Search the Community

Hello, My name is Ardeshir and I have a master’s degree in traffic and transportation engineering. I have to admit that I am new and I don’t have a lot of information and experience in choice models. Right now I am working on a paper to study the heterogeneity in my statistical sample. My study has two alternatives (Pat and car) and I have used the MNL and mixed logit models to study the heterogeneity. The image attached illustrates the ML model outputs. I have 4 questions that I have mentioned below and I would be truly thankful to have your help and guidance: 1- What interpretations can I have from the output of my two models separately? 2- How can I find the mathematical procedure (formula) of heterogeneity by MNL and ML models in order to write them down in the method section of my paper? 3- How can I know that my overall model is statically significant by P-value and chi-squared? 4- Can I say that those parameters which have an unexpected sign are not statically significant? (as an example the value of heterogeneity for in-vehicle time x job trip purpose has a positive sign)

When I read the documentation regarding the "group" element, it looks like it would be used similar to the "all", "any" or "choice" tags. It also says that it has optional "name" and "ref" attributes that cannot be used simultaneously. However, when I try to create a group within a sequence, I get a validation complaint that the required "ref" attribute is missing. The page at http://www.w3schools.com/schema/el_group.asp does not say anything about a situation where this would be required. It does have an example where a group is declared with a name and then used, by reference, later within a complex type definition. There it makes sense that the "ref" attribute is required to reference the pre-defined group. Is this something like a macro that can be used to name some elements to be used in multiple type definitions by reference? Can a group be created within a sequence without a name or ref attribute?