PASS 12 Update

Update Release Notes

This page lists the changes that have been made to PASS 12 since it was released.

Version: 12.0.11

Released: July 28, 2015

1. Fixed a problem that occurred when the computer language setting (e.g. Spanish, Colombia) uses a comma as both the decimal separator and the list separator. In this case, none of the procedures would run.

2. Fixed “Overflow” error in plots with extremely large axis values.

3. Fixed Anderson-Darling Normality Test and Range Normality Test calculation for large sample sizes. The power was incorrect for the Anderson-Darling Test when N > ~4000 (it was reported as 0.000 when it should have been 1.000 for very non-normal data). The Range Test is only available when N < 1000, but results were still appearing.

4. Fixed example templates for Confidence Intervals for One Proportion procedure. They were generating a "Type mismatch" error.

5. This is the final update to PASS 12.

Version: 12.0.10

Released: October 2, 2014

1. Fixes calculation problems with all Conditional Power procedures. We discovered that the book Chang (2008) on which all conditional power calculations are based has typos in the formulas for both conditional and predictive power on page 73. These typos were confirmed by the author, Mark Chang, himself. The correct formulas are given in Jennison and Turnbull (2000), pages 207-213. PASS has been updated so that the conditional and predictive power values are now calculated using the corrected formulas. The documentation was also updated with the correct formulas.

Version: 12.0.9

Released: September 18, 2014

1. Fixes a problem with the search for sample size in Tests for One Proportion. In some instances when the proportions are very small, the calculated sample size was slightly overestimated.

2. Fixes an error in several Tests for Two Proportions procedures (i.e. Inequality, Non-Inferiority, Superiority, Equivalence). Plots had wrong values on the Y-axis when solving for Odds Ratio, Ratio, or Difference.

3. In the Logrank Test (Input Mortality) procedure, the “Solve For” drop-down text was corrected.

4. Fixes a problem in all procedures that include Futility Boundaries. Setting the Number of Skipped Futility Looks to something greater than 1 generated an overflow error.

5. Fixes a problem in the Tests for the Difference of Two Hazard Rates Assuming an Exponential Model procedure. The decimals for power and proportions were not being set correctly so the reports did not display the correct number of decimals. Sometimes, power was displayed as only 0 or 1. This bug did not affect calculations, only the report format.

Version: 12.0.8

Released: March 27, 2014

1. Fixes a problem with the search for sample size in Tests for One Proportion. In some instances when the proportions are very small, the calculated sample size was slightly overestimated.

Version: 12.0.7

Released: March 4, 2014

1. Fixes a problem with the Confidence Intervals for One Proportion procedure. The summary statements were not being shown when the option to show them was checked.

2. Fixes a problem with the Mixed Models procedure for more complicated models. Some models were deemed non-estimable when they are, in fact, estimable.

3. Fixes a problem with the cluster-randomized procedures involving proportions. The sample sizes were being rounded before the ICC adjustment was made resulting in slightly different power results.

4. The documentation for the two-sample proportions tests procedures (Inequality, Non-Inferority, Superiority, and Equivalence) was updated to correct for typos in the power formulas.

Version: 12.0.6

Released: January 24, 2014

1. Fixes a problem with the Group-Sequential Tests for Two Means (Simulation) procedure that was generating a "subscript out of range" error in rare cases.

2. Fixes a problem with the search for ICC in procedures that use ICC (e.g. Non-Inferiority Tests for Two Proportions in a Cluster-Randomized Design). The search algorithm was not finding a non-zero solution.

Version: 12.0.5

Released: November 8, 2013

1. Fixes a problem with the Tests for Two Groups of Pre-Post Scores procedure. The power and sample sizes were not being computed correctly when σ(T1) ≠ σ(T2).

Version: 12.0.4

Released: October 24, 2013

1. Fixes a problem with the Repeated Measures Analysis procedure. The search for sample size terminated prematurely when using Mean Input Type = Means in Spreadsheet. This error did not affect results when entering Means and Interaction Effects directly.

Version: 12.0.3

Released: August 6, 2013

1. Fixes minor issues in Inequality Tests for Two Means in a Cluster-Randomized Design procedure documentation.

2. Fixes report for Tests for Two means in a Repeated Measures Design when setting Covariance Type = Simple. Rho was being reported as -1.7E+308 instead of 0.

3. Fixes calculation errors in 4 survival routines:
-
Tests for Two Survival Curves Using Cox's Proportional Hazards Model
Non-Inferiority Tests for Two Survival Curves Using Cox's Proportional Hazards Model
Equivalence Tests for Two Survival Curves Using Cox's Proportional Hazards Model
Superiority by a Margin Tests for Two Survival Curves Using Cox's Proportional Hazards Model.
-
We discovered that the book on which the routines are based had a typographical error such that the sample size reported was 2x the correct sample size. This has been corrected so that the correct sample sizes are now reported.

Version: 12.0.2

Released: February 19, 2013

1. Fixes minor issues in Lin's Concordance procedure documentation.

Version: 12.0.1

Released: January 8, 2013

1. Initial Release of PASS 12.