Clashes in Timetable Scheduling

Barney Clive Mathe

(Student number: 1355692)

Supervisors:
Dr Joke Bührmann

A 4th year Research Project submitted to the School of Mechanical, Industrial and Aeronautical
Engineering in the Faculty of Engineering and the Built Environment, at the University of the
Witwatersrand, in fulfilment of the requirements for the Bachelor’s degree in Engineering (Industrial).

09 September 2021

i
Abstract

In this study timetable scheduling in the School of Mechanical, Industrial, and Aeronautical
Engineering at the University of Witwatersrand was investigated. The purpose of the study was to
address the clashes that exist in the timetables by identifying these clashes and researching on ways
they can be reduced thus allowing students to register for the maximum number of courses possible.
This included examining methods used for scheduling in the school and in other institutions in order
to understand the challenges and constraints that come with scheduling. And an investigation on the
optimal number of courses that students can take and hours they can handle in a day without being
overloaded by work was also done. Students in their second, third and fourth year of study within the
school and lecturers who lecture in these respective years, were considered as participants. The study
was delimited to only the school’s academic schedules, students, and staff. Also, the research only
considered cases of online and contact learning. Students participated in a survey while lecturers were
interviewed. Other tools used included Simio to detect clashes in the timetables. Email and WhatsApp
messenger were used for the distribution of the survey and interview invitations. And for uncertainty
analysis, a sample size calculator was used. Excel was used for recording and processing of data. And
for further processing, a coding tree was developed for interview responses. ANOVA was used to
compare students’ scores for the academic schedules based on course load. And a t-test was
performed to compare students’ and lecturers’ ratings of the academic schedules and differences in
lecture duration preferences. Regression analysis was also performed to assess the correlation of
scores for the tests and class timetable. And it was found from Simio experimentation that all the
academic schedules for all streams had clashes and through the surveys, it was found that students
surveyed have also experienced clashes in their undergraduate years. From the surveys it was also
found that most students are willing to work extra time to reduce clashes. It was identified that free
afternoons and free lectures can be used to schedule courses that cause clashes. It was also discovered
that courses can be rearranged to avoid clashes which includes the use of the timeslots mentioned
above. So, it was concluded that clashes can be reduced in the timetables and some of the ways to do
this include the rearrangement of courses using excel solver, proactively changing timetables based on
failure rate, an asynchronous learning system which offers students courses which they have clashes
for and the use of a Kanban system that allows for easy scheduling. So, proceeding from this research
study the recommendations above will be explored to see their feasibility in reducing clashes.

ii
 INDIVIDUAL DECLARATION WITH JOINT TASK SUBMITTED FOR ASSESSMENT

I (name) _____Barney Mathe ___________________ (student number) __1355692_____________,

am registered for (course code) _____MECN4006___ (course name) _Research Project in the year

2021.

I herewith submit the following task, “Timetable Scheduling Project Report

_______________________________________________________________________”

in partial fulfilment of the requirements of the above course.

I hereby declare the following:

o I am aware that plagiarism (the use of someone else’s work without their permission and / or
without acknowledging the original source) is wrong;

o I confirm that the work submitted herewith for assessment in the above course is my own unaided
work except where I have explicitly indicated otherwise;

o This task has not been submitted before, either individually or jointly, for any course requirement,
examination or degree at this or any other tertiary educational institution;

o I have followed the required conventions in referencing the thoughts and ideas of others;

o I understand that the University of the Witwatersrand may take disciplinary action against us if it
can be shown that this task is not our own unaided work or that we have failed to acknowledge
the sources of the ideas or words in our writing in this task.

