Volume 15 Number 2

MELBOURNE 2017

93

World Transactions on Engineering and Technology Education

Editorial

The organisation of the 5th Mediterranean Seminar on Engineering and Technology Education is in its final stages. The Seminar will be held at the Piraeus University of Applied Sciences (PUAS) in Piraeus-Athens, Greece, between 11 and 15 September 2017. More than 20 papers from authors in 10 countries worldwide have been proposed, and are included in the Seminar’s preliminary programme and scheduled for presentation at the Seminar. The paramount objective of the Seminar is to address and discuss a multitude of issues, problems, opportunities and challenges that contemporary academia faces in an effort to prepare future technologists and engineers capable of taking on the increasingly advanced, complex and sophisticated technology within the contexts of the changing world and professional practices. It is envisaged that several significant issues will be raised by the Mediterranean Centre for Engineering and Technology Education (MCE&TE), co-organiser of the Seminar with the World Institute for Engineering and Technology Education (WIETE), with a view to cultivating the knowledge and skills essential for high quality maritime and environmental engineering and technology education.

We are indeed fortunate that the Seminar enjoys the patronage of Professor Lazaros Vrizidis, President of the Piraeus University of Applied Sciences (PUAS), formerly known as Technological Educational Institute (TEI-Piraeus).

Our readers may have noticed the Call for Papers on the WIETE site concerning the organisation of the next annual meeting of the WIETE. This is for the 9th WIETE Annual Conference on Engineering and Technology Education, under the major theme Effective Methods in Engineering and Technology Education. The principal objective of the Conference is to bring together partners, members, associates, supporters and friends of the WIETE from all over the world, and to continue discussions on issues of importance to engineering and technology education, with particular emphasis on the prevalent theme. The Conference will be held between 19 and 23 February 2018 at Cinnamon Residence in Bangkok, Thailand. It has been decided that the presented papers will be published predominantly in WIETE’s World Transactions on Engineering and Technology Education (WTE&TE), Vol.15, No.4 or Vol.16, No.1, but participating colleagues will have the option of designating their papers to be published in the Global Journal of Engineering Education (GJEE), if they so desire. It should be stressed that both WIETE journals are indexed by Scopus.

WIETE members, associates and sympathisers are cordially invited to propose a paper and attend the Conference. It is envisaged that those colleagues born in 1943 will have the opportunity of getting together with the WIETE Director over a gala dinner to celebrate the 75th year of their productive lives. An after-conference excursion (3 days, 2 nights) will also be offered at a minimal cost to those colleagues and their spouses wishing to visit Pattaya, a world-class seaside resort in Thailand. Interested parties will find the Conference announcement, including a call for contributions, at: http://www.wiete.com.au/conferences/9wiete/index.html

I am pleased to advise our readers that the present issue of the WTE&TE, marked as Vol.15, No.2, includes 18 highly informative articles, with most of them coming from Indonesia (11). The other articles have come from countries such as the People’s Republic of China (3), Israel (1) and Taiwan (1). There are two collaborative articles, one coming from Russia, Canada and the United States of America, and the other from Slovenia and Egypt. It is an interesting observation that there is growing interest by Indonesian academics to publish articles on engineering and technology education, which may indicate a new era for Indonesian higher education in this important area of academic endeavour.

I wish to thank the authors for their contributions to this issue and their willingness to share their research and development achievements with likewise minded academic colleagues. Sincere thanks are extended to the referees for their prompt and efficient assessment of the articles. Also, I would like to express my gratitude to the WIETE editorial team that on this occasion included Dr Dianne Q. Nguyen, Mrs Dorota I. Pudlowski and Dr Ian R. Dobson, for their invaluable assistance in the release of this issue.

