I.J.

Education and Management Engineering, 2015, 5, 12-19

Published Online October 2015 in MECS (http://www.mecs-press.net)
DOI: 10.5815/ijeme.2015.05.02

Available online at http://www.mecs-press.net/ijeme

A Decision Tree Approach for Predicting Students Academic

Performance
Kolo David Koloa, Solomon A. Adepojub, John Kolo Alhassanb

a
Department of Computer Science, Niger State College of Education, Minna
b
Department of Computer Science, Federal University of Technology Minna, Nigeria

Abstract

This research is on the use of a decision tree approach for predicting students‟ academic performance.
Education is the platform on which a society improves the quality of its citizens. To improve on the quality of
education, there is a need to be able to predict academic performance of the students. The IBM Statistical
Package for Social Studies (SPSS) is used to apply the Chi-Square Automatic Interaction Detection (CHAID)
in producing the decision tree structure. Factors such as the financial status of the students, motivation to learn,
gender were discovered to affect the performance of the students. 66.8% of the students were predicted to have
passed while 33.2% were predicted to fail. It is observed that much larger percentage of the students were
likely to pass and there is also a higher likely of male students passing than female students.

Index Terms: Prediction, Data Mining, Performance, Decision Tree, Academic.

© 2015 Published by MECS Publisher. Selection and/or peer review under responsibility of the Research
Association of Modern Education and Computer Science.

1. Introduction

Knowledge discovery in databases is the process of identifying valid, novel, potentially useful and ultimately
understandable patterns in data. Data mining is a step in the knowledge discovery process consisting of
particular data mining algorithms that, under some acceptable computational efficiency limitations, finds
patterns or models in data. In other words, the goal of knowledge discovery and data mining is to find
interesting patterns and/or models that exist in databases but are hidden among the volumes of data [6]. Data
mining is applied in various fields of human endeavors including education; this is referred to as educational
data mining. Educational data mining is a discipline, which is concerned with developing methods for
exploring the unique types and pattern of data that come from educational institutions as well as using those
methods to better understand students, and the settings which they learn in.
There have been increasing research interests in using data mining in education. This new emerging field,
* Corresponding author. Tel.: +2348035956346
E-mail address: [email protected], [email protected]
 A Decision Tree Approach for Predicting Students Academic Performance 13

called Educational Data Mining is concerned with developing methods that discover knowledge from data
originating from educational environments. Data mining methods or approaches includes Classifications,
Clustering, Naïve Bayesian, decision trees, neural networks and Fuzzy rules
All participants in the educational process (teachers, students, academic planners and administrators) could
benefit by applying data mining on the data from the higher education system. Data mining represents the
computational data process from different perspectives, with the goal of extracting implicit and interesting
samples trends and information from the data. So, it can greatly help every participant in the educational
process [1].
According to [17] the concept of educational data mining may be classified into the following classes:
Prediction which includes the following sub classes: classification, regression, density estimation; Clustering
which includes relationship mining, association, correlation mining, sequential pattern mining, causal data
mining; distillation of data for human judgment and discovery with models. The following are areas of
advantages of educational data mining in educational institutions [5]:

 Accurate analysis and visualization of data

 The ability to provide feedback for teachers/ instructors
 The ability to predict student performance
 The production of functional student behavioural models
 The ability to detect undesirable student behaviours
 It also helps to group students into classes based on various performance characteristics
 The ability to analyze the students‟ social networks
 Developing concept maps
 It also aid in the construction of curriculums
 It helps in the education planning and administration

