Fiction in the space between II: you use data but it doesn't mean, you're not just telling stories7/24/2024  The University of North Carolina System Board of Governors today unanimously approved the program eliminations at UNCG and UNC-Ashville. My personal opinion matters not, but I don't think it was particularly smart for UNCG to eliminate physics while trying to build a small nanoscience school (nano without physics?), or to not have a physics program beyond teaching introductory physics in a world where climate change, energy technology, and artificial intelligence are the future. It's also kind of counter intuitive that one of our strongest PhD programs, Computational Math, was cut with a narrative, based on a fictional, or at least data-free, story that undergraduate teaching will be better without the distraction of a relatively strong and distinctive doctoral program and that cutting the doctoral program will save money (that is its own complicated story because of the role of doctoral students in teaching and mentoring undergraduates).
Tracy Chapman in her song "telling stories" sings "There is fiction in the space between; the lines on the page of your memories; you write it down but it doesn't mean; you're not just telling stories." I think you could change the last two lines to read "you use data but it doesn't mean; you're not just telling stories." That would sum up why I get so angry about the stories that are being told about higher education by higher education leaders and consultants like rpk. For example, WFAE quoted in their story the UNCG Chancellor's remarks regarding program elimination (I did not hear the remarks, so I am just reacting to this one quote): "In his address to the board, Gilliam pointed out that UNCG’s math department has the same number of tenured or tenure-track faculty as psychology — a major that has ten times the number of students." A colleague who had access to some data sent me the data for academic year 2022-2023 (apparently the most recent year) and compared Math/Stats vs Psychology. Based on the data I was sent, the chancellor was correct about a 10-fold difference in the number of undergraduates majoring in the two disciplines, but those data don't tell the story the quote seemed to be intended to support. Here are the data from the 2022-2023 academic year that I was sent so cannot fully verify: The department of math/statistics is the first number and the second number is the department of psychology.
So, yes, these data I was sent suggest that the chancellor was correct regarding the number of majors. The rest of the data show that the two departments in 2022-2023 are roughly the same in the number of credit hours they taught and the number of credit hours taught per faculty member, Since universities generate revenue from student credit hours and not majors, a reasonably good interpretation of that data would be that Math and Psychology should have roughly the same number of faculty. If one looks only at the net revenue number, perhaps the most important number with respect to financial stability, one might conclude that math might be a better investment than psychology with respect to return on investment. But, net revenue data are complicated. For example, psychology majors also take credit hours in math/stats, and vice versa, so they contribute to each other's positive net revenue. *(see footnote below regarding how complex university budgets are and a recommendation for a terrific book that explains them so well). The data the chancellor presented on majors was also a red herring with respect to program closures. UNCG is not eliminating all Math/Stat undergraduate degrees (at least not in the program eliminations the BoG approved on July 24, 2024) and there are no plans that have been announced to reduce the size of the math faculty (perhaps that is coming) In fact, by eliminating the computational math PhD program, it is may be possible that math/stats may need to hire more faculty or staff to replace the teaching and mentoring done by Math graduate students. The moral of that story is that the chancellor, at least in that one quote, "used data but it doesn't mean, he's not just telling stories." To be fair, though, this was the only quote I heard that was related to program eliminations. I was pretty surprised that his team didn't find a better anecdote of data that support one of the actual cuts. Two of my favorite books are "How to lie with statistics (a 1954 classic that is as true, if not more so, today)" and "How to lie with maps" (I was a colleague of the author at Syracuse Universities in the early 1990s, so read it very early when the first edition was released in 1991). It is sad that these books are so relevant to higher education, an industry that should know better. The case with psychology and math is a fitting example. You can use essentially the same group of data to argue three very different perspectives. You can look at student credit hours and you can reasonably declare, "Math should have the same number of faculty as psychology!" Look at the number of majors and you can reasonably declare "Math should only have 10% of the number of faculty as psychology!". Or you can look at net revenue and reasonably declare "We need to invest in math!" Although not relevant to this particular example, how data are presented can also affect their interpretation (e.g., the scale one uses on the Y-axis on a bar chart can exacerbate or minimize differences, leading to different conclusions with the same data. The moral of the story is that