Universities are often treated as independent and autonomous entities. After all, it is much simpler to attribute their success or failure in fulfilling their mission—whatever that might be—to their independent actions. Not unlike the view that firms fail or succeed because of their leaders’ weaknesses or foresight, universities are often evaluated by their ability to graduate students on time (using the graduation rate metric) and their reputational scores, and institutions often link these outputs directly to their leaders, faculty actions and intent.
This view is at best incomplete and in most cases incorrect.
Each university is a system that takes in high school graduates (at least for its undergraduate portion) for input and, through a regulated and complicated process, produces its own graduates four, five or six years later. Just like any dynamic system, its output depends both on its input as well as its own initial conditions and internal dynamics. In the parlance of dynamic systems, these are called the natural and forced responses. For linear systems, one can actually decompose the output into the sum of two terms: one due to the input (the forced response), and one due to the system itself (the natural response). Such decomposition is not generally possible for nonlinear systems but the general argument remains valid. This proposition can be easily verified using the data currently available for many universities.
At my own university, using the data collected over multiple years on our incoming students’ characteristics (e.g. High School GPA, ACT score, etc.), I can easily verify that increasing the selectivity of applicants will allow us to adjust the graduation rate to any desired level (Heileman 2015). Of course, the price paid is limiting access by reducing the number of admitted students. This will come as no surprise to most, but the point of the exercise is to show that the university can achieve almost any desired output (e.g. graduation rate) if the input is properly selected. The more useful conclusion is to show that adjustments to the university process itself, while maintaining the same input characteristics, leads to improved outcomes. This also can be proved using our data, where a curriculum is streamlined, advising is optimized, and the interventions are matched to the students’ characteristics. In fact, for similar students’ characteristics, one can demonstrate much better graduation rates from two universities, depending on their internal characteristics. Better yet, one need only examine the performance of student athletes versus the general student population at any large public flagship university. The two groups share in their academic characteristics, but proceed along different paths; student athletes, while juggling their academic and athletic duties, are provided with more academic support and thus graduate at higher rates.
So what is the point of considering the university as a dynamic system? It is to highlight that the output is dependent on the interplay between the input and the internal dynamics. Highly selective colleges filter out students—affecting the output portion due to the external inputs— while effective universities are adapting their own characteristics to affect the output portion due to their internal dynamics. The two knobs need to be concurrently adjusted.
However, the above discussion misses one important aspect: simply focusing on the easy-to-measure-and-manipulate graduation rate is both shortsighted and limiting. In fact, the value-added by any university should be measured by the learning that takes place within its boundaries, and the ability of its graduates to perform in their chosen fields. “Learning”—a student’s learning experience and their academic results—is again affected by the interplay between the input characteristics (an academically better-prepared student will take better advantage of a university’s learning experience) and the university’s internal dynamics (a more challenging curriculum, academic support). The one caveat is that measuring learning remains elusive at least in a predictive manner. In other words, it is difficult to predict that a certain graduate of a specific school will eventually perform at a pre-specified level. One must rely on delayed measurements (alumni and employers’ surveys) to determine the effectiveness of specific interventions and curricula. Recent approaches (CLA+, Purdue-Gallup index, OECD, etc.) are attempts to measure what we truly value.
The same conceptual framework may be interpreted differently by reversing the roles of the student and the university: a student is a system with its own initial conditions, while the university experience provides the inputs needed to achieve a desired outcome. Thus again, the graduation rates, the learning outcomes, etc. become the combined outputs resulting from the interplay between the student’s initial conditions and the university’s interventions.
This discussion is not meant to provide a mathematical description of the dynamics of learning, but only to illustrate that focusing our attention (or ire!) within the university’s boundaries misses the big picture. On the other hand, attempting to improve naïve measures, such as graduation rates, leads to the perverse incentives of filtering out a large number of incoming students and decreasing the number of eventual college graduates. Only by focusing on the input and the dynamics concurrently can we achieve the ultimate goal of improving the quality and quantity of the universities’ outputs.
– – – –
References
Heileman, G.L., Babbitt, T.H., & Abdallah, C. T. (2015). Visualizing student flows: Busting myths about student movement and success. Change: The magazine of higher learning, Vol. 47, No. 3, (May/June 2015). 30–39.
You Might Also Like