Signature______ ___________________Date__09 September 2021____

iii
Contents

Contents............................................................................................................................................iv
List of Figures..................................................................................................................................vi
List of Tables.................................................................................................................................viii
Nomenclature..................................................................................................................................ix
Chapter 1 ........................................................................................................................................1
1. Introduction..................................................................................................................................1
1.1 Research Background.............................................................................................................1
1.2 Problem Statement and Motivation........................................................................................2
1.3 Research Question..................................................................................................................3
1.4 Research Objectives...............................................................................................................3
1.5 Research Method Summary...................................................................................................3
1.6 Delimitations..........................................................................................................................4
1.7 Outline of Chapters................................................................................................................4
Chapter 2........................................................................................................................................5
2. Literature Review........................................................................................................................5
2.1 Constraints.............................................................................................................................5
2.2 Students Interests...................................................................................................................6
2.3 Timetable design....................................................................................................................6
2.4 Number of courses.................................................................................................................8
2.5 Time in the timetable design..................................................................................................8
2.6 Class size and space...............................................................................................................9
2.7 Assessments timetable..........................................................................................................10
2.7.1 Time for assessments ...................................................................................................10
2.7.2 Number and arrangement of assessments.....................................................................11
2.8 Timetable scheduling during Online Learning.....................................................................11
Chapter 3.......................................................................................................................................13
3. Research Method .......................................................................................................................13
3.1 Research Design...................................................................................................................13
3.2 Instrumentation.....................................................................................................................13
3.3 Sampling ..............................................................................................................................18
3.4 Data collection procedures...................................................................................................18

iv
 3.5 Data Processing and analysis Procedures...........................................................................25
3.6 Validity and reliability........................................................................................................26
3.7 Ethical Considerations........................................................................................................26
3.8 Risks Mitigation and Contingency Plans ...........................................................................27
Chapter 4 ....................................................................................................................................28
4. Data Analysis and Results........................................................................................................28
4.1 Data.....................................................................................................................................28
4.2 Data Processing ..................................................................................................................32
4.3 Results.................................................................................................................................40
Chapter 5 ....................................................................................................................................59
5. Discussion................................................................................................................................59
Chapter 6 ....................................................................................................................................64
6. Conclusions and recommendations..........................................................................................64
7. References................................................................................................................................67
8. Appendices...............................................................................................................................70
A. Survey..................................................................................................................................70
B. Interview questions..............................................................................................................78
C. Interview Protocol................................................................................................................79
D. Coding Tree.........................................................................................................................80
E. Simio Timetable models......................................................................................................84
F. Clashes in timetables….......................................................................................................88
G. F tables ...............................................................................................................................89
H. t-tables.................................................................................................................................91
I. Ethics Certificate..................................................................................................................92
J. Registrar Approval Letter....................................................................................................93
K. Margin of error calculations...............................................................................................94

v
List of Figures

Figure 1: Survey Structure .................................................................................................................... 14

Figure 2: Interview Structure ................................................................................................................ 14
Figure 3: Simio Object Resources ........................................................................................................ 16
Figure 4: Source Object ........................................................................................................................ 16
Figure 5: Server Object ......................................................................................................................... 16
Figure 6: Simio Path ............................................................................................................................. 16
Figure 7: Sink Object ............................................................................................................................ 17
Figure 8: Simio model comparison of courses for one timeslot ........................................................... 19
Figure 9: Simio course comparison model for a semester .................................................................... 20
Figure 10: Simio Course Comparison model for a single Test day ...................................................... 22
Figure 11: Experiment for the Industrial Engineering stream .............................................................. 32
Figure 12: Experiment for the Mechanical Engineering stream ........................................................... 33
Figure 13: Experiment for Aeronautical Engineering stream ............................................................... 33
Figure 14: Experiment for the tests timetable ....................................................................................... 33
Figure 15: Coding tree Structure ........................................................................................................... 34
Figure 16: Anova Excel ........................................................................................................................ 35
Figure 17: T-test on excel ..................................................................................................................... 38
Figure 18: Correlation Coefficient ........................................................................................................ 39
Figure 19: Clashes in each stream of engineering ................................................................................ 40
Figure 20: Clashes in the tests timetable............................................................................................... 40
Figure 21: Students with clashes in undergraduate years ..................................................................... 42
Figure 22: Cross Registered students with clashes ............................................................................... 41
Figure 23: Students with clashes for each stream ................................................................................. 41
Figure 24: Students willingness to do extra work to reduce clashes..................................................... 42
Figure 25: Timeslots for extra work to be done .................................................................................... 43
Figure 26: Effect of course load on students’ performance .................................................................. 44
Figure 27: Students rating of the current class timetable ...................................................................... 44
Figure 28: Students preferred number of courses per day .................................................................... 46
Figure 29: Students preferred number of hours per day ....................................................................... 46
Figure 30: Constraints that affect students’ attendance of classes ........................................................ 47
Figure 31: Students preferred lecture duration...................................................................................... 47
Figure 32: Students rating of the tests timetable ................................................................................... 48
Figure 33: Regression Analysis ............................................................................................................ 49
Figure 34: Students preferred number of tests per day ......................................................................... 50
Figure 35: Students preferred timeslot to write tests and exams .......................................................... 50

