Abstract:

In-class quizzes are often used as a formative assessment tool, to assess student learning, facilitate feedback and identify to what extent learning is taking place. However, do in-class quizzes have any measurable effect upon learning, or are they just for the teacher’s benefit? Does it matter whether students get the answer right or wrong in classroom polls, or is it simply the act of taking part which is important for student learning?

This case study will show how the simple task of in-class engagement with classroom quizzes can be demonstrated to enhance students learning and performance in final summative assessment. Attendees will learn about the importance of using classroom polls to drive classroom engagement, the student demographics who are less likely to respond, the importance of regular polling through a course, and the importance of ensuring that students continue to keep answering polls through the course.

We monitor the responses to classroom polls in a series of 6 workshops of a large undergraduate finance course with over 1000 students. In each workshop, we invited students to answer a qualitative question and a quantitative question, and monitored students' responses. We link this to a pre-course demographic questionnaire and students' final exam results, and draw a number of useful inferences about the use of in-class polling:

1. We identify a strong, positive relationship between answering in-class quizzes and final exam mark; we find that the act of just responding has a significant effect on final exam mark, making the effect of two whole grades' differences between 0% and 100% response.
2. We identify a strong, positive relationship between answering questions correctly and the final exam mark, but it depends whether the questions are quantitative or qualitative. In the case of quantitative questions, it matters much more whether the student gets the answers correct. The additional effect of just trying, but getting the answer wrong, doesn't provide any significant extra increase in final exam mark.
In the case of qualitative questions, instead, it matters much more whether the student just tries to answer the question. The additional effect of getting the answer correct, doesn't provide any significant extra increase in final exam mark.
3. We find that poll responses in the second half the course matter more, that is it is not sufficient for students to be engaged at the start of the course, but rather, they need to continue to remain engaged through to the end of the course.
4. We also identify the particular student demographics who may need to be encouraged more to respond in class: we need to encourage younger students, UK-origin students, male students and native language students to respond more.

Theory:
The closest study to ours in Einig (2013), who finds a statistically significant correlation between regular MCQ usage and higher examination performance, controlling for a number of other confounding variables. However, that study only uses a binary variable of engagement to predict exam performance, which takes the value 1 when students complete 5 or more sets of MCQs, and zero otherwise. By contrast, our study includes the exact number of questions answered.

More widely, our study contributes to the literature on Formative Assessment, e.g. Johansson et al (2021), Marriott and Lau (2008), Paschal (2022), William and Thompson (2007).

Keywords:

Classroom Polling, Formative Assessment, Feedback, Quizzes.