Zenon J. Pudlowski

Volume 15, Number 2, 2017

Melbourne 2017

World Transactions on

Engineering and Technology Education

Editor-in-Chief Zenon J. Pudlowski

World Institute for Engineering and Technology Education (WIETE) Melbourne, Australia

90

World Transactions on Engineering and Technology Education

Association of Taiwan Engineering Education and Management

(ATEEM) Taipei, Taiwan

Commonwealth Science and Technology Academy for Research

(C-STAR) Chennai, India

Faculty of Architecture Slovak University of Technology

in Bratislava Bratislava, Slovakia

Polytechnic Institute of New York University

New York, USA

Piraeus University of Applied Sciences

Piraeus-Athens, Greece

University of Botswana Gaborone, Botswana

Published by:

World Institute for Engineering and Technology Education (WIETE) 141 The Boulevard, Ivanhoe East, Melbourne, VIC 3079, Australia

E-mail: info@wiete.com.au Internet: http://www.wiete.com.au

© 2017 WIETE

ISSN 1446-2257

Articles published in the World Transactions on Engineering and Technology Education (WTE&TE) have undergone a formal and rigorous process of peer review by international referees.

This Journal is copyright. Apart from any fair dealing for the purpose of private study, research, criticism or review as permitted under the Copyright Act, no part of this Journal may be reproduced by any process without the written permission of the publisher.

Responsibility for the contents of these articles rests upon the authors and not the publisher. Data presented and conclusions developed by the authors are for information only and are not intended for use without independent substantiating investigations on the part of the potential user.

91

World Transactions on Engineering and Technology Education

EDITOR-IN-CHIEF

Prof. Zenon J. Pudlowski World Institute for Engineering and Technology Education (WIETE) 141 The Boulevard, Ivanhoe East, Melbourne, VIC 3079, Australia Tel/Fax: +61 3 94994339, E-mail: zenon.pudlowski@wiete.com.au Internet: http://www.wiete.com.au

MANAGER & ASSOCIATE EDITOR

Dr Dianne Q. Nguyen World Institute for Engineering and Technology Education (WIETE), Melbourne, Australia

ASSOCIATE EDITORS

Dr Ian R. Dobson Monash University, Melbourne, Australia Mrs Dorota I. Pudlowski World Institute for Engineering and Technology Education (WIETE),

Melbourne, Australia Ms Krystyna Wareing Independent Communications Consultant, Crawley, West Sussex, England, UK

ADVISORY BOARD MEMBERS

Prof. Hosni I. Abu-Mulaweh Indiana University-Purdue University Fort Wayne, Fort Wayne, Indiana, USA Dr Stanislav Avsec University of Ljubljana, Ljubljana, Slovenia A/Prof. Kristijan Breznik International School for Social and Business Studies; College of Industrial

Engineering, Celje, Slovenia Prof. Samuel C-C. Chang National Taiwan Normal University, Taipei, Taiwan Prof. Colin U. Chisholm Glasgow Caledonian University, Glasgow, Scotland, UK A/Prof. Chih-Hsien Huang Ming Chi University of Technology, New Taipei City, Taiwan Prof. Ahmad Ibrahim Yorkville University, Toronto, Ontario, Canada Prof. Krzysztof Kluszczyński Silesian University of Technology, Gliwice, Poland Prof. Romanas V. Krivickas Kaunas University of Technology, Kaunas, Lithuania Prof. Sabina Kuc Cracow University of Technology, Kraków, Poland Dr Wojciech Kuczborski Private Consultant & WIETE, Perth, Australia Prof. Viljan Mahnič University of Ljubljana, Ljubljana, Slovenia Prof. George Metaxas Piraeus University of Applied Sciences Piraeus-Athens, Greece Prof. Andrew Nafalski University of South Australia, Mawson Lakes, Adelaide, Australia Dr Zorica Nedic University of South Australia, Mawson Lakes, Adelaide, Australia Prof. Derek O. Northwood University of Windsor, Windsor, Canada Prof. Robert Špaček Slovak University of Technology in Bratislava, Bratislava, Slovakia Prof. David W.S. Tai HungKuang University, Taichung, Taiwan Prof. Steven Thatcher Central Queensland University, Cairns, Australia Prof. Jacek Uziak University of Botswana, Gaborone, Botswana Prof. Algirdas V. Valiulis Vilnius Gediminas Technical University, Vilnius, Lithuania

