red M&M and multiple testing

if you heard of multiple testing problem you have probably seen this cartoon. If you haven't you should check out these links : link1, link2, link3

But did you know that the MARS company pulled the red M&M's (link) and then reintroduced them (link)

and I wonder.

some time ago i met with researcher that wants to see if stem cells are generating new cells in their favorite system.

the data is collected by labeling certain cells and then counting them at number of time points.

literature search reveals that it is mostly modeled by a differential equation that is then fitted to data.

as someone raised in a stochastic simulation world i wonder why it is not modeled by a stochastic birth and death process. Has anyone compared the fit of these 2 approaches?

hi, i do statistics for living but it is my hobby too.

i use the term statistics in a very wide sense. it includes biostatistics, bioinformatics, computational biology, machine learning, ...

i plan to write on any and all topics related or not so much to statistics, its theory and application.