Evaluating technology-based interventions to enhance the learning of first year Statistics students

Abstract

This retrospective study focuses on the unique contribution of the various technology-based interventions to improve the learning experience and performance of first-year statistics students. Eight consecutive cohorts from 2011 to 2018 were divided into five intervention cohorts for comparison of performance measured by the examination mark, of which four included the initial introduction of a technology intervention and one formed the base year. The other three cohorts were non-initial technology intervention years.

Departing from a traditional teaching model in 2011 the technology-based intervention strategies implemented were firstly the Aplia™ interactive online homework system in 2012 and then, in addition, the flipped classroom in 2013. The flipped classroom as pedagogical model, grounded in constructivism, substituted the traditional teaching model. The impact of the flipped classroom versus the traditional model was measured and the role of the online homework system in both teaching models. QT-clickers were subsequently introduced in 2015 to enable active and cooperative learning for face-to-face engagement inside the classroom. The pedagogical influence of the QT-clicker and the effect of partial grade crediting were investigated. Due to students’ different personality types and learning styles it was decided in 2018 to finally implement and evaluate peer learning activities using South African content and lecturer-switching within the flipped classroom.

To make the consecutive cohorts comparable, it was decided to divide the students in each cohort into two samples; the prerequisite sample consisting of only new students with at least 60% for Grade 12 mathematics and an AP Score of at least 26, while the remaining students formed the non-prerequisite sample. The non-prerequisite sample was further divided into new and repeat students.

The results from the various graphs, ANOVA and chi-square tests confirmed that the differentiator between the technology-based interventions is pedagogy. The flipped classroom model cohorts outperformed the traditional model cohorts. The uniqueness of this study is that it illustrates the benefits of the online homework system, QT-clickers, and peer learning activities through the lens of a flipped classroom. The positioning of the online homework system as a self-paced preparation tool concentrates on a student-centred environment, with self-directed learning and self-knowledge, helped to control cognitive overload. This study further demonstrates the pedagogical influence of the QT-clicker and the effect of partial grade crediting in a large, flipped classroom setting. Most of the new students in both prerequisite and non-prerequisite samples benefited from the flipped classroom and QT-clicker usage, with the stronger students’ final marks being moved into higher percentage brackets. Surprisingly, most of the repeat students also gained from the flipped classroom and peer learning activities with localised content and lecturer-switching, which indicate their need for alternative learning activities and styles.

Several general linear models (GLM) and a machine learning algorithm, XGBOOST (XGB), were fitted to the data to investigate how the technology-based interventions are related to the students’ examination performance. The outcome of the GLM models and XGB indicates that the association of higher examination marks hold for the flipped classroom cohorts, i.e., 2013, 2015 and 2018. A qualitative component of the study reports on the student voice testifying to positive experience.