92

World Transactions on Engineering and Technology Education

Contents

Z.J. Pudlowski Editorial 93

T.P. Tsai, J. Lin, L.C. Lin & J. Li A blended learning lesson design for an EPUB3 e-book-based course 94

Nizaruddin, Muhtarom & Sugiyanti

Improving students’ problem-solving ability in mathematics through game-based learning activities

102

K. Breznik & H. Rezk The mission statements of public research centres in Egypt 108

L.S. Riza, R. Awaludin, H. Sutarno, Munir & A.P. Wibawa

A model for auto generating sets of examination items in educational assessment by using fuzzy c-means

114

A. Gero, Y. Stav & N. Yamin Use of real world examples in engineering education: the case of the course Electric Circuit Theory

120

T.Y.E. Siswono, A.W. Kohar, A.H. Rosyidi & S. Hartono

Primary school teachers’ beliefs and knowledge about mathematical problem-solving and their performance in a problem-solving task

126

M. Ayub & O. Karnalim Predicting outcomes in Introductory Programming using J48 classification 132

Y. Luo, X. Li & R. Yin The dictionary use strategy for writing in English by engineering students - a case study

137

H. Hendriana Teachers’ hard and soft skills in innovative teaching of mathematics 145

M.A. Mizar, M. Amin, M. Ashar & Marsono

Performance test of machine groundnuts husk peeling using a rotary multi-disc system - a case study

151

Y. Zhu, Y. Chen, T. Pan & H. Liu SPIED: an international maker education practice of China, Japan and Korea 157

R.D. Mahande & Jasruddin Utilisation study of mobile technology at a vocational high school 162

Y. Zu & M. Xu Research on the teaching reform of a Single Chip Microcomputer course based on the concept of CDIO

169

N. Amaliyah, Sapriya & E. Maryani

A trans-disciplinary approach and inquiry-based learning model of social studies

174

Ratnadewi, R.P. Adhie, Y. Hutama, J. Christian & D. Wijaya

Implementation and performance analysis of AES-128 cryptography method in an NFC-based communication system

178

M. Kuimova, D. Burleigh, X. Maldague & D. Startseva

Academic exchange programmes to enhance foreign language skills and academic excellence

184

S. Maf’ulah, D. Juniati & T.Y.E. Siswono

The aspects of reversible thinking in solving algebraic problems by an elementary student winning National Olympiad medals in science

189

H. Upu, Djadir & S. Asyari The fifth graders’ mathematisation process in solving contextual problems 195

Index of Authors 200

World Transactions on Engineering and Technology Education 2017 WIETE Vol.15, No.2, 2017

132

INTRODUCTION

Introductory Programming is a course, which becomes a key factor for computer science (CS) students, especially first-year students who have just completed high school and wish to continue their journey through the CS major. After spending several weeks on the Introductory Programming class, some students realise that they can hardly understand programming logic and syntax due to their lack of computational thinking ability. Consequently, it may generate an issue in student retention in the CS major. Some students tend to give up on their CS major since they think that they have not assimilated such skills. In addition, despite the fact that most students may pass the course, they are not guaranteed to pass advanced programming courses in the next semester, if they do not pass this course with distinguished results. Recognising the difficulty of programming, a course is proportional to the assigned semester of a given course. The higher its assigned semester is, the more difficult its course material is.

