In business schools, there have long been terms for two kinds of successful students: “quants” and “poets.” Quants are the finance and accounting whizzes, deft with numbers and technical systems. Poets shine in marketing and management, good with words and with understanding people.
Educational research about business schools has largely ignored this folk belief, equating success with overall grade point average. GPA is a convenient metric and often the only one available to researchers, but it collapses different kinds of academic performance into one dimension, potentially obscuring important differences in how students think, learn, and lead.
Michael Morris, the Chavkin-Chang Professor of Leadership at Columbia Business School, has long emphasized the role of social intelligence. His recent best-seller Tribal shows how social smarts underlie culture and cooperation—for early humans who formed the first tribes and for us in our complex organizations today—so it is key to leading groups of all kinds.
For a decade, Morris ran Columbia’s Program on Social Intelligence, which worked across different parts of the school (admissions, core classes, electives, clubs, career services) to foster social skills. His new research paper is based on some uniquely rich datasets about MBA students compiled during this work.
Together with Aaron Wallen, a professor at Columbia University’s School of Professional Studies, and Zachariah Brown, an assistant professor at the Hong Kong University of Science and Technology, Morris analyzed the dimensionality of MBA grades in core classes. Their research corroborated the MBAs’ self-view that success comes in two dimensions: socially oriented and technically focused. Further, consistent with two kinds of intelligence, the researchers found that social and technical success correlate with different aptitude scores (verbal and quantitative GMAT tests) and different pre-MBA experiences and interests. Their analyses also revealed leadership to be more associated with the social dimension of grades and more predicted by attendant aptitudes—but not in all the ways that admissions officers expected.
The findings have big implications for both business schools and companies choosing who to bring on board.
Two Paths to Success
To put student achievement under the research microscope, the team analyzed two unusually comprehensive datasets. The first was an exit survey of a graduating class, where students self-reported grades and their pre-MBA scores and experiences. A factor analysis of grades revealed that academic performance was well captured by two dimensions, technical and social course success, only weakly correlated. And consistent with a bivariate model of talent, these performance factors were differentially predicted by different aptitude measures (GMAT-Q and GMAT-V) and experiences. “This set of findings showed basic confirmation for the quant/poet model of MBA performance,” Morris says.
To corroborate this picture, the researchers turned to a more comprehensive and objective dataset. They drew on records from the classes of 2009 and 2010 that had been compiled in a consulting project for the school. The dataset incorporated records from offices across the school: admissions, career services, registrar, and student life. In addition to academic grades, leadership performance was assessed subjectively through classmate appraisals and objectively through attainment of positions in student organizations. In addition to the aptitude scores of GMAT-Q and GMAT-V, a writing section (GMAT-W) that was included in those years was also included in the dataset. Further, it included admissions application questions about interests, as well as a standardized vocational interest inventory used in career counseling. Past studies of MBA performance have relied on small samples and very limited metrics. This rich dataset was the educational research equivalent, Wallen suggests, of “paleontologists uncovering an entire dinosaur skeleton with skin attached rather just than the usual one or two desiccated bones.”
The second study replicated the findings of the first: MBA performance fell on two dimensions, technical and social. Once again, GMAT-Q predicted grades in technical classes, whereas GMAT-V predicted grades in socially oriented classes. Interestingly, GMAT-V also predicted grades in technical classes. Further predictiveness of academic success came from pre-MBA measures of interests. Wallen notes that interest level is often a factor in studies of elite performance. “Raw ability without the motivation and application doesn’t cut it,” he says
The additional success metric of leadership attainment tracked grades in social classes more than technical classes. It also shared the same antecedents, such as interests in social topics. (Interests in technical topics was a negative predictor). Additionally, GMAT-W scores emerged as an independent predictor, suggesting that not just verbal comprehension but also expression enables students’ emergence as leaders.
Implications for Admissions and Beyond
These studies have a few important findings for researchers, educators, and corporations.
For educational researchers, they challenge the reliance on overall GPA. Distinguishing technical and social courses enables a much clearer picture of the predictors and correlates of academic success. These studies also highlight interests as an additional indicator of potential.
Admissions offices would be well served by testing its hunches about “diamonds in the rough.” Some of the received lore within the MBA admissions profession about spotting leaders did not pan out in the performance data. Military experience and college athletics, for example, did not significantly predict leadership attainment. Business schools have the data—from admissions applications to alumni achievements—to identify the real indicators of leadership potential in their application pools.
For those running business schools, this research suggests the value of promoting multiple success metrics. “Instead of one ‘Dean’s List’ based on overall GPA, administrators should note success in different core classes to recognize different kinds of talent,” Morris says. After all, business schools aim to cultivate different kinds of talent needed for different business roles. Most core classes are technical subjects, but social intelligence may be more crucial for moving up. Compared to technical expertise, social expertise is also likely harder to replace with AI.
The message for companies is also clear. Some hiring procedures rely on simple metrics like GPA. Says Brown, “You can have two very different candidates with an identical GPA but not identical talent. For many companies or many roles, it is worth the effort to try to tell them apart.”
These implications are even more important given the increased use of automatized assessment. “Assessing talent is different than it was in our parents’ generation. We should be open to the idea that best practices change across the decades,” Morris noted. Added Wallen, “In the age of AI in the workforce, adeptness at social judgment may become an important differentiator of human employees as opposed to AI agents.”
