2024 AIRUM Concurrent Sessions Listing
Thursday, November 7, 2024

9:00 - 9:45 am1:30 - 2:15 pm2:30 pm - 3:15

Downloadable PDF version: Sessions Listing

9:00 - 9:45 am

13. The use of AI in Data Analytics

Presenter: Jim Sorenson, University of Mary

Description: In this presentation, we delve into the transformative impact of conversational AI models like ChatGPT on data analysis and data science. These sophisticated tools are reshaping how we extract insights, automate processes, and make decisions, significantly enhancing productivity and fostering innovation. I will share my personal experience with utilizing ChatGPT to design and implement predictive analytics models. Specifically, I'll demonstrate how this technology assists in overcoming language barriers commonly encountered by data professionals transitioning between SQL and more complex programming languages like R or Python. 

ChatGPT has proven indispensable in simplifying the complexities of data manipulation and analysis. By utilizing a tool, such as ChatGPT, I have been able to refine my R knowledge by interacting with the AI and translate my 'wants and desires' into scripts in R. However, the adoption of conversational AI models does not replace the need for education or understanding of data analytics. These tools will still require human intervention to ensure a successful project. 

We will discuss the management of privacy concerns when handling confidential information. Emphasizing the importance of understanding the data privacy policies of AI tools is crucial for ensuring the security of sensitive data.

14. Growing Data Governance at MPTC

Presenter: Laura Waurio, Moraine Park Technical College

Description: This session will cover the structure of Moraine Park Technical College's Data Governance Cross-Functional Team, including its purpose, objectives, and membership. The presentation will feature accomplishments of the cross-functional team and continued areas of effort. The presenter will showcase the data governance toolkit and self-assessment on the Wisconsin Technical College System (WTCS) resource page. The WTCS self-assessment is divided into four sections: Data Understanding, Data Quality, Data Security, and Strategic Data Use. Key content will include a comparison of self-assessment results from Spring 2021 compared to Spring 2024, highlighting causes for celebration and how the data have been used to inform future plans.

15. Effectively Using Continuous Improvement and Innovation to Facilitate Organizational Change

Presenters:
Meridith Wentz, University of Wisconsin-Stout
Brenda Krueger, University of Wisconsin-Stout

Description: The University of Wisconsin-Stout has discovered the need for both continuous improvement and innovation to facilitate organizational change. Both are needed to effectively facilitate organizational change; however, the strategies and tools needed for each approach are different. We will discuss why some ideas do not take off and what can be done to provide more equitable access and support. We will guide you through the strategies and tools that we use for continuous improvement and innovation. We will examine the organizational challenges the approaches are designed to address and share how the approaches fit in with our FOCUS2030 Strategic Plan and are applied across the organization. 

1:30 - 2:15 pm

16. Data Visualization Dashboard Showcase

Dashboard: Dashboards galore! Presenter: Christopher Petrie, Northwestern Health Sciences University 

Dashboard: Enhancing Survey Awareness: Visualizing Survey Calendars with Power BI Presenter: Ian Dahlinghaus, MSOE

Dashboard: The St. Olaf NSSE and HEDS Alumni Survey Dashboards Presenter: Kelsey Thompson, St. Olaf College 

Dashboard: Survey Hub Presenter: Ling Ning, Universities of Wisconsin

17. When Python Met R: A SASsy Love Story Unfolds!

 Presenters: 
Alyssa Brinkley, SAS Education Manager
Josh Sheinberg, SAS Customer Advisor

Description: Fall head over heels for data analytics in "When Python Met R: A SASsy Love Story Unfolds!" In this romantic tale of code collaboration, we introduce the ultimate trio—SAS, Python, and R—coming together in a unified, cloud-based programming playground. While R is fashionably late to the party (joining us next year), we'll showcase the seamless chemistry between SAS and Python as they tackle data ingestion, cleaning, and modeling with high-performance flair and trustworthy results. Discover how this dynamic duo (soon to be a terrific trio) ignites collaboration across diverse teams, making workflow harmony more than just a fairy tale. And for those who prefer point-and-click adventures over coding quests, we'll sprinkle in the magic of SAS Viya. Whether you're a data cupid or an analytics adventurer, this session promises a whirlwind romance with the future of institutional research!

18. The CIP Wizard and the Philosopher's Stone

Presenter: Allan Joseph Medwick, Ursidae Analytics, LLC

Description:  This presentation is designed for institutional researchers, educational administrators, policymakers, and assessment professionals involved in program development, strategic planning, and resource allocation in higher education institutions. It is particularly relevant for those who use the Classification of Instructional Programs (CIP) codes in their work. 

CIP codes play a critical role in how educational programs are classified and funded, impacting everything from student enrollment to institutional funding and accreditation. This presentation will provide valuable insights into the recent updates in CIP 2020, and offer a glimpse into the upcoming changes slated for 2025 and 2030. By understanding how institutions use these codes to align with market demands and benefit from policy changes (e.g., gaming STEM OPT), attendees will gain practical knowledge that can be applied at their institutions.

