What is student success? Let’s start with the most fundamental definition: completing and receiving a degree. For decades, national completion rates have hovered around 20-30 percent in three years for an associate’s degree and 50-60 percent in six years for undergraduate degrees.
Here’s the problem: The data is not actionable and is enabling low-expectations creep, such as setting expectations that a six-year completion rate is a success metric for first-time full-time students in a four-year program. Low-expectations creep can best be summed up by Eberhardt Rechtin:
“High expectations, because they are unlikely to be fulfilled, define failure… low expectations, because they are likely to be accomplished, define success.”
Our messy definition of success
1. It is very hard to find applied definitions of student success that are not, at the heart, derivatives of failure or risk. A classic student-success initiative begins by collecting data on dropouts and uses that as a proxy for reducing failure, rather than for scaling success.
4 reasons why student success is misdefined in higher ed and how data can fix it
2. When risk data is analyzed to personalize success plans, too often socio-demographics such as income, zip code, gender, and race become the focus. A very real consequence of this practice is that risk labels are assigned to student sub-groups before they even arrive at an institution. By age 17 or older, students are savvy at recognizing how they are being esteemed by an ecosystem, so this approach is limiting and counter-productive for both the student and the institution.
3. Many definitions deconstruct the student experience into a single process or point of retention. For example, randomly asking representatives from admissions, retention, and career services departments for their definition of or data on student success yields unique, short-term answers.
4. Rarely do definitions incorporate student expectations, or that of their families. Success is a grade, a score, or something done to them, rather than with them.
(Next page: Why we need to collect data on a different subset)