College Admissions: Should universities use practical models or experts to determine the qualifications of potential students?

An important role of the admissions office is to choose students who will fit well with the school based on their academics and extracurricular activities. By using models to predict the success of potential students, they can have more consistent results that expresses the higher chance for these students to do well at UCR. Experts are better than models in terms of having common sense. They can catch errors that models may not be able to do. Comparatively, models prove to have more consistency and reliability. By using specific variables that affect their decision making, they are able to target the specific needs they require from future students. Some of these variables include test scores, grades, letter of recommendations, written essay, extracurricular activities, and consistency in behavior. Problems may arise if the model begins to overgeneralize based on their sample size. It may force other students out if they do not necessarily match all variables in the model but have strong skills that may prove to be useful in other fields. The use of predictive models can lower the risk of any biases that an expert admissions officer may have. From an outside perspective, the use of models in admissions process will appear to be more fair in its decision. A common mistake that an admission officer can make is to create a narrative based on the given information of the student. For example, if they look at their admissions essay that explains their journey to higher education, there could be more room for them to make a decision based on their compassion for the student. Due to the constraint of how many students the school can offer admission to, students who have higher scores and meet every requirement that the school lays forth may not receive an admission letter compared to another student who had lower test scores but did more meaningful extracurricular activities that gave back to the community. If the model can choose a range of potential students based on the amount of extracurricular activities attempted and a range of acceptable test scores, they will be able to achieve a range that factors that make students unique whether they had high scores or received many honors for notable work. This can help to diversify the pool of UCR students. A model may not be able to identify the other factors that make the student outstanding even if their grades may not be as high as other students. Admissions officers could care more about the work ethic of students and their motivation to succeed rather than their past scores. A student who is hardworking will have higher chances of succeeding compared to a student who had high scores in the past because there is no guarantee that past behavior will carry forward into the future. A problem that may arise with predictive models is that they may choose students based on test scores and grades but not factor in the situation they were in. Models can present data, but experts are needed to change data into information they can use. An alternative to using just a model or an expert alone, I think it could be quite useful for admissions officers to use a model that can identify the pool of test scores they are willing to accept. After, these potential students should be re-evaluated by experts who view how well the students did given circumstances written in their essays.