There have been several pieces of research, which aimed to identify the issues experienced by students in learning programming, especially in the introductory programming course. Byrne and Lyons [1], and Bergin and Reilly [2] have studied some of the factors, which influence the success of novices in the introductory programming course; namely, previous computing experience and prior academic performance. According to Rountree et al, decision tree classifier is used to identify combinations of factors that interact to predict success or failure in the introductory programming course [3]. Also, a study by Wiedenbeck et al has shown that self-efficacy and the mental model have a direct effect on overall success in an introductory programming course [4]. A study by Wilson indicates that a formal class in programming and game playing …promote success in an introductory computer science course, there is a significant gender difference particularly for game playing [5].

In this study, prediction attributes that determine the success of novice students in computer programming will be explored further. Such prediction will be conducted using data mining technique, particularly the J48 classification technique. As a case study, this work incorporates student data from the Introductory Programming course, which was held in the Computer Science (Informatics) major, at Maranatha Christian University. The attributes cover four aspects, which are personal, prior education, admission and assessment data. The result of this study is expected to become a supplementary data source for handling the student retention issue. For instance, providing a more-sophisticated learning method or course syllabus based on high-valued prediction attributes. Moreover, these data can also be used by university admissions in terms of student recruitment. They can filter the students recruited, based on given prediction patterns.

CLASSIFICATION IN DATA MINING

The data mining task can be divided into two categories, descriptive and predictive [6]. The classification method is a predictive data mining task, which is defined as a predictive method that is used to classify unseen data [6][7].

Predicting outcomes in Introductory Programming using J48 classification

Mewati Ayub & Oscar Karnalim

Maranatha Christian University Bandung, Indonesia

ABSTRACT: In a computer science (CS) major, Introductory Programming becomes a substantial course, which determines whether students can complete that major or not. This study evaluates the correlation between student data with the students’ capacity to pass that course. Such correlation is exploited according to a data mining technique called J48. For each student, the work incorporates personal, prior education, admission and assessment data. Based on an evaluation of 41 pieces of student data, the national test score for mathematics in Indonesia presents the most promising attributes, followed by the admission test score. The results of this study are expected to provide a brief insight for CS lecturer and the university, so that they can handle emerging issues in CS education, especially, the low retention rate.

133

The predictive model is generated based on the analysis of a training data set. Prediction of a new data should be done using the model. There are some techniques that can be used for classification, such as decision tree induction, Bayes classification or rule based classification [6]. In this study, the authors applied decision tree induction as a classification technique; namely, J48 classification. The J48 classification is Weka’s implementation of C 4.5 decision tree learner. Weka implements a later and slightly improved version, which called C4.5 revision 8 [7].

For each leaf in a decision tree derived from the J48 classification, there are two numbers (n/m), which mean that n instances reach the leaf, but m instances are classified incorrectly [7]. The percentage of correctly classified instances will determine whether the generated model is sufficiently good. When the amount of data for training and testing is limited, stratified tenfold cross validation techniques can be used to ensure the performance of the classifier [7].

Rules in the IF-THEN form can be extracted from a decision tree. Each path from the root to a leaf node can be written as one rule. The rule antecedent (IF-part) is formed by combining the splitting criterion along a given path using the AND connection. The leaf node which contains the class prediction forms the rule consequent (THEN-part) [6].

METHODOLOGY

The class under study was a first-year course; namely, Basic Programming. It is based on the Python programming language in procedural style programming. All students had entered the course directly from high school. After the students have completed the course, they should be able to specify, design, code and test a computing solution. The course was composed of two sessions: theory of 150 minutes’ duration and laboratory practice of 210 minutes’ duration. These sessions were conducted once a week during that semester. In term of course material, Basic Programming was divided into logic and advanced-technique material.

Logic material was taught in the first seven weeks (before a mid-test). It covered data types, variables, conditional statements and looping. On the other hand, advanced-technique material was taught during the last seven weeks (after the mid-test). It covered many advanced techniques that are frequently used in programming. Such techniques include functions, arrays, searching and sorting. The assessment of this course consists of a mid-semester written examination (25% marks), a final written examination (25% marks), mid-semester and final laboratory examinations (25% marks), and 12 weekly laboratory assignments (25% marks).