it is easy to spin, obfuscate or lie with data, even when one doesn't intend to (e.g., by not interpreting data in the context of other data, or letting computer software define the scale of of an axis on a graph). Data are inherently objective but they are inanimate. They come to life when they are interpreted subjectively by somebody. Although data are inanimate, I love them anyway (I don't have a pet rock, though). They ignite my curiosity more than they confirm my stories. In fact, they more often than I would like to acknowledge, make me admit that I was wrong. Data interpretation can be very humbling- any conclusions are always tentative in the face of new data or new interpretations. I often said as an administrator that nobody should ever be afraid of data; it's just data. The more transparent the better, even if it counters a narrative, because more people looking at data, asking questions about it, reanalyzing it, etc., will lead to more ideas being generated and perhaps less fiction in the space between a narrative and reality. Reducing the fiction in the space between narrative and reality means that more learning is occurring. But data become dangerous when one defines their purpose simply to support a story. When the narrative is more important than a complex truth to someone, then the only good data to them are data that support that narrative. Ignoring or being intellectually lazy with data leaves infinite hectares of fiction in the space between. And, nothing good happens in that space. I tend to think that adjusting one's narrative when the subjective interpretation of data is counter to the narrative is a better approach than ignoring the data or finding ways to essentially lie with it. We are all subject to confirmation bias. And we all suffer from the evolution of our cognitive abilities that makes it easy to tell a story with little data, and much more difficult and energy consuming to adjust the story when the data might suggest that one should do so (this is paraphrasing a point I took away from reading Daniel Kahneman's "Thinking Fast and Slow"). But we should try. I mean every university I know has critical thinking skills as a foundation of their general education curriculum, and in many ways, critical thinking is about adjusting a narrative to data, not forcing data into a pre-existing narrative. It would be nice if academic leaders and political leaders and boards (and everybody else including me) modelled this approach. Justifying controversial decisions is a challenging task and the tendency will be to find any data that can justify a decision, even if there are a lot of other data that are contrary, or, as in the statement the UNCG chancellor made, choosing to rely on data to justify a decision that aren't relevant to the actual decision being justified, at least on that day. Admittedly, that does not mean the decisions were wrong, but people who see incongruencies between data and decision get frustrated, perhaps angry, and often lose trust. Not many bettors would put their money down on a ship getting through a storm (or terrible headwinds to use an overused metaphor) when the crew responsible for sailing the ship, doesn't trust the captain and their leadership team. I will just try to stay humble about how I interpret data and will try to remember my adaptation of Tracy Chapman's lyrics for the remainder of my career (and in thinking about every political statement I ever hear or read)- "There is fiction in the space between; the lines of the page of your memories; you use data, but it doesn't mean; you're not just telling stories". I hope that readers will give me coordinates for my GPS should they find me lost in my own fiction in the space between me and reality, and I hope that every once in a while my blog does that for you. * University budgets are a very complex system (I encourage everyone to read Former U. Arizona provost and current chief academic officer for the Arizona Board of Regents Andrew Comrie's, book "Like Nobody's Business: An Insider's Guide to How US University Finances Really Work)." Also, in general, cutting expenses by eliminating academic programs only increases net revenue if those cuts do not lead to similar decreases in student credit hour revenue. One way to consider this in estimating future net revenue is to only consider the drop in student credit hour revenue directly associated with programs that were eliminated. The challenge in the projection, though, is that it is harder to estimate any decreases in enrollment that occur because of a decline of a UNCG's reputation due to perceived declines in academic quality or budget issues relative to peer schools (reputation's tend to change slowly), increases in student:faculty ratio, perception of prospective students and their parents that the school focuses on efficiency of credit hour delivery over academic quality, and/or loss of programs that may not attract majors but have courses that non-majors want to take (e.g., astronomy, a certain language, archeology). It is also hard to estimate positive changes in net revenue that might occur from investing in other programs. For UNCG, currently there is a relatively small proportion of programs where student demand exceeds supply. Those programs where demand exceeds capacity are generally expensive to teach perhaps because they require low student:faculty ratios (e.g., Genetic Counseling, Nursing, Music Performance at UNCG), so don't necessarily generate a lot of positive net revenue with increased enrollment.
0 Comments
Leave a Reply. |