2. Choice Models Family of statistical models that attempt to capture the underlying rational decision process by which individuals choose among different options
10. MLM is intended for use when the dependent variable takes on more than two outcomes and the outcomes have no natural ordering (e. g. university majors, soap brands, political parties)
11. All predictors in the model should measure individual characteristics that are hypothesized to affect the outcome choice
12.
13. MLM allows the effects of the independent variables to differ for each distinct outcome category
14. By default, the base outcome category is the one most commonly selected but the model may set to use as base category any other that is meaningful for the researcher
18. It estimates the effects of change on variables measuring individual characteristics
19. It is commonly used when a group characteristic is hypothesized to have effect on the choiche (e.g. health condition, political party affiliation)
20.
21. It allows controlling for unobserved heteregeneity when it is constant over time (e.g. gender, ideology)
22.
23. Outcome choices are expressed as functions of the characteristics of the alternatives themselves as well as functions of characteristics of the choosers (as it occurs in the MLM)
24.
25. Outcome choices are expressed as functions of the characteristics of the alternatives themselves as well as functions of characteristics of the choosers (as it occurs in the MLM)
33. Stereotype logistic models can be used when the researcher is unsure of the relevance of the ordering, as is often the case when subjects are asked to assess or judge something (e.g. Likert scales for customer satisfaction)
34. Stereotype logistic models can also be used when the researcher suspect that some of the alternatives are similar (e.g. job positions)
37. Nested logit should be used for analyzing models in which the choice has a two-level or three level structure (e. g. deciding first whether or not to buy a car, second why type of car to buy, and third, specific car model)
38. The nested logit model can be explained as the product of a series of MNL choice models defining each level in a tree structure
39.
40. The model is specified in series of equations for each choice level
41. Dependent variables should include case-specific variables (individual characteristics) and alternative-specific variables (choice characteristics)
42.
43. Post-estimation techniques generate predicted probabilities for specific individuals profiles, discrete changes in probabilities and factors changes in the odds depending on the change of the value of any specific variable
50. Minimal bibliography Cameron, A. C. and P. Trivedi (2009). Microeconometrics Using Stata . College Station, Texas, Stata Press. Koppelman F & Sethi V (2000) Closed-form discrete-choice models. In: Hensher DA & Button KJ (eds) Handbook of Transport Modelling , Volume 1, of Handbooks in Transport (pp 211–222). Oxford: Pergamon Press. Long, J. S. (1997). Regression Models for Categorical and Limited Dependent Variables. Thousands Oaks, CA: Sage Publications. Long, J.S., and Freese, J. (2001). Regression Models for Categorical Dependent Variables Using Stata . College Station, TX: A Stata Press Publication. McFadden, D. (1978) Modeling the Choice of Residential Location. Transportation Research Record 672, TRB, National Research Council, Washington, D.C., pp.72-77.