19. From State to Local: Generalizable Methodology for Setting Performance Targets

Presenter: Russell Dahlke, Minnesota State System

Description:  Attendees will learn about Minnesota's approach to setting postsecondary state-determined performance levels (SDPLs) for the core indicators in their 2025-28 Perkins V state plan. This methodology, while developed for Perkins V, is generalizable to a wide range of strategic planning and target-setting contexts. It covers how Minnesota established state targets and how these were used to determine local targets. Important considerations were differences in institutional characteristics such as size, performance levels, and trends in performance along with aligning methodologies to federal legislative requirements. 

Key elements of the methodology include the use of linear regression and the construction of index numbers, making the approach accessible and straightforward to implement. Participants will be able to adapt these techniques to various contexts, ensuring their applicability beyond Perkins V. Metrics covered include the three Perkins V postsecondary core indicators of postsecondary placement in jobs and continuing education, completion rates, and enrollment in gender non-traditional programs determined by CIP code. 

By the end, participants will gain strategies and tools to enhance their target-setting processes, assess institutional performance, and ensure compliance with legislative requirements. This presentation aims to equip institutional researchers with a scalable methodology that can be applied across different educational and strategic planning contexts.

2:30 pm - 3:15

20. 5 Tips and Exercises for Improving Your Productivity With R

Presenter: Tyler Platz, Southwest Wisconsin Technical College

Description:  As a data analyst in institutional research, I am always trying to find ways to maximize time spent on high-impact work. I carefully manage my calendar and (admittedly) multi-task in meetings. Even with new AI tools designed reduce repetition in tasks I engage in, there are many remaining opportunities I have leveraged to improve my productivity through automation. 

Some of my largest productivity gains come from my use of R as a general-purpose tool rather than a simple statistical programming language. Going beyond performing t-tests and creating data visualizations, I use R to manipulate folders, clean common data files, create and edit Excel spreadsheets, and short-cut tasks all while documenting my workflow in the readable, reproducible format of an R script. To supercharge my productivity, I invest time in creating code short-cuts (known as snippets) and repeatable, custom functions (known as wrappers) to reduce duplication when coding. 

I believe everyone, not just data analysts, can benefit from the productivity gains that come from using R as a general-purpose tool. My goal in this workshop, through engaging in hands-on exercises, is to allow participants to leave with tangible tools in addition to a new perspective on how to hack their productivity.

21. Utilizing Labor Market Data to Explore Cutting Edge Opportunities for Graduates and the Skills Needed to Succeed 

Presenter: Melissa McKenney, Lightcast

Description:  Lightcast will be presenting on how to use labor market data to identify areas of rapid growth, new and emerging jobs and skills, and how to ensure there is program alignment for these new opportunities. Institutions will be able to use this knowledge and data to be on the forefront in order to teach students the necessary qualifications and skills needed to succeed in these rapid growth roles. Lightcast will also present on how to identify where these opportunities are located for potential employer partnerships, internships and transitional skills for future career advancement. 

22. Leveraging Predictive Analytics for Admissions: The Lead Scoring Project at UW-Eau Claire

Presenters:
Casey Rozowski, University of Wisconsin-Eau Claire
Colton Lunemann, University of Wisconsin-Eau Claire

Description:  In Fall 2021, UW-Eau Claire joined the Common App, making it easier for high school students to apply. This led to a nearly 50% surge in applications, overwhelming the admissions department and decreasing the yield rate by about 30%. To address these challenges, we initiated the Lead Scoring Project to leverage predictive analytics for identifying serious applicants. 

We developed two predictive models using data from Slate and PeopleSoft, which contain applicant information, and the National Student Clearinghouse (NSC). The logistic regression model predicted the likelihood of an applicant depositing at UW-Eau Claire, achieving an accuracy of 79.9%. Significant variables included campus visits, incoming credits, and ACT scores. The machine learning model, utilizing a random forest algorithm, provided high true positive rates for predicting enrollments and differentiated non-deposit types. 

These models streamlined the admissions process, reducing staff workload and increasing the yield rate. The results of the logistic regression were integrated into Slate, providing admissions with easy-to-use predictors. Future plans include updating the models with new enrollment data and integrating financial aid information to further refine predictions. This project demonstrates the transformative potential of predictive analytics in higher education admissions.

23. The Demographics of Student Loans:  Understanding Differences in Student Debt Burden by Sex and Race and Ethnicity of Minnesota College Graduates, 2021-2022

Presenters:
Maxwell Herrera, Minnesota Office of Higher Education
Nicole Whelan, Minnesota Office of Higher Education

Description:  Student loan debt continues to be an area of interest at all levels of policymaking. At both the state and federal level, there is an ongoing conversation around college affordability and access to higher education for all Americans. Nationally, Minnesota ranks high in overall educational attainment, but significant disparities exist for historically excluded student populations, including those who identify as part of the BIPOC communities. For Minnesota to remain globally competitive, the state's education systems need to ensure every higher education graduate is on track to pursue the education and training necessary for careers of the future. As such, the intersection of access, disparities, and debt has been a concern for researchers, lawmakers, and other community stakeholders. By better understanding these distinctions, policymakers have needed information to enhance future state policies for higher education. The data presented in this presentation represents an analysis of Minnesota students by award levels and sectors for the student's 'highest obtained award' level during the 2021-2022 academic year. This presentation examines differences in median cumulative loan debt and the percentage of students with debt at graduation amongst students of different sexes, races, and ethnicities.