Explanation of the Log likelihood criterion


New Member
Hello all,

I'm a counseling doctoral student working on dissertation with a mixed methods approach (though qual dominant)

Part of my exploration involves a cluster analysis around a number of variables to determine in which groups I should conduct the following discourse analysis.

Because my data set uses both continuous and discontinuous variables, I get that I have to use the 2-step cluster analysis approach. I also understand overall why this needs to be done due to the data.

My question is, in layman's terms, what is the log likelihood criterion doing to the data to allow for the cluster analysis to account for both categorical and continuous data. Most of the stuff online, understandably, is very notation heavy so I can't follow it. My general impression is that it involves estimating probability but I don't really understand. Any clarification or additional resources anyone can offer?


New Member
The overall is study is mixed methods as the cluster analysis is being used to setup a subsequent discourse analysis in the emerging groups

the cluster analysis contains both continuous and discontinuous variables