Memorizing marketing definitions through real-time feedback in an intelligent tutoring system

Abstract

Teachers in Dutch bilingual Marketing programs at MBO level have several difficult tasks. Among these is preparing students for their exams, which often requires the students to memorize definitions of marketing concepts in English. As these students prefer listening, watching and practical work to reading [Scheltinga et al., 2020], and as English is not their native language, they often struggle to memorize these definitions to any degree of accu-racy. With personnel shortages on the rise, pandemic lock-downs and increased adminis-trative workloads, teachers have less and less time to engage in one-on-one tutoring with students [Kim et al., 2022], while such tutoring has been proven to be the most effective means of helping students learn [Bloom, 1984].
For repetitive tutoring tasks such as helping individual students memorize a definition, a solution might be found in the form of an Intelligent Tutoring System (ITS) specifically tailored to that task. In this paper, we present MemorAIze, a prototype built to test the effectiveness of a Natural Language Processing (NLP) based ITS for students who need tu-toring to memorize Marketing definitions in a foreign language. MemorAIze incorporates several basic gamification elements and was designed to provide immediate feedback to students while they are typing their definitions, interrupting them as a human tutor would, to correct and encourage them as they reproduce memorized text.
Using Regular Expressions (RegEx), MemorAIze can take students through three steps in a specific memorization strategy [Little and McDaniel, 2015] tailored to developing vo-cabulary. These steps are:
1. Expose the learner to keywords
2. Put the keywords in the correct order
3. Form coherent sentences with the keywords
It encourages students to find a set of important words, then helps them form the cor-rect coherent sentence with those words. The tool aims to help students remember [Bloom, 1956] text, but offers no route to understanding the text.
In this study, we evaluate MemorAIze as a teaching tool. We assign students the task of memorizing and exactly reproducing Marketing related definitions, then expose them to MemorAIze as a digital tutor. While we found no significant data on long-term learning ef-fects based on this experiment, we do see that short-term learning effects can be achieved using this tool when short definitions are used. We also gather several suggestions for im-provement of the tool that can be used in later studies.

Date of Award	2 Feb 2023
Original language	English
Supervisor	Bastiaan Heeren (Examiner) & Johan Jeuring (Co-assessor)