What does censoring actually do to data?

Hi all,

I have no statistics background but recently started to use MATLAB's statistical toolbox to fit weibull functions to right censored data. I feel extremely uncomfortable in putting my data into a black box. So what does, for example, the function wblfit do if I pass a censoring vector?

Best wishes, Max

Answer 1

[the cyclist]

It is a bit difficult to know how much detail you want, when you "have no statistics background", but want an explanation of a statistical concept. I can point you to a couple resources.

First, the Wikipedia article on censoring gives a general introduction to the concept of censoring in statistics.

Second, to pry open the black box, you can open up (most) MATLAB functions in the editor, and see exactly what they are calculating. For example

edit wblfit.m

will open up that function so that you can see what is going on. In this case, you are very quickly going to realize that wblfit just does a transform and then call evfit, so

edit evfit.m

will show you where the real work is being done.

[Maximilian]

Thank you! The book cited in the evfit.m explained everything pretty clear.