Prediction of student performance is helpful in order to provide a student with the necessary assistance in the
learning process [3]. Knowing the possible outcome of the learning process based on results of prediction can
help an institution to change the outcome of the new set of students by adjusting the factors that contributed to
the past performance. The ability to predict students‟ performance can also aid educational planners and
administrators to adequately plan for the change in students population in any direction, i.e. increase of
decrease. However, it is usually difficult to come up with a manual set of rules that are needed to predict
students‟ performance. Hence there is a need to look for readily capable methods for dealing with the task. This
research has looked specifically at personal, social and economic factors that may affect students‟ performance
and used same to predict the performance of students in the coming session or semester as the case may be.
The objective of this study is to identify factors that will affect the academic performance of students and
predict the academic performance of students as either pass or fail by using the decision tree and ordinal
regression approaches of the SPSS software. Decision Tree technique has been found to be a very adequate
technique to generate a comprehensive and precise analysis. Decision trees are used in data mining to study
historical data and on the basis of the data analysis and its rules, one can predict the result [18]. SPSS combines
the features of data mining activities which is a combination of Statistical analysis and database management.
The SPSS package is easily available as a commercial off the shelf software and can easily be copied from one
computer to another due to its robust portability. It also has a good graphics user interface thereby making it
easy to learn and use by teachers/instructors who may want to use it for further analysis. The results are also
easy to view and interpret for decision making.

2. Review of Related Work

In order to predict the performance of students the researcher took into consideration the work of other
14 A Decision Tree Approach for Predicting Students Academic Performance

researchers that are in the same direction. Other researchers have looked at the work of predicting students‟
performance by applying many approaches and coming up with diverse results.
Three supervised data mining algorithms, i.e. Bayesian, Decision trees and Neural Networks which were
applied by [1] on the preoperative assessment data to predict success in a course (to produce result as either
passed or failed) and the performance of the learning methods were evaluated based on their predictive
accuracy, ease of learning and user friendly characteristics. The researchers observed that that this methodology
can be used to help students and teachers to improve student‟s performance; reduce failing ratio by taking
appropriate steps at right time to improve the quality of learning.
[12] described the process of knowledge discovery from databases using a practical example of a current
actual problem. They developed two models based on decision tree which were successfully used to predict
student success based on GPA criterion and time student needs to finish the undergraduate program (time-to-
degree) criterion.
Bayesian classification method was also used by [4] in their work on student database to predict the students‟
grades on the basis of previous year performance. The researchers concluded that the study will help the
students and the teachers to improve the grades of the student. The study also helps to identify those students
which needed special attention to reduce failing ratio and taking appropriate action at right time.
[2] compared four different classifiers and combined the results into a multiple classifier. Their research
divided the data into three (3) different classes weighing the features and using a genetic algorithm to minimize
the error rate improves the prediction accuracy at least 10% in the all cases of 2, 3 and 9-Classes. In cases
where the number of features is low, the feature weighting worked much better than feature selection. The
successful optimization of student classification in all three cases demonstrates the merits of using the LON-
CAPA data to predict the students‟ final grades based on their features, which are extracted from the homework
data. However, the research in this case was based on an online course as opposed to the regular classroom
class that the present study considers.
Furthermore, [3] observed that in the problem of prediction of performance, it is possible to automatically
predict students‟ performance. Moreover by using extensible classification formalism such as Bayesian
networks, which was employed in their research it becomes possible to easily and uniformly integrate such
knowledge into the learning task. The researchers‟ experiments also show the need for methods aimed at
predicting performance and exploring more learning algorithms.
[8] addressed the prediction of secondary school students‟ performance in two core subjects of mathematics
and Portuguese by using their past score in the previous session and other demographic factors and employed
four data mining methods of Decision trees, Random Forests, neural networks and Support Vector machines
approach. The results show that the prediction was achievable provided the grades of the previous session were
known. This confirms that the prediction of students‟ performance is premised on past performance and hence
shows that a student‟s performance is closely related to the performance in previous course (most likely a
prerequisite course).
[11] concluded that Decision trees are so popular because they produce classification rules that are easy to
interpret than other classification methods. Frequently used decision tree classifiers are studied and the
experiments are conducted to find the best classifier for Student data to predict the student‟s performance in the
end semester examination. The experimental results show that Classification and Regression Tree (CART)
CART is the best algorithm for classification of data. From the study conducted by [7], by using a case study
on educational data mining to identify up to what extent the enrolment data can be used to predict student‟s
success. Two algorithms CHAID and CART were applied on student enrolment data of information system
students of open polytechnic of New Zealand to get two decision trees classifying successful and unsuccessful
students. The result obtained showed that the accuracy obtained with CHAID and CART was 59.4 and 60.5
respectively.
In their study [9] used classification task to predict the final grade of students. It was done by the use of ID3
decision tree method. Another study undertaken by [10] showed that Data Mining Techniques (DMT)
capabilities provided effective improving tools for student performance. The study further showed how useful
 A Decision Tree Approach for Predicting Students Academic Performance 15

