Predictors of engagement with mathematics and statistics support

This paper investigates demographic predictors of engagement with mathematics and statistics drop-in support at Coventry University which has an extensive and well-used mathematics and statistics support provision. The analysis focuses on students from a set of 12 disciplines where there is clear mathematics and/or statistics content. It explores the effect of gender, entry requirements, course stage, ethnicity, age, nationality and disability. Two dimensions of engagement with mathematics and statistics support are examined: the binary dimension of engaged (at least once) or not and, for those who did engage, the number of times they did. These two dimensions are modelled through a two-stage hurdle model using a binomial logistic regression model to predict engagement and a negative binomial distribution model to predict the number of visits made by a student who does engage. These results provide valuable insights into how higher education institutions can tailor their support depending on the demographic population and engagement at their universities.