Investigation of Prospective Elementary Mathematics Teachers’ Learning Styles and Relationships between Them Using Data Mining

Abstract

One of the general discussion in the studies about learning style is what degree of students whose learning style is determined, have other learning styles. In this context, the aim of this study is to determine the learning styles of prospective elementary mathematics teachers and to explore the relationships between these styles by using data mining techniques. Data mining can be defined as applications of different algorithms to identify patterns and relationships in a data set. For this purpose, Grasha-Reichmann Learning Styles Inventory was applied to 400 prospective elementary mathematics teachers at Dokuz Eylul University. Cronbach's alpha reliability coefficient of the scale was found as 0.83.Results show that more than 50% of female students have "independent'' learning style. At the same time students who have competitive learning style had the least number of students. The male students who have collaborative and dependent learning styles were the majority.. From Class 1 to Class 4, it was observed that the number of students who have individual learning styles was decreasing and the number of students who have cooperative learning styles was increasing. In network graph, it was found that one of the strongest relationships was between the students who have cooperative and independent learning style with high level. On the other hand the relationship between the students who have passive and independent learning style with low level was not seen in graph. The decision tree indicates that the most effective attribute is independent learning style to identify which level of the learning style students have. Besides in the Data mining, learning styles, Mathematics Education association rules model several rules are constructed with %75 confidence. Extended SummaryWith the transition to learner-centered education, approaches related to teaching and learning began to be questioned and developed in education system. The fundamental basis of the change is individual features of students and effect of these features to educational environment.Teachers can use individual differences in planning learning environments and processes (Tomlinson, 1997: 25). These individual differences consist of a lot of feature such as learning style. Kolb (1984) defines learning style as a preferred way of gathering information, whereas for Dunn (1984) learning style is an individual way of absorbing and retaining information or skills. Özer (1998) reported that teachers must know students' learning styles for effective learning.

The role of individual differences on epistemic curiosity (EC) and self-regulated learning (SRL) during e-learning: the Turkish context

“…This limitation prompted researchers to study in-depth for new analysis methods. As a result of these studies, a new data analysis method, data mining has emerged, which enables the analysis of data from different angles and to summarize this data by converting it into useful information (Delavari et al, 2008;Narli et al, 2014). The researchers defined the data analysis method as the process of discovering meaningful information from data stacks using methodologies such as artificial intelligence, statistics, and machine learning (Tan et al, 2006;Aran et al, 2019).…”

“…Data mining includes different analysis models within itself. Many studies have categorized these models in different classifications (Ayık et al, 2007;Baker & Yacef, 2009;Baradwaj & Pal, 2012;Delavari et al, 2008;Luan, 2002;Narli et al, 2014). The most general definition of the association rule is categorized in the descriptive model which tries to reveal which events can occur simultaneously by examining the relations of the variables in the dataset with each other.…”

A Model Proposal to Determine Learning Styles of Students by Using Machine Learning Techniques and Kolb Learning Styles Inventory

“…Truong (2015) is another researcher who presented a proposal for the development of a system that classifies learning styles through students' online learning behaviors. Narlı, Aksoy, & Ercire (2014) aimed to determine learning styles of students and to explore the relationships between them. For this purpose, Turkish translation of Grasha-Riechmann Student Learning Styles Inventory was applied to prospective elementary mathematics teachers at Dokuz Eylül University.…”