Designing The Future Of Student Success Using Artificial Intelligence

• By: WU, Maryann (University Of Southern California, United States)
• Co-author(s): Dr Maryann Wu (University Of Southern California, Los Angeles, United States)
  Prof Ying Wang (University Of Southern California, Los Angeles, United States)
  Dr Ian Haworth (University Of Southern California, Los Angeles, United States)
  
• Abstract:

  Background
  Artificial intelligence (AI) is rapidly transforming the world enabling us to rethink how information is integrated and analyzed, and how resulting insights can be used to improve overall decision making. AI is currently being used to detect diseases faster, help provide personalized treatment plans for patients, and automate certain processes such as drug discovery. However, few studies have been done on how AI can be used to provide personalized educational plans for students to improve overall success outcomes upon graduation.
  
  Purpose
  Our objectives are to: 1) collect and curate curricular and co-curricular data related to student outcomes at multiple stages of pharmacy school, and 2) build a multi-step AI model called AI-SiPS: Artificial Intelligence - Success in Pharmacy School to identify variables that can be used to predict student success upon graduation.
  
  Method
  The current AI-SiPS model is based on data from a Doctor of Pharmacy program in the United States using the classes of 2019 to 2022. Data includes course grades (n=745 students), a third-year survey on perceived readiness for advanced pharmacy practice experience (APPE) rotations (n=261), student rotation assignments (n=740), perceived experience in the APPE year (n=564), and initial professional step students took after graduation (also n=564). Student success outcomes are divided into broad categories of “Residency”, “Industry”, “Community/Hospital” (RICH).
  
  Data is being analyzed using Konstanz Information Miner (KNIME ver. 4.5.1). Using this platform, we have examined 1) relationships of APPE rotation order with residency matching, 2) performance in didactic courses with subsequent pre-APPE confidence, and 3) pre-APPE confidence with RICH outcomes. Decision tree analysis was used to find data breakpoints in each relationship.
  
  Analysis has indicated that: 1) an early Acute Care APPE leads to a higher residency match rate than a late Acute Care APPE (70.2% vs. 58.0%), 2) students who perform well in certain therapeutic didactic courses, such as cardiology, have a higher chance of matching for residency than those who do not perform well (77.8% vs. 51.7%), and 3) students who feel confident about starting their APPEs have a higher chance of matching for residency than those who report not feeling confident (77.1% vs. 66.7%). Earlier didactic courses focused on practical application of knowledge also show impact on self-perceived confidence.
  
  Results
  These results have been used to implement changes in processes, including earlier identification of students with a desire to pursue residency. This information has permitted appropriate reorganization of students’ APPE rotation order, which is predicted to lead to higher residency match rates. Similar changes may be possible to support other career goals.
  
  Conclusion
  Our AI-SiPS model represents a conceptual approach to using AI to support student success and desired post-graduation outcomes. Results from our research may allow for educators to advise students on specific actions that need to be taken in order to maximize their chances of success based on individual career aspirations. Future plans include adding admissions, co-curricular, and mentorship data to the AI-SiPS model and providing AI workflows to other educational institutions interested in using AI to support student success.