College Entrance Exams Example

Abstract

This chapter highlights an example of Bayesian Belief Network (BBN) in an academic scenario by evaluating the variables of “Freshman Status” and “ACT Scores.” Here, student retention has great human capital economic utility in the university’s ability to make profit and maintain accreditation and high academic standards to maintain their reputation as an industry leader. It also provides the experimental protocol for conducting the BBN, which includes the following 11-Steps: (a) Step 1: identify a population of interest, (b) Step 2: slice through this population and identify at a minimum two mutually exclusive or disjoint (unconditional) events, which are the subsets of our population, (c) Step 3: determine prior (a priori) or unconditional probabilities, (d) Step 4: identify the conditional event and its subset of mutually exclusive or disjoint (unconditional) elements, (e) Step 5: conduct the random experiment, (f) Step 6: determine frequency counts, (g) Step 7: determine likelihood/conditional probabilities (relative frequencies), (h) Step 8: determine joint probabilities, (i) Step 9: determine posterior probabilities, (j) Step 10: draw a tree diagram, and (k) Step 11: run a Netica replication. In addition, it provides a conclusion, which includes a discussion of posterior and inverse probabilities as they pertain to this scenario