Nationally, only about half of students who start college actually complete and earn college degrees, according to the National Student Clearinghouse. It is much worse for students who start at a community college or 2-year degree program. This means many students are increasing their debt (from student loans) and not reaping the financial benefits of a college degree and higher earnings. In addition, colleges are losing money when students dropout. It’s a lose-lose game.
Colleges have been getting increasing pressure from the federal government and others to improve their college completion rates. They have turned to data analytics to better understand how they could intervene earlier with students who might be at risk of not completing their degree and dropping out.
When I first heard this podcast from The Hechinger Report on how “Colleges are using big data to track students in an effort to boost graduation rates, but it comes at a cost” it made me want to listen closer. As a #datalover and #datanerd, this topic certainly peaked my interest. I am also a first generation college graduate and can personally relate to this topic of college dropouts since many (over half) of my high school peers didn’t finish their college degree.
This is not a new strategy (“predictive analytics”) as companies like Amazon and Google are doing this all of the time with our digital footprint. However, it’s new for colleges and non-profit organizations.
Georgia State is a case study example of how their university has embraced data analytics to improve college completion rates. Their university now has one of the highest rates of college graduation for public universities in the country, and they have closed the racial equity gap. Students of color are graduating at the same rate as white students at Georgia State.
So how did Georgia State get there? They used their data to drive and inform program improvement. We’ve talked about this here and here.
The podcast also shares the struggles that other universities face in implementing these changes. It’s not enough to purchase the data analysis software, but you also need trained staff who are able to analyze and interpret the data to take action on it.
We’ve put together 3 tips to get started with using your data to drive program improvement that’s not only based off of the success of Georgia State but our work with other clients who want to improve their impact.
- Have a system in place to collect and track meaningful data. Georgia purchased a data system to help them bring all of their data together and identify patterns. At Transform Consulting Group (TCG) we are big fans of using Tableau Software (see more here and here). We love Tableau, because it can make your data easier to review and understand. We are also adept at using whatever data systems our client has to pull out the information that we need to inform decision-making.
- Have trained staff who can analyze and interpret the data. Getting the data from your system is the first step. Then you need to have individuals who know how to identify patterns, ask inquisitive questions and develop recommendations. We are big fans of forming an “Impact Team” at an organization who is trained on analyzing and reviewing your data and can help drive action based on the results. Learn more here.
- Have a process to determine program improvement changes. Once you have your data and results, you are now at the fun part – you get to take action on changes to make to improve your results! We find that sometimes organizations don’t have a clear process in place to determine what changes – such as a new curriculum, staff training, client outreach – they will make based on the results. We follow a Continuous Quality Improvement Process using the “Plan – Do – Study -Act or (PDSA)” framework to determine what action steps we will take. Other similar processes might be “Lean/Six Sigma”. At Georgia State they decided to hire additional advisors to use the data to reach out to students at risk. This is one of the many changes they have made, based on their data results, to help students stay on track with college graduation. The goal here is to be intentional on how you will use your data to make programmatic changes and then study the result of those changes to know if it’s working or not.
What are some goals that you are not currently satisfied within your organization? Have you considered how data analytics could help you focus in on improving your results? Bill Gates said, “Without measurement there is no shared accountability.“ We would love to help you improve your impact through data analytics. Contact us to learn more.
We worked with a group of community leaders in Jay County to complete the feasibility of converting an old elementary school building into an early childhood center. Like many rural communities, Jay County has a declining population that has impacted their local schools in continuing to operate multiple school buildings, which has resulted in school consolidations and closures. At the same time, their rural community also struggles with attracting new employers due to a lack of child care for a growing workforce. Their community leaders had the idea of converting a closed elementary school into an early childhood center but wanted assistance in completing a feasibility study first.
Does your evaluation process accurately represent the population you serve? The marginalized and less privileged? We shared in this 
shared examples about how equity is undermined based on traditional evaluation practices. She then shared an overview of the core equitable evaluation principles. The work of evaluation is for and contributes to equity. Evaluations should answer critical questions about the effect of a program on populations, underlying drivers of inequity, and why history and culture might be affecting the work and potential changes. In addition, evaluation design and follow-through should be based around the values of equity, meaning multi-culturally valid and set toward participant ownership.
Indiana’s Early Learning Advisory Committee (ELAC) released its 
have multiple sites and/or facilitators. During observations, it’s crucial to have a specific rating sheet or checklist of what you should expect to see. If a program has a fidelity checklist, you can use it during observations! If not, you should create your own rubric.