Hi everyone
I have a dependent categorical variable of three levels, corresponding to three sectors of activity in the agricultural field, A, B and C and within each of them there are sub-levels, for example under A there are four sub-levels a1, a2, a3 and a4.
Over time, for more profit, farmers change their activity, this change is subject to several factors like demands, financial support, etc. (we are talking here about independent, quantitative and categorical variables).
For example, a farmer who practiced activity A, changed his activity to B, i.e. He has completely changed the activity sector or can only change to a subsector, for example moving from “a2” to “a1” (the same applies to other farmers.
Is there a statistical technique that can be used to model these changes?
I think you can answer your research questions by using multinomial models. However I am a bit unsure how to incorporate sub-levels. Are they ordered ? Depending on your hypotheses I would recommend estimating different models for main and sub-levels, respectively
@Sven. Wouldn't the sublevels be treated as in a split plot design. I don't do much ag stuff but Montgomery's DOE Book should handle this. My second guess would be a Latin square
Best to all, David Booth
You can use the SEM method (Structure Equation Modelling) using AMOS software
thanx Mohamed T. A. Abdelmawgoud David Eugene Booth Sven Greving , for your contribution.
Mohamed T. A. Abdelmawgoud Sir, SEM won't be possible because there is a change of activity in each individual and we have to test for changes. And since it's categorical we should go for non-parametric tests
Ali Mhamed Sir, see A person is doing an activity which is categorized and at some point of time this changes. So if you think of the distribution for such kind of data you won't find a parametric distribution.
HI, may be you should consider a Markov (or semi-Markov) chain model to treat your data. Parametrization could be tricky as it may lead to numerous parameters but some are more parsimonious than a direct one.
I think you should use decision trees before looking to mathematically model your problem. These will guide you to obtain an explanation of the decisions made and thereby model the problem using a multinomial approach.
Thankyou Mamerto Reyes Hernández , Jean Maccario and Arpit Singh for these propositions,
I've done a little research, and it looks like the Markov chain might be the closest solution.
I will keep you informed of any progress,
- Badges
- Science topic
- Similar topics
- Engineering
- Modeling