematica 2.2 (better for graphics) or 3.0 and requires the files listed below.
Downloads
Download All (zip, 6kb)
ptd.m (m, 11kb) |
Sample size calculations (and loads other files) |
ptdtime.m (m, 7kb) |
Compute optimal interval between screening tests |
ptddelta.m (m, 5kb) |
Create symbolic delta based on model parameters |
To reproduce the calculations in the manuscript, load ptd.m.
To obtain the sample size tradeoff curve, use SampleSizePlot[{pN, pA, deltaN}], where pN and pA are the values of p under the null and alternative hypotheses, respectively, and deltaN is the value of delta under the null hypothesis. In the manuscript pN=pA=.09 and deltaN=-.05.
To obtain the minimum value of m* use SampleSizeLimit [{pN,pA,deltaN},n], where n is symbolic. To reproduce the table of optimal times between screens, type OptimalTimeList[deltasym,parsym,psilist]. To plot delta over time, use PlotDelta[deltasym,parsym,{.1,.5}].
Back to top