Will Machines Decide Which College You Get Into?

March 2, 2019

 Photo Source: Wikimedia Commons

 

     Every year, a new swath of high school seniors go through the agonizing process of applying to college, waiting to see which ones accept them, and finally deciding among the acceptances which one to attend.  Colleges receive thousands of applications.  In the United States alone, 750,000 students submitted over 3 million applications to colleges and universities last year according to an article by Kayla Webley. (time.com) The volume of applications has multiplied in part due to the electronic application filing system known as the Common Application.  The Common App makes it much easier for applicants to apply to multiple schools because much of their information and essays can be reused.  They do not have to deal with the slow tedious process of filling out paper applications for each school of interest. Moreover, the competition for spots in the top schools continues to escalate evidenced by the lawsuit against Harvard University admissions brought by a group called Students for Fair Admissions. The suit claims that Harvard discriminates against Asian-Americans in the admissions process.  According to an article by Carrie Jung, “The plaintiff's analysis shows Asian-Americans routinely perform better in academic and extracurricular ratings in this system, but they consistently fall behind other ethnic groups in what is known as a ‘personal score.’” (npr.org)  The suit challenges the admission practices of Harvard and will most likely end up at the Supreme Court of the United States to revisit affirmative action.  

 

      A new arbiter of admissions has arisen beyond the people in admissions departments in colleges in the form of artificial intelligence. It has not happened yet but the future may see the use of artificial intelligence to determine the appropriateness of candidates for admission to certain schools.  A company called vibeffect (thevibeffect.com) developed an artificial intelligence driven program that uses many personal attributes about individual students and features of over 1,000 colleges combined with survey information from “High Thriving” Students to predict at which colleges a student would best perform.  In an article in eCampusNews titled “Science develops an algorithm for college selection—but does it work?,” the author describes the variables as follows, “vibeffect’s individual variables include things like whether someone has held a job, whether he or she likes working independently or on a team, and if the person is apt to ask for help (or not). On the college side, vibeffect factors in a school’s use of innovative teaching techniques, transportation options, and social opportunities.”  vibeffect claims to have 90% success rate when students use this tool to select schools where they will thrive.  Another company, Naviance, describes itself as college and career readiness software that helps schools plan and develop careers.  Naviance begins collecting information such as performance and class selection in middle school.  Naviance also collects college acceptances from a particular high school to model the acceptance rate into different colleges for students with different grade point averages.  It also helps guidance counselors to recommend courses to help students achieve their goals. 

 

      Colleges every year get inundated with thousands of applications, which has been facilitated by the advent of electronic filing.  Additionally, the competition for acceptance to the top schools gets fiercer, as evidenced by law suits such as the case with the Students for Fair Admissions v. Harvard University.  The shear volume of applications may tempt schools to automate the selection process to narrow down applicants.  Today, several companies use artificial intelligence and statistics to help students find an appropriate school for them.  It serves to reason that if a machine can recommend the right college based on individual attributes and college qualities, then it can be turned around and colleges using the same information could identify the best students for them.   Such systems may replace the admissions board, but such systems need careful monitoring to avoid being contaminated by bias.  A number of cases in facial recognition and the law have witnessed racial bias because the data that trained the machine contained bias. In one example, a product called COMPAS (Correctional Offender Management Profiling for Alternative Sanctions) has been developed by Northpointe Inc to evaluate the risk that a criminal will commit another crime to help judges set sentences. However, it turned out that the data that trained COMPAS had bias towards African Americans’ likelihood of committing another crime over other racial groups.  Admissions boards and students may be tempted to adopt such complex tools like vibeffect and Naviance to reduce the work to sift through thousands of applications, but they should remain vigilant against bias in the machine.

 

 

 

Dr. Smith’s career in scientific and information research spans the areas of bioinformatics, artificial intelligence, toxicology, and chemistry. He has published a number of peer-reviewed scientific papers. He has worked over the past seventeen years developing advanced analytics, machine learning, and knowledge management tools to enable research and support high level decision making. Tim completed his Ph.D. in Toxicology at Cornell University and a Bachelor of Science in chemistry from the University of Washington.

 

You can buy his book on Amazon in paperback and in kindle format here.

 

 

Share on Facebook
Share on Twitter
Please reload