vi
Figure 36: Constraints affecting students tests/exams writing conditions ............................................ 51
Figure 37: Number of Courses for each lecturer ................................................................................... 51
Figure 38: Amount of work performed by lecturers ............................................................................. 52
Figure 39: Other Duties performed by lecturers ................................................................................... 52
Figure 40: Lecturer's thoughts on the scheduling process .................................................................... 53
Figure 41: Perceived constraints on tests and class timetables by lecturers ......................................... 54
Figure 42: Constraints on the lecturing schedule .................................................................................. 54
Figure 43: Preferred Class Size by lecturers ......................................................................................... 55
Figure 44: Number of hours spent on academic activities by lecturers ................................................ 55
Figure 45: Lecturers Preferred timeslot ................................................................................................ 56
Figure 46: Lecturers Preferred lecture duration .................................................................................... 56
Figure 47: Students willingness to work on an integrated system for clashes ...................................... 57
Figure 48: Students preferred learning method to reduce clashes ........................................................ 58
Figure 49: Students response to timetable that changes with failure rate ............................................. 58
Figure 50: Mechanical Engineering Model .......................................................................................... 84
Figure 51: Industrial Engineering Model .............................................................................................. 85
Figure 52: Aeronautical Engineering Model......................................................................................... 86
Figure 53: Tests timetable Model ......................................................................................................... 87
Figure 54: Margin of error calculation for students responses ............................................................. 93
Figure 55: Margin of error calculation for lecturers responses ............................................................. 94

vii
List of tables

Table 1: Real-life representation of Simio objects ................................................................................ 18

Table 2: Second to Fourth year courses ................................................................................................ 28
Table 3: Responses Per stream.............................................................................................................. 30
Table 4: Responses per year of study.................................................................................................... 30
Table 5: Students responses per course load ......................................................................................... 30
Table 6: Responses per question ........................................................................................................... 30
Table 7: Students rating of class timetable matched with course load ................................................. 45
Table 8: Anova analysis on students’ ratings of the class timetable ..................................................... 45
Table 9: Students rating of the tests timetable matched with course load ............................................ 48
Table 10: Anova analysis on students’ ratings of the tests timetable.................................................... 49
Table 11: Regression Coefficient.......................................................................................................... 50
Table 12: T-test for ratings on Timetables ............................................................................................ 53
Table 13: T-test on lecture duration ...................................................................................................... 57
Table 14: Clashes in the timetables....................................................................................................... 88
Table 15: F tables .................................................................................................................................. 89
Table 16: T-tables ................................................................................................................................. 91

viii
Nomenclature

Symbol Meaning Units

𝑥̅/𝜇 Mean -
df Degrees of Freedom -
S Standard Deviation -
𝑛 Number of Samples/ responses -
t95 t value at 95% confidence level -
ni Number of responses for each -
level in ANOVA

N Total Number of responses for -

all levels in ANOVA.
xi Response value -
a Total Number of levels in -
ANOVA
yij Response value in a particular -
level in ANOVA

Ai Total summed value for all the -

responses in a particular level
in ANOVA

T Total summed value for all the -

response values in all levels in
ANOVA

SST Total Sum of squares -

SSF Sum of squares due to factor -
SSE Sum of squares due to error -
MSF The mean square of the factor -
MSE The mean square of the error -
VF Degree of freedom for the -
factor

