Using Data Mining Techniques Examination of the Middle School Students’ Attitude towards Mathematics in the Context of Some Variables

Abstract

The aim of this study is to examine middle school students" attitude towards mathematics in the context of their mathematic learning preferences using data mining which is data analysis methodology that has been successfully used in different areas including educational domains. "How do I actually learn?" questionnaire and attitude scale were applied to 702 middle school students studying in three different cities of Turkey. Demographic data (gender, grade level, parents" education level, pre-school education) were also gathered. Data mining techniques such as decision tree was implemented. Furthermore web graph was used for visualization of relationship between mathematic learning preference and attitude towards mathematics. Constructed decision tree models with C5.0 algorithm revealed that all of variables used in this study are related to the attitude towards mathematics but the most effective attribute is found as grade level. Using web graph, it was found that the strongest relationship was between reflective learning preference and positive level attitude towards mathematics.

Students’ Activity in an Online Learning Environment for Mathematics: The Role of Thinking Levels

“…This is indeed our approach in the current study, relying on logged data from an online, applet-based environment for elementary school mathematics, and analyzing it to study differences in students' behaviors when attempting to solve questions of different thinking levels; the latter variable was collected manually and was triangulated with the data from the log files. As for the importance of age in the context of mathematics education -potentially impacting students' achievements either because of their psychological development or attitudes towards mathematics -and as age was not included in the data we analyzed, the current study includes this variable, mediated by grade level (Idil et al, 2016;Zhou et al, 2019).…”

Educational Data Mining Studies in Turkey: A Systematic Review

“…According to the results in Table 5, the topics studied in EDM are investigation the causes of students' failures (Baran & Kilic, 2015;Birtil, 2011;Ucgun, 2009), text mining (Afacan Adanir, 2019;Akcapinar, 2015;Sohsah et al, 2015), determination of variables that affect attitude (Hark, 2013;Idil et al, 2016), determination of reasons for absenteeism (Dalkilic & Aydin, 2017), estimation of instructor performance (Cifci et al, 2018), determination of familial variables affecting reading skill (Avsar & Yalcin, 2015), prediction of the department that students will prefer (Coskun, 2013), modeling of video navigations (Akcapinar & Bayazit, 2018) and prediction emotional states with facial recognition and (Ayvaz et al, 2017).…”

“…When Table 4 is analyzed, it is observed that 46.77% of the studies are for prediction task, followed by the tasks of classification, clustering and association rules respectively. Yelegin (2012), Saygili (2013), Sengur (2013), Akcapinar (2014), Aksoy (2014), Ozbay (2015), Uysal (2015), Cebi (2016), Sahin (2018), Guruler et al (2010), Sohsah et al (2015), Idil et al (2016), Ayvaz et al (2017), Afacan Adanir (2019) 24.19 Prediction Aydin (2007), Y. Aydogdu (2011), Sengur (2013, Bahadir (2013), Yildiz (2014), Akcapinar (2014), Coskun (2013), Uysal (2015), Yildiz Aybek (2016), Cebi (2016), Barngrover (2017), Yagci (2018), Yorganci (2018), Kentli and Sahin (2011), Turhan et al (2013), Akcapinar (2015), Demir (2015), Kayri (2015), Sohsah et al (2015), Bahadir (2016) Note.…”

“…Another technique commonly used in EDM studies is the decision trees. It is seen that decision trees are used in the studies for estimating student achievement (Aydin, 2007), identifying the variables affecting the attitude towards a specific purpose (Idil et al, 2016) and classifying students according to certain characteristics (Guruler et al, 2010;Yelegin, 2012). Other techniques commonly used in EDM research include clustering techniques (Avsar & Yalcin, 2015;Baran & Kilic, 2015;Dogan & Camurcu, 2010), regression (Bahadir, 2016;Coskun, 2013;Turhan et al, 2013), association rules techniques (Hark, 2013;Onan et al, 2016;Yucel, 2012), bayes classifiers (Cebi, 2016;Cifci et al, 2018;Sahin, 2018;Uysal, 2015), and support vector machines (Kentli & Sahin, 2011;Sohsah et al, 2015;Tekin, 2014).…”

The Mathematics Learning Styles of Vocational College Students

“…An important research that characterizes students' learning preferences in mathematics is also proposed by Forster (1999). Forster (1999) considers a model according to four fundamental dimensions helping educators to plan learning environments in mathematics course (Idil, Narli & Aksoy, 2016): reflective, inquisitive, diligent, and user.…”