Elimination-By-Aspects (EBA) Model
The elimination-by-aspects (EBA) model (Tversky 1972) starts by establishing the attribute that is most important then retrieving a cutoff value for the attribute. All the other alternatives that have values below the cutoff are then eliminated. The process then continues with the other remaining important attributes till there is only one alternative remaining. The EBA model fits a choice probabilistic model by maximum likelihood.
The model represents the ultimate decision selection and explains it as a set of eliminations. Each alternative that is available for choice is then viewed as having a set of aspects for instance product features.
At every stage of making a choice process, there is an aspect that is selected with the probability been proportional to the aspect that the consumer has placed on it. All the other alternatives that are considered but they do not contain the aspects selected are then eliminated. This process goes on till there is only one alternative remaining and the remaining alternative is considered the one the consumer has selected (Tversky 1972)
While EBA is difficult to apply and calibrate compared to Luce’s Choice Axiom model for choices, it goes a long way to avoid independence from alternatives property that are irrelevant of the latter model. It is because of this reason that the EBA is seen as an appealing model. In addition to this, it has also led to development of other sequential/hierarchical choice models which include the Generalized Elimination Model (Hauser 1986) and the Hierarchical Elimination Model (Sattath and Tversky 1979). Some of the arguments made in regard to EBA include the following:
- M- A data frame of square matrix consisting of total choice frequencies row stimuli are selected over column stimuli
- A-List of vectors that consists of stimulus aspects in which case the default is 1:1 where the number 1 is the stimuli.
- S- The beginning vector that has a default 1/J for all parameters in which case J is the number of parameters.
Usually, A is a list of vectors with the first element of each being any number from 1 to I and additional elements which specify aspects that are shared by numerous stimuli. A should have numerous elements so that it can act as the stimuli.
This paper was written by a writer at assignmentwritinghelp.net and should not be passed off as your own work. However, you can buy a custom paper that is of the same quality and written by experts by contacting our support team.
Once you leave your assignment task with us, rest assured that we will do everything within our power to ensure the paper is free from grammar and typo errors. We will also ensure it is 100% unique and tailored to exceed your expectations.
This will be achieved with the help of our quality assurance team and editors who commit their effort towards ensuring you get a quality paper that gives you value for your money. Don’t content with getting low grades! Contact us today and it will be the best decision you ever made!