Logistic regression and modal value.

Hello users of Shazam,

I use the logit-estimator (logistic regeression). Alle my variables (dependent & independent) are dichotomous. Shazam calculates the marginal probabilities of the variables. The notion "typical case" is important when the marginal probabilities are calculated. According to the Shazam manual - see last version, p. 313:

"An approach is to set values to represent a "typical" case. A "typical case" can be defined by setting all dummy variables to their modal values and all other variables to their mean values".

Question: What is meant by "modal value"? Is it the most frequent value of the random dummy variable?

edit retag close merge delete

Ken - thanks for your confirmation. I have two follow up questions to this topic: Let's say we have a sample of N observations and that the dummy variable has the value 1 in 50% of the cases and value 0 in the rest (50%) of the sample. (a) In this case It is not possible to conclude what is the modal value - the most frequent value of the dummy variable? (b) Why not use the average value of the dummy variable? In this particular case the average value would have been 0.5?

( 2012-01-09 05:43:45 +0000 )edit

Sort by ยป oldest newest most voted

In this case there is no 'typical case' you have two possibilities so just do it twice to show the two typical cases. Dummy variables should always be 0 or 1 not .5

more

Thanks Ken!

( 2012-01-10 07:53:26 +0000 )edit

Yes, modal is the most frequent value

more