Predicting HE higher and higher

Academic success at university could depend on the changing interaction between students’ habits over time rather than fixed traits such as intelligence or total study hours. This conclusion is discussed in the International Journal of Computational Systems Engineering in a paper that challenges the conventional methods of predicting and measuring educational success.

In the research, the team looked at why some students consistently perform better than others and have developed a statistical model that treats learning behaviour as dynamic rather than static. The study suggests that standard approaches to educational analysis commonly overlook the fact that student routines, motivation, and workloads change during their time at university. Student habits frequently fluctuate in response to deadlines, stress, extracurricular commitments, and changing levels of engagement. Moreover, these factors influence each other dynamically from term to term, and static models cannot, by definition, take this into account.

The research used an extended linear regression model to estimate how strongly particular variables, such as attendance, study time, and motivation, affect examination results or scores. One of the clearest findings from this kind of analysis involved cramming before examinations. Educational advice often portrays intensive last-minute revision as inherently inferior to consistent long-term study. The study’s findings suggest a more nuanced relationship. Short-term intensive study was associated with stronger immediate improvements in results than long-term study habits alone. However, the researchers stress that cramming was only really effective when supported by stable routines and regular review throughout the term. The study also found that too many extracurricular activities reduced the effectiveness of cramming by limiting both available time and mental energy.

The study raises questions about how educational institutions understand student achievement. Universities frequently rely on static indicators such as attendance rates, exam results, and cumulative study hours when assessing academic potential. The researchers argue that these measures may overlook the importance of timing, behavioural change, and the interaction between short-term and long-term learning strategies.

Huang, R. (2026) ‘Analysis of factors affecting college students’ academic performance based on linear regression’, Int. J. Computational Systems Engineering, Vol. 10, No. 8, pp.1–13.