Find Reliability from bimodal data

[#1]

Hello,

I am currently analyzing a few data sets which seem to have a bimodal distribution. I am currently using the 'Life Distribution' under 'Reliability and Survival' but the distributions that best fit are Log Generalized Gamma, Generalized Gamma, Threshold Weibull, Frechet, etc. which don't really make much sense for the given data and where it comes from. Below is a snapshot of what I am trying to model. Is there a way to analyze bimodal distributions in JMP and how can I do that? I don't have JMP Pro which I believe you need to analyze mixed distributions, so just wondering if there are other ways to analyze the data. I am trying to calculate the reliability from the data set.

[#2 · Accepted solution]

Is it possible that you have two different populations represented in this data? If so, you might try a mixture model with two distributions. The two models may be the same or different.