data mining can be in higher education particularly to predict the final performance of student. The researchers
collected data from student by using questionnaire to find the relationships between behavioural attitude of
student and their academic performance. Data mining techniques were then applied. They obtained the
prediction rule model using decision tree as well as implementing the rules into Support Vector Machine (SVM)
algorithm to predict the students‟ final grade. Also the students were clustered into groups using kernel k-
means clustering. The study expressed the strong correlation between mental condition of student and their
final academic performance.
Also, [16] used Iterative Dichotomiser 3 (ID3) decision tree algorithm to predict the university students‟
grade of a university in Nigeria. A prediction accuracy of 79,556 was obtained from the model. They further
suggested the use of other decision based model to predict student‟s performance.

3. Methodology

The data collected was for a course CSC214 (Data Structures) which is a 2 nd year course for NCE in
computer science in Nigerian Colleges of Education. The course was chosen because of the familiarity of the
researchers with the course and the grades considered were the semester results for the course CSC214 for three
previous years. A questionnaire was also distributed to students to collect data about the other factors
considered in the prediction such as the students‟ financial strength and motivation to study. The data collected
from the result sheet was entered into SPSS Version 20 for analysis.
Table 1. Factors of the Prediction

Factor Description Attributes SPSS

value
Score Student‟s grade >40 and <40 0 – 100
Status Student‟s status Pass, Fail 1,2
Gender Students gender Male, female 1,2
Finance Financial strength Low, Medium, High 1,2,3
Motivation Attitude to learning Low, Moderate, High 1,2,3

The factors that are considered in predicting the performance of students in this study are;

 Scores: this is the total scores obtained by the students in the previous session of the course under
consideration. The scores‟ attributes are within the range of >40 and <40. The scores are however entered
into SPSS as obtained from the results sheets (i.e. 0 – 100). For the purpose of this study,the classification
was reduced from the six class classification of scores i.e. A, B, C, D, E, F to a two-class classification of
Pass and Fail in order to reduce the margin of error for the research as it regards to the prediction
 Status: Status refers to the remark obtained for the scores of the students, i.e. „Pass‟ or „fail‟. In which
case a score less than 40 is entered as fail while a score greater or equal to 40 is entered as pass. The
status is coded as 1 or 2; where 1 stands for pass and 2 stands for fail.
 Gender: Gender is used to classify the students as either male or female. This factor is quite important
because it helps to know the effect of gender of the students on predicting. This is in agreement with [13]
who established that gender plays a great role in the performance of female
 Students in a large public Turkish University. Though gender and age are predictors in the academic
performance of students, gender is a better predictor [14].
 Finance: this refers to the financial status of students while in school, this factor is important because it
affects the stability of the students as it relates to their comfort during the academic session. Finance is an
avenue trough which the students are able to settle their bill. Students without adequate financial strength
are affected adversely [15]. The attributes of the factor include, High, Medium mad Low are they are
16 A Decision Tree Approach for Predicting Students Academic Performance

coded in SPSS as 1, 2 and 3. Where 1 is for high, 2 is medium and 3 is Low.

 Motivation: this refers to a factor that pushes a student to work hard or serves as reinforcement for the
student; such as studying in groups, rewards and prices etc. the attributes are high, moderate and low.
These are coded as 1, 2 and 3; 1 for low, 2 for moderate and 3 for high.