VE Degree of freedom for the error -

F Factor level mean -

ix
F crit Critical factor -
t T statistic calculated from the t- -
test

H0 Null hypothesis -
H1 Alternative hypothesis -

x
Chapter 1

Introduction

1.1 Research Background

Over the years, timetables have been an integral part of the learning environment. Timetabling
documents have played a crucial role in the facilitation and organization of activities in different
workspaces, without these documents order would not be possible mostly in systems that involve
human beings as participants. Their significance can be seen with their use in workplaces, schools,
universities, transportation, etc. By definition, “timetables are schedules listing the events and the
times at which these events are going to take place” [1, p.1]. As noticed in the definition there is an
emphasis on the order and duration of activities to be performed in a given workspace, thus
facilitating how work should be done and when it should be done. So, it is clear to see that timetables
have the ability to control activities in a workspace, and all entities within the workspace whether
human or machine are subject to following these schedules. Therefore, it is imperative to fully
understand these documents, the factors which affect them, and how they are scheduled. So, this
research study will focus on timetable scheduling and its crucial impact in facilitating learning and
ensuring that there is a continuous flow of activities in the learning environment.

The context in which this research study will be conducted will be through the school of Mechanical,
Industrial, and Aeronautical Engineering situated and functioning at the University of Witwatersrand.
As in the name, this school comprises of three streams of engineering which includes the Mechanical
Engineering stream, Industrial Engineering stream, and the Aeronautical Engineering stream. Since
the school functions within the premises of the University, it makes use of university properties such
as lecture halls, laboratories, and exam halls just to name a few. These resources are used for lecturing
or instructional learning and other academic activities, they are used in competition with thirty-two
(32) other schools at the University of the Witwatersrand [2].

The school generally makes use of three scheduling documents to organize most of the academic
activities for students. These documents include the class timetables and tests or assessments
timetable for all years of study in the school and the faculty rule book. These documents are also of
interest in this study. The class timetables for all years of study include an arrangement of scheduled
lectures with the time durations which each lecture is supposed to take, which shows the starting and
ending time of each lecture. Along with these time slots, it also indicates venues in which each lecture
is supposed to take place. All these are arranged differently on different days of the week and within
the first two years, all streams share the same timetable while for the last two years of study the
timetables are specialized to suit each stream. The assessments timetable also follows a similar
arrangement in that it also indicates the tests and assignments to be completed for each course and the

1
dates on which these have to be done. It shows all the tests and assignments for all years of study and
streams. It is also very significant to note that even though these documents still have the same
primary function, there has been a change in how they are used in this period of the covid19
pandemic. This change has greatly affected the way in which people follow these documents and their
impact on the learning environment. So, the role of these documents in the school and also in the
general learning environment will be evaluated in this research study, even the way in which they are
scheduled.

The last document of interest in this study is the faculty rule book which is a document that outlines
the rules pertaining to admission, registration, attendance, curriculum, and progression. In this
research only the progression rules and guidelines will be used, these govern the progression of
students from one year of study to the next year. The combination of these documents is essential in
understanding how scheduling has been executed in the school of Mechanical, Industrial, and
Aeronautical Engineering to coordinate work or academic activities.

Individuals of interest who are stakeholders in the school that this report focuses on are those that are
affected by or those that affect the scheduled documents. This includes members of staff such as
lecturers who lecture courses within the three streams. These lecturers are also responsible for
students’ personal academic matters which can be matters related to course-specific problems and
class-related problems. The class-related problems encompass aspects such as lectures, tests,
assignments, tutorials, venues, and the scheduling of documents to facilitate these aspects. Students
are another group which this study focuses on, this is because students are subject to following the
scheduled documents. They are also expected to abide by the rules and guidelines expressed in the
faculty rule book. So, all the stakeholders indicated above are impacted by the scheduling process in
some way.

1.2 Problem Statement and Motivation

As expressed above students are affected by the three scheduling documents considered in this study.
So due to students failing certain courses in certain years of study within the school of Mechanical,
Industrial, and Aeronautical Engineering, this has caused clashes in cross-registered students’
timetables. These cross-registered students are those according to the rule book on progression have
been limited on the number of courses to be admitted in and are enrolled in courses of two different
years of study. As these students are able to register for courses of different years of study, a timetable
of one year of study might clash with timetables of other years of study in courses students are
allowed to register for. And due to these clashes’ students omit these courses which cause clashes
even when they qualify for them according to the requirements in the rule book such as prerequisites
and corequisites. And as the rule book stresses the importance of constraints such as prerequisites
which have valuable content that is needed for higher years of study courses. This then creates a

2
problem in that students who omit certain courses due to clashes tend to forget the content acquired
from these courses which were passed but could not be carried on with.