In this case, the data set has been extracted from 41 students who have a minimum 75% of attendance for the Basic Programming course. Each piece of student data consists of personal, admission, prior education and assessment data. Pre-processing for these data was done by resolving inconsistencies, and transforming data to obtain quality data that are feasible for classification.

As described in Table 1, a group of attributes has been selected for student classification. These attributes consist of:

a) personal data, such as gender and student’s home town (province);b) prior education data, such as high school major, national test score for mathematics (NTM);c) admission data, such as admission test score (ATS);d) assessment data from basic programming course, such as final written examination score (WES) and final

laboratory examination score (LES). The gender, province, major, NTM, ATS attributes will be utilised to predictWES or LES.

Table 1: Student data set.

Attribute name Description Possible values Gender Student’s gender [M=male, F=female] ATS Admission test score [E : excellent, G : good, F : fair] Province Student’s home town [J = Java, L = outside Java] Major High school major [1 = major A, 2 = outside major A]

NTM National test score for mathematics [E : excellent, G: good, F : fair]

WES Written examination score [E : excellent, G : good, N : not passed] LES Laboratory examination score [E : excellent, G : good, N : not passed]

Based on the student data set, this study explored the final examination scores as class attributes against personal data, prior education data and admission data through classification technique. The experiment was performed twice, one for final written examination scores and the other for final laboratory examination scores.

RESULTS AND DISCUSSION

In Table 2, the data set is grouped according to gender, major, province, ATS and NTM. Descriptive statistics (means and standard deviation) for each group are shown in Table 2.

134

Table 2: Descriptive statistics.

Grouping n WES LES

Means SD Means SD Gender Male 33 56.12 24.88 53.45 19.54

Female 8 48.75 18.07 43.00 19.73 Major Major A 28 59.00 20.65 55.00 17.54

Outside major A 13 45.38 27.83 43.69 22.75 Province Java 31 56.13 25.33 53.68 20.75

Outside Java 10 50.20 18.07 44.40 15.23 ATS Excellent 13 67.23 23.85 61.77 20.34

Good 25 52.68 19.36 49.08 16.69 Fair 3 17.00 10.58 26.00 16.37

NTM Excellent 4 88.75 13.15 76.50 19.28 Good 20 61.30 18.82 56.35 15.34 Fair 17 38.88 18.44 39.71 17.20

To examine the differences between WES and LES for each of the group attributes, the authors used a t-test for gender, major and province. For example, the WES mean for male students was compared to the WES mean for female students. In carrying out the t-tests, assumptions of normality and equality of variances were confirmed. In each group, the t-tests suggested no significant differences between any of the factors and the final examination scores.

The authors inspected the differences between final examination scores for the ATS group and the NTM group using the one-way Anova test. The results of the Anova tests showed that there was a significance difference (p-value < 0.01) between the attributes’ value in both groups. Table 3 indicates the F-value for each group with a critical F-value of 5.21. They used MaxStat Lite version 3.6 to calculate descriptive statistics, t-tests and Anova tests.

Table 3: F-value of Anova tests for ATS and NTM.

Grouping WES LES F-value F-value

ATS 7.59 5.40 NTM 14.62 9.82

This study further explored the relationship between final examination scores with the group attributes using the J48 classification of the students’ data set using WEKA version 3.8.1 as a data mining toolkit. The first classification was done using written examination score (WES) as the attribute class. The second classification used laboratory examination score (LES) as the class. Because of limited data, this study utilised a ten-folds cross validation to ensure the validation of the results.

The result of the first classification is shown in Figure 1 in the form of a J48 pruned tree with 70.73% accuracy in WES predicting. There are 29 correctly classified instances and 12 incorrectly classified instances. Figure 1 indicates that the NTM attribute is the most affective attribute in predicting written examination score (WES). The result of WES is accordingly determined based on NTM. If NTM is excellent, then, WES is excellent. Similarly, if NTM is fair, then, WES is not passed. However, if NTM is good, major and gender attributes also contribute to predict the WES. Table 4 resumes the rules generated from the J48 decision tree in Figure 1.