In building the predictive decision tree using the SPSS program based on the criteria in Table 1, the SPSS
predictive approach used is the CHAID approach. The dependent variable is Score which implies that the study
is concerned with predicting the score of the student using the existing (previous) score. The factors that
influence that scores are listed under the independent variables as status finance and gender while the variable
chosen as the influence variable is the motivation.
Table 2. Model Summary

Growing method CHAID

Dependent Variable SCORES
Independent Variables STATUS, FINANCE, GENDER

Specifications Validation None

Maximum Tree Depth 3

Minimum Cases in Parent 100

Node
Minimum Cases in Child Node 50

Independent Variables STATUS, GENDER

Included
Number of Nodes 5
Results
Number of terminal Nodes 3

Depth 2

Fig. 1. Decision Tree Structure for Predicting Students‟ Academic Performance

 A Decision Tree Approach for Predicting Students Academic Performance 17

4. Results and Discussion

Fig 1 shows the decision tree for the prediction of the students‟ performance in the study. The first test was
based on the status of the students which was determined based on the number of passed and failed students.
The result shows that 66.8% passed and 33.2% failed and based on the principle of decision trees it shows that
the percentage of those who passed is higher and therefore the new consideration is the passed node (that is
why the next decision tree node is under the passed node).
The next factor for consideration is the gender (Male, Female) and based on the principle of decision tree
structure the gender division shows that 55.10% of the male students are predicted to pass while 59.03% of the
female is predicted in the predictive analysis. Therefore this shows that the study is showing that more female
students are predicted to do better in the course than the male students.
Table 3. Gain Summary for Nodes

Node N Percent Mean

4 42 19.9% 65.52
3 99 46.9% 60.81
2 70 33.2% 31.80
Growing Method: CHAID, Dependent Variable: SCORES

Table 3 shows the summary of the gain for the nodes in the predictive decision tree structures. It shows that
node tree has a higher gain than the other nodes at 46.9% which is an indication of the level of influence of the
third note factor on the predictive analysis of the students‟ performance.

5. Conclusion and Recommendations

The research has shown that some factors such as finance level, motivation level, gender and grades obtained
in previous (prerequisite) courses can affect the performance of students in academic institutions. This was also
established in some of the reviewed literature in this paper. More so, a more automated system can be used to
achieve the prediction of student‟s performance through data mining, by using software with such capabilities
such as SPSS used in this work. The research would have produced a clearer and more accurate result if a
higher version of SPSS was employed which could have shown a larger view of the tree model without
affecting the MS Word window. More research involving additional factors other than the ones used in this
research could still be carried out. This would have helped to have a clearer and more detailed predictive
analysis.
The researcher wishes to recommend that the public and private educational institutions in Nigeria should
pay more attention to efficient prediction analysis in education. This could be achieved by initiating and
funding research into the prediction analysis in education, as it will aid educational administration and planning.

References

[1] Osmanbegovic E., Suljic M. “Data mining approach for predicting student performance” Economic Review-
Journal of Economics and Business. Volume 10(1) (2012)
[2] Behrouz, M, Karshy, D, Korlemeyer G, Punch, W. “Predicting student performance: an application of
data Mining methods with the educational web-based system” Lon-capa. 33rd ASEE/IEEE Frontiers in
Education Conference. Boulder C.O. USA, (2003).
18 A Decision Tree Approach for Predicting Students Academic Performance

