Group homogeneity tests - difference when using right censor and interval censored data

[#1]

Hi,

can you explain how JMP handles Group homogeneity tests (Logrank, Wilcoxon, Peto-Peto) with different type of event censoring? Are there any differences in methods used by JMP to calculate those tests under different event censoring?

I made a 2 populations comparisons in JMP platform Life Distribution - Compare Groups, using same dataset, but with different event censoring definition. For uncensored events, p-value for tests was around 10%, whereas for interval censored events tests' p-value dropped to 1%, which changes the analysis conclusion.

Below are JMP results:

Events Interval Censoring

Uncensored events

Additional question: For uncensored event case, sometimes in statistics summary table an unnamed column is added with term "Biased" in it. What does it mean in this context? I couldn't find any information on this in JMP documentation. 

Thank you for your help.

[#2]

Hello,

I am facing the same problem, does anyone know the answer?

Thank you,

MM