I am an Excel 2007 user and am having some issues using Solver to obtain Probit estimates.
If you are not familiar with the Probit model, what I am trying to do is find parameters that maximize the likelihood of observing my sample. This likelihood is simply the product of the likelihoods of observing each individual datapoint. The tricky part is that each observation's likelihood is determined by an "if" statement taking on the value based on the cumulative normal distribution if true, and 1 minus that value if false.
Please see attached. I am maximizing cell G27 by changing the cells M1 through M3... this process gives accurate estimates if the number of factors making up G27 is less than 25, but if I try using more it just returns zeros for the parameters.
Thanks a lot for any help you can give!
Bookmarks