1.3 Research Question

Therefore, the purpose of this study is to unravel the following question:

How can clashes in the timetables be reduced while ensuring that students who meet requirements are
registered to the maximum number of courses as possible and that there is maximum use of the time
available for teaching and learning?

1.4 Research Objectives

The main objective of this research is:

• To identify ways to reduce clashes in the school’s academic schedules by studying the
existing schedules such as tests, assignments, and class timetables, and how these documents
are scheduled.

This will then require the research study to:

• Examine the methods used in scheduling timetables within the school and in other institutions
of learning to fully understand the challenges or constraints which come with the scheduling
process.
• Identify the number of clashes and where they appear in the timetables.
• Investigate the optimal number of courses that students can take while ensuring that they are
not overloaded by work.
• Investigate the optimal number of hours suitable for students to do their scheduled activities
and for lecturers to work on their courses.

1.5 Research Method Summary

The method used for the research study in trying to unravel timetable scheduling will begin with
analyzing previously performed studies on the subject matter. This will include consideration of
literature relevant to the research study and it will allow for further understanding of the subject
matter. Then in order to conduct research on timetable scheduling in relation to the School of
Mechanical, Industrial, and Aeronautical Engineering, the type of data needed for this study will be
identified including the methods to be used for data collection and experiments. This includes the
method used to identify and show clashes in the timetables.

After the methods for data collection and for conducting experiments have been identified, tools to be
used for the experiments or data collection will be selected, depending on their ability to measure the
research variables. Participants in the research study should also be selected based on the system of

3
interest and the needs of the research study. Then experiments and data collection will be performed
with the use of the methods considered and tools selected. After the data collection and
experimentation, the outcomes will be processed and analyzed with the tools selected. Then the
results or outcomes of the analysis will be discussed in relation to the research topic, problem,
objectives, and in relation to the critical research question. This discussion also includes making
comparisons of the outcomes obtained with literature on similar research already existing. Based on
the discussion an informed conclusion should then be made on the research problem and
recommendations on how to address the problem should be stated.

1.6. Delimitations

Since the research study is conducted at the School of Mechanical, Industrial, and Aeronautical
Engineering at the University of Witwatersrand, the research study will only concentrate on this
system boundary. Meaning that participants considered for this study will only come from the school.
The research will also focus on factors that relate to the school’s scheduling system and also the
experimental data will mainly be reflective of the school’s system. This includes the use of only the
school’s scheduling documents to do investigations on the research topic and problems. Though
research will be done on previously performed studies on the research topic in other institutions, this
will also be compared to the school’s system to see if the information considered is applicable. The
study will focus on both online and contact learning in trying to understand timetable scheduling,
other forms of learning will not be considered for this study.

1.7. Outline of Chapters

Chapter 1: Introduction- this includes information on the background of the research, the problem
statement or motivation, the research question and objectives

Chapter 2: Literature Review- includes consideration of previously performed studies on the subject
matter, information on similar studies performed is recorded based on its applicability to the study.

Chapter 3: Research Method- this comprises of the research design, the instrumentation used,
sampling or the sampling plan, data collection, and analysis procedures, information on the Validity
and Reliability of the research, and Ethical considerations.

Chapter 4: Analysis and Results- this includes the analysis of the processed results obtained through
the application of the research method, it also includes the display of the analyzed results.

Chapter 5: Discussion-This includes a thorough discussion on the research results and their relation to
the research problem, objectives, and question, and also in relation to the literature review.

Chapter 6: Conclusions and Recommendations- This includes conclusions on the research problem
and potential recommendations on how the problem can be addressed.

