The flipped classroom aims to improve learning by engaging students in educational activities outside of traditional lessons. Flipped classrooms have steadily gained popularity in the last decade and are a topic of discussion in teaching and learning forums. However, its adoption in mathematics and statistics has been subdued. Most higher education mathematics and statistics are still delivered through traditional lectures where students are passive participants. In this study, experiences of flipping a large first-level statistics class are presented. The implementation included a combination of peer learning and tutor-assistance in lectures. Student performance, in the form of final examination and overall marks over four semesters (two with traditional delivery and two flipped), were analysed for differences with respect to the two teaching modes after adjusting for demographic differences. In addition, student survey data were analysed with a view to revealing any relationship between attitude towards a flipped classroom and performance. The results showed that students' performance improved and an increased understanding of concepts was achieved through the flipped classroom approach. Evidence also indicated an increase in learner engagement. Student feedback indicated a higher preference for a flipped mode overall and in particular for ages 20 and below.