The college and university admissions process has become increasingly complex over the past several decades. As application volumes have grown and institutions have expanded the information they collect from applicants, admissions professionals are tasked with making decisions based on a wide range of academic, personal, and experiential factors. Depending on the institution and program, admissions decisions may incorporate grade point averages, standardized test scores, interviews, essays, letters of recommendation, extracurricular involvement, research experience, leadership activities, and numerous other indicators of aptitude.
Despite the amount of information available, admissions decisions are rarely straightforward. A small number of applicants may clearly exceed the institution's expectations, while others may clearly fail to meet minimum requirements. However, a substantial portion of applicants often fall somewhere in the middle, possessing different combinations of strengths and weaknesses that make direct comparisons challenging. One applicant may have exceptional grades but limited extracurricular involvement, while another may demonstrate substantial leadership experience despite a more modest academic record.
As a result, monitoring admissions processes for compliance with applicable legal requirements can be challenging. At their core, however, admissions processes are fundamentally decision-making systems, akin to those used to make employment decisions. Institutions must determine not only what information to collect, but also how that information should be evaluated, combined, and weighted throughout the admissions process. Understanding how data are used to make admissions decisions is critical for evaluating the effectiveness, consistency, and fairness of an admissions system.
This article provides an overview of how admissions processes typically operate, the different types of data used in admissions decisions, common approaches for combining information from multiple sources, and considerations for analyzing and evaluating an institution's admissions practices.
At its core, an admissions process is a funnel. Institutions typically begin with a large applicant pool and gradually reduce the number of candidates as applications move through various stages of review. While the specific process varies between institutions, most admissions systems involve three broad decision points.
First, applicants who clearly fail to meet minimum qualifications are screened out. These decisions are often based on objective criteria established by admissions leadership, such as minimum GPA requirements, prerequisite coursework, or standardized test performance.
Second, some applicants are clearly exceptional. These applicants often possess outstanding academic credentials and accomplishments that make them highly competitive relative to the broader pool.
The greatest challenge typically involves applicants in the middle of the distribution. These individuals often possess different combinations of strengths and weaknesses. One applicant may have exceptional test scores but weaker grades. Another may have strong academic performance but limited research or extracurricular experience. A third may demonstrate substantial adversity or unique experiences that are not easily quantified.
At this point, institutions must determine how different pieces of information should be weighed and combined to reach a decision.
Admissions decisions are informed by a combination of quantitative and qualitative information. Common quantitative measures include overall GPA, relevant coursework GPA, LSAT scores, MCAT scores, SAT or ACT scores, advanced coursework, and research productivity. Qualitative information may include personal statements or essays, letters of recommendation, interviews, extracurricular activities, leadership experiences, community service, and evidence of resilience or adversity. The challenge is rarely a lack of information. Instead, the challenges are recording or tracking data so that it can be consistently evaluated through statistical analysis, then determining how the information should be interpreted and combined.
Before evaluating an admissions process, it is important to understand the nature of the data being collected. Some measures are dichotomous, meaning applicants either possess a characteristic or they do not. Some examples of dichotomous variables include completed prerequisite coursework (Yes/No), state residency (in-state, out of state), and prior military service (Yes/No). Others are continuous and can take on a wide range of values. Examples of continuous variables include GPA, standardized test scores, or number of years of work experience. Still others are categorical and reflect membership in a group or classification, such as undergraduate major and institution attended. Beyond these three basic levels of data, continuous variables can be coded as dichotomous variables when an organization uses them in that manner. For example, with GPA or test scores, the institution can treat them as pass/fail when applicants score above or below a particular cutoff. In that case, GPA would appropriately be evaluated as a dichotomous not a continuous variable.
The distinction matters because different types of data support different decision-making approaches and analytical methods. Often, a continuous variable may be treated as a dichotomous variable when there is a point of diminishing return. For example, with a standardized test score, an applicant with a 1400 SAT may be no more qualified than an applicant with a 1380 SAT; they both possess the minimum level of cognitive ability to be able to complete the coursework at an institution.
Although institutions collect many forms of information, there are only a handful of ways that data are typically incorporated into selection decisions.
Hurdle Models: Applicants must satisfy a specific requirement before advancing. Minimum GPA requirements, minimum test scores, and prerequisite coursework are common examples.
Compensatory Models: Strengths in one area offset weaknesses in another. Composite scores frequently operate under this framework.
Holistic Review: Admissions committees evaluate the applicant as a whole person rather than relying solely on formulas or numerical rankings.
Most admissions systems are not purely holistic, purely compensatory, or purely hurdle-based. Instead, they are combinations of all three.
One of the most common mistakes when evaluating admissions outcomes is assuming that all applicants were evaluated using the same decision rules.
Meaningful analysis requires careful mapping of the admissions process onto a statistical model that allows you to look at how the decision process is operating through a data-driven perspective. Institutions should identify the stages of review, the information available at each stage, the criteria used to advance applicants, and whether different applicant pools are evaluated differently.
Once the process has been mapped, institutions can identify the variables that reflect actual decision-making. This step is particularly important because data recorded in an admissions database may not accurately reflect how decisions were made.
Similarly, institutions frequently convert continuous measures into categories or thresholds during decision-making. Accurately modeling admissions decisions requires creating statistical variables that mirror how information was actually used by decision-makers.
While college, medical school, and law school admissions serve different purposes than employee selection, they rely on many of the same underlying principles. Both involve evaluating multiple predictors, balancing strengths and weaknesses across applicants, and making decisions under conditions of uncertainty.
For that reason, many of the concepts developed within I/O psychology, including hurdle models, compensatory models, holistic evaluation, predictor measurement, and process mapping, can provide valuable insights into how admissions systems function. Understanding these concepts can help institutions better evaluate their admissions practices and make more informed decisions about the students they seek to admit.
To learn more about how Berkshire can help with proactive review of admissions processes, please visit https://www.berkshireassociates.com/college-and-university-admissions-data-analytics-services.