4
Chapter 2

Literature Review

In addition to the definition of timetables considered in the introduction, Oude Vrielink et al. [3]
further define timetabling as the “allocation of given resources to specific objects being placed in
space-time in such a way as to satisfy as much as possible a set of desirable objectives and subjected
to constraints’’[3, p. 148]. The two definitions are very significant in that they emphasize on the role
of tangible and intangible resources in the scheduling of timetables. The availability of these resources
is associated with meeting constraints in the scheduling process, so the role of these constraints is an
important aspect to consider and one which is investigated in this research. Further research on the
algorithms used to construct timetables will be performed in the design project succeeding this
research study.

2.1 Constraints

According to Oude Vrielink et al. [3], constraints that impact timetable scheduling include classrooms
and housing, the number of staff available to lecture, diminishing resources which are resources that
become scarce as they are used up over time. And other considerations such as the timetable practices
used for timetabling [3]. Similar to [3], research study [4] also indicates that there are three variables
that need to be considered, these variables include lecturer availability which is regarded as the
product of the lecturers and the time they are available. Thus, showing whether lecturers are available
for a specific lecture or not in a given period of time [4]. The second variable which is similar is the
class availability, which in this reference is defined as the product of the classes and the time in which
they are available within a given space [4]. So, this then shows whether certain classes are available
for specific lectures or not. Then the third variable which is considered in the paper and which is
different from the ones already mentioned is class requirements, which define the total number of
hours each lecturer is to meet each class [4].

This class requirement is also addressed by [5] as the time slot of each lecture. The paper also concurs
with the other references in that it considers lecturer availability and venue, and also states that each
lecture period must at least have one lecturer available [5]. The paper further states that the challenge
is generating an optimal timetable solution that takes into account all the various constraints [5]. Some
of the problems discussed in the paper that may arise in creating an optimal timetable include both
teacher and course clashes [5]. This idea of an optimal timetable can be associated with the timetable
practices as mentioned in [3], in that the best practices in creating an optimal timetable need to take
into account course and teacher clashes.

Other constraints that are considered to have an influence on timetable scheduling include the length
of the school day, week, and year, and also the number of subjects in the curriculum [6]. The amount

5
of academic support is also considered by [3], this reference also states that all these constraints or
impeding factors are ultimately related to the budget which is available to allow for these services [3].

So as mentioned in [5] that the biggest challenge is to create an optimal timetable to cater to all these
constraints. Due to these differing constraints, diverse problems, and unique requirements of different
institutions this has made it difficult to generate a generally applicable model for all timetabling issues
in all institutions [3]. So, timetabling has remained unique for different contexts, but it can be noticed
that there are general themes or trends that are discussed by each reference above which includes
time, space, number of courses/ workload, and availability of lecturers. Different models developed
need to take into account these factors and their relation to performance and students’ interests. So
further literature on students’ interests and performance is essential, which includes how the common
factors as mentioned above influence these two aspects.

2.2 Students Interests

According to Larabi-Marie-Sainte et al. [7], students’ academic performance is the point of interest
for both students and the academic institution. This performance is said to be subject to factors such
as student absences, number of courses per semester, the average number of lectures per day, the
average number of free timeslots per day, and the length of each lecture [7], which to some extent are
similar to the general factors mentioned above. Performance is also associated with the assessments
given to students and how these assessments are facilitated. So, in addition to the above factors,
assessment-related factors such as the number of tests written in a day, time given for each
assessment, and arrangement of the assessments are crucial aspects of timetabling that affect the
performance of students [8]. In [9] it is also indicated that students seek to enrol in the maximum
number of courses allowable and that enrolment should be subject to students’ preference or choice on
which courses they would like to pursue or not due to certain limitations [9].

2.3 Timetable design

In consideration of the above, the design of timetables should ensure that students’ interests are
satisfied. This means satisfying performance objectives that are expected from timetables. The first of
which is flexibility in the timetabling system [3]. This includes increased accessibility to resources
and the ability to participate in multiple courses within the curriculum [3]. Thus, an increase in the
number and types of courses students can take in specific years of study which is one of the students’
interests mentioned above. This flexibility also means that the timetable is adjustable to suit students
and staff interests such as suitable times for learning, proper arrangement of courses, a manageable
workload, and working times for lecturers.