[3] Bekele, R., Menzel, W. “A bayesian approach to predict performance of a student (BAPPS): A Case with
Ethiopian Students”. Journal of Information Science (2013).
[4] Bhardwaj, K., Pal, S “Data Mining: A prediction for performance improvement using classification”.
International Journal of Computer Science and Information Security. Volume 9(4). (2011).
[5] Romero, C, Ventura, S. “Educational Data Mining: A Review of the State-of-the-Art. IEEE Transaction
on Systems, Man, and Cybernetics, Part C” Applications and Reviews. Volume 40(6) (2012).
[6] Bae, E., Bailey, J: “COALA: A novel approach for the extraction of an alternate clustering of high quality
and high dissimilarity”. Proceedings of the Sixth International Conference on Data Mining. Pp. 53 – 62.
(2006).
[7] Kovacic, Z. “Early prediction of student success: Mining student enrollment data” Proceedings of
Informing Science & IT Education Conference. (2010).
[8] Cortez P, Silva A. Using data mining to predict Secondary school student performance. Journal of
information science Volume 2(6). (2013).
[9] Ahmed, A. B. E, Ibrahim S. E.. "Data Mining: A prediction for Student's Performance Using
Classification Method." World Journal of Computer Application and Technology Volume 2(2) (2014).
[10] Sembiring S, Zarlis, M, Hartama, D. Ramliana S, Elvi W. “Prediction of student academic performance
by an application of data mining techniques.” International Conference on Management and Artificial
Intelligence IPEDR Volume.6, (2011).
[11] Surjeet K, Yadav, Bharadwaj, B. Pal B.” Data Mining Applications: A comparative Study for Predicting
Student‟s performance.” International journal of innovative technology & creative engineering. Volume
1(12). (2012).
[12] Mladen D., Mirjana P. B., Vanja Š., “Improving University Operations with Data Mining: Predicting
Student Performance”, International Journal of Social, Behavioral, Educational, Economic and
Management Engineering Volume 8(4), 2014.
[13] Meltem, D. “Gender difference in academic performance in a large public university in Turkey”.
Economic Research center working papers in economics. 4(17). Pp. 22-23, (2004).
[14] Abubakar, R. B. and Oguguo, O. D. “Age and gender as predictors of academic achievements of College
mathematics and science students.” Proceedings of the International Conference of teaching, learning and
change. International Association of Teaching and learning. (2011).
[15] Nnamani, C. N, Dikko, H. G and Kinta, L. M. “Impact of students‟ financial strength on their academic
performance: Kaduna Polytechnic experience”. African Research Review 8(1), (2014).
[16] Ogunde A.O., Ajibade D.A. “A data Mining System for Predicting University Students F=Graduation Grade
Using ID3 Decision Tree approach”, Journal of Computer Science and Information Technology, Volume 2(1)
(2014).
[17] Ryan S.J.D. Baker, Kalina Yacef. “The State of Educational Data Mining in 2009: A Review and Future
Visions”, Journal of Educational Mining, Volume 1(2009).
[18] Undavia, J. N., Dolia, P. M.; Shah, N. P. “Prediction of Graduate Students for Master Degree based on
Their Past Performance using Decision Tree in Weka Environment”. International Journal of Computer
Applications; Volume 74 (21), (2013).

Authors’ Profiles

Kolo David Kolo holds a Bachelor of Technology (B.Tech) Degree in Mathematics/Computer

Science from Federal University of Technology (FUT) Minna, Nigeria, in 2000, Postgraduate
Diploma in Education (PGDE) from the Ahmadu Bello University (ABU) Zaria, Nigeria and
Master of Technology (M.Tech) degree in Computer Science from the Federal university of
Technology, Minna, Nigeria respectively in 2015. He works as a lecturer in the Department of
Computer Science, College of Education, Minna, Nigeria. His area of interest includes database
 A Decision Tree Approach for Predicting Students Academic Performance 19

management and mobile learning.

Solomon. A. Adepoju holds B.Tech (Maths/Computer Science) and M.Sc (Computer

Science) from the Federal University of Technology, Minna and University of Ibadan
respectively. His research interests include Human Computer Interaction, Web Mining and
ICT4D is a member of Computer Professional Association (CPN) as well other international
association.

Dr. J. K. Alhassan obtained B.Tech (Mathematics/ Computer Science) at the Federal

University of Technology, Minna. M.Sc (Computer Science) at University of Ibadan, Nigeria
in 2006, and PhD (Computer Science), at Federal University of Technology, Minna, Nigeria.
He carried out part of his PhD research at United Institute of Informatics Problems, National
Academy of Sciences of Belarus (UIIP NASB) Minsk, Republic of Belarus. He is currently
the Acting Head, at the Department of Cyber Security Science, Federal University of
Technology, Minna, Niger State, Nigeria. His research interest includes Artificial
Intelligence, Data Mining, Internet Technology, Database Management System, Software Architecture,
Machine Learning, Human Computer Interaction and Computer Security. He is a member of Computer
Professionals (Registration Council) of Nigeria (CPN).

How to cite this paper: Kolo David Kolo, Solomon A. Adepoju, John Kolo Alhassan,"A Decision Tree
Approach for Predicting Students Academic Performance", International Journal of Education and
Management Engineering(IJEME), Vol.5, No.5, pp.12-19, 2015.DOI: 10.5815/ijeme.2015.05.02