Figure 1: J48 pruned tree for WES predicting.

135

Table 4: Rule for WES.

Rule # Rule’s premise WES

Percentages of instances Excellent Good Not passed

1 IF NTM = Good and Major = 1 and Gender = Male

- 71.43% -

2 IF NTM = Good and Major = 1 and Gender = Female

- - 50%

3 IF NTM = Good and Major = 2 - - 50% 4 IF NTM = Fair - - 88.23% 5 IF NTM = Excellent 75% - -

Figure 2: J48 pruned tree for LES predicting.

The result of the second classification is shown in Figure 2 in the form of the J48 pruned tree with 58.54% accuracy percentage of LES predicting. There are 24 correct classified instances and 17 incorrectly classified instances. Figure 2 indicates that the NTM attribute is also the most affective attribute in predicting the laboratory examination score (LES). The result of LES is accordingly determined based on NTM. If NTM is excellent, then, LES is excellent. Similarly, if NTM is fair, then, LES is not passed. Nearly the same as predicting WES, if NTM is good, gender attributes also contribute to predict the LES. Table 5 reports on the rules generated from the J48 decision tree in Figure 2.

Table 5: Rule for LES.

Rule # Rule’s premise LES

Percentages of instances Excellent Good Not passed

1 IF NTM = Good and Gender = Male - 68.75% - 2 IF NTM = Good and Gender =

Female - - 75%

3 IF NTM = Fair - - 70.59% 4 IF NTM = Excellent 50% - -

CONCLUSIONS

Student success or failure in the Introductory Programming course will influence the continuity of their study in computer science. In this study, five attributes were used to predict outcomes in introductory programming: gender, student home town (province), high school major, national test score for mathematics (NTM) and admission test score (ATS).

Written examination score (WES) and laboratory examination score (LES) are used as class attributes. The statistical test results of this study show two predictive factors in the following order of importance: national test score for mathematics and admission test score. Using J48 classification, the authors obtained a confirmation that the national test score for mathematics becomes the most impactful prediction attribute. From the J48 decision tree, the combination of factors to predict success or failure can be resumed.

ACKNOWLEDGMENT

The authors would like to acknowledge the financial support provided by the Maranatha Christian University Research Committee, by means of the Maranatha Christian University Grant.

136

REFERENCES

1. Byrne, P. and Lyons, G., The effect of student attributes on success in programming. Proc. ITiCSE 6th AnnualConf. on Innovation and Technol. in Computer Science Educ., New York, USA, 49-52 (2001).

2. Bergin, S. and Reilly, R., Programming: factors that influence success. Proc. 36th SIGCSE Technical Symp. onComputer Science Educ., Missouri, USA, 411-415 (2005).

3. Rountree, N., Rountree, J., Robins, A. and Hannah, R., Interacting factors that predict success and failure in a CS1course. Proc. ITiCSE Conf. on Innovation and Technol. in Computer Science Educ., Leeds, United Kingdom, 101-104 (2004).

4. Wiedenbeck, S., Labelle, D. and Kain, V.N.R., Factors affecting course outcomes in introductory programming.Proc. 16th Annual Workshop of the Psychology of Programming Interest Group, Carlow, Ireland, 97-110 (2004).

5. Wilson, B.C., A study of factors promoting success in computer science including gender differences. ComputerScience Educ., 12, 1-2, 141-164 (2002).

6. Han, J., Kamber, M. and Pei, J., Data Mining Concepts and Techniques. Waltham: Elsevier Inc. (2012).7. Witten, I.H., Frank, E. and Hall, M.A., Data Mining Practical Machine Learning Tools and Techniques.

Burlington: Elsevier Inc. (2011).