In [3] it is also stated that the design of the timetable needs to maintain the quality of teaching and
learning which refers to quality in the activities that promote learning. This quality includes

6
institutional behaviours and activities that advance the institution’s educational objectives which as
expressed above includes increased performance by students [3]. This quality is subject to the amount
of academic support that can be offered to students even outside the classroom setting to encourage
continuous and proper learning which also depends on the availability of lecturers which is one of the
constraints mentioned above. This support further expands to consider the support given to lecturers
to improve the quality of teaching which reflects in students’ performance. It is also influenced by the
availability of services such as course development units, advising, and supervision of student
research [10].

Another performance objective that needs to be met by the timetable design is efficiency. The
timetables need to maintain a certain degree of efficiency while ensuring that educational quality
levels are also maintained [3]. In [9] the freedom of choice available to students is used to develop a
measure of timetable efficiency. In this paper, this efficiency balances the desirability of having both
popular and rarer course options available against the need to deny some students their chosen courses
[9]. The reference indicates that a practical timetable is one that accommodates some allowable
courses and excludes others [9]. So, efficiency in this regard is dependent on the students’ choice of
courses which need to be incorporated in their timetables. The choice is first made based on students’
interests then followed by consideration of the facility’s or organization’s constraints [9]. The paper
defines the timetabling process to consider distinct courses allowable under the rules of the institution,
then students’ preferences, and then the faculty [9]. So, in [9], an efficient timetable is one that caters
to a subset of all allowable programs or courses in accordance with the rules of the institution or
faculty. The more courses are omitted from the timetable the less efficient the timetable becomes [9].

Proper student engagement is another performance objective emphasized by [11] and is considered
crucial in the timetable design. Student engagement is considered as key to unlocking students’
potential and thus improving students’ performance [11]. A balanced timetable in this case is one that
enables student engagement in the teaching and learning process. It is a customized timetable that
considers students’ abilities, interests, and learning styles [11]. It considers the alignment of what is
taught and how it is taught to the abilities, interests, and learning styles of students [11]. So, this paper
concurs with the flexibility as considered in [3] and asserts that timetabling may vary based on
students’ needs. In relation to the academic support mentioned above, this reference also indicates that
necessary support needs to be provided for students on factors that may hinder their learning
experience [11]. This may include academic-related matters such as not coping with schoolwork, may
also include family and community factors [11]. In the text, this is expressed in what is termed as
‘down time’ which is the time provided to students experiencing negative emotions [11]. This is an
important aspect even in higher education as students may be stressed or overwhelmed by course
material, so consideration of this aspect is important for healthy teaching and learning. Another
classified time mentioned in the reference is called social time, this is a period of time where students

7
participate in interactions with other students [11]. This promotes collective learning and teamwork,
the reference also states that it improves social skills and social connectedness which are essential in
the real workplace [11]. So, in [11] emphasis is on the importance of a timetable that does not only
promote lecturer and student engagement but also student to student engagement which allows
students to learn from one another. This objective of increasing student engagement and support is
also echoed in [10].

2.4 Number of courses

The number of courses done by students is one of the factors that play a crucial role in timetable
scheduling, and as mentioned in the literature above it can also affect the performance of students and
the kinds of timetable practices employed by institutions. According to a research study [10], the time
to degree can be reduced if students take more credits each term, it indicates that this may increase the
academic performance of students. In a study conducted in the paper, it was discovered that there is
no evidence that proves that high course loads have a negative impact on students’ grades even for
students with low academic performance [10]. The reference indicates that students tend to substitute
some of their leisure time away from studies with academic activities when the course load increases
[10]. Some of the other advantages stated in the paper are that it increases students’ persistence and it
is cost-effective because it reduces time to degree [10]. According to the paper, this increase in course
load helps align students with their education, which results in them exerting more effort into their
studies [10].

The reference states that taking higher credit courses than lower credit courses badly affect students’
academic performance but taking additional lower credit courses per semester leads to an increase in
course grades/ performance [10]. This increase includes a 0.011 standard deviation increase per
semester course added [10]. Though the paper acknowledges that