Estimating the optimum number of attempts for any university examination: a negative binomial approach

This study sought to provide an empirical answer to the question of how many attempts should be allowed to sit for a given University module. We analyzed data from Makerere University, spanning five admissions from 2007 to 2011, for 14 selected modules. We modeled the number of attempts students made on examinations as a count variable using the negative binomial distribution within a Bayesian framework. Our results indicate that while there are differences in the number of attempts across student and academic-specific factors, the number of attempts generally does not exceed three. The study concludes that, given the main objective of university education—to produce well-prepared and competitive graduates, a policy of a maximum of three attempts appears to be reasonable and prudent since it balances the need for student success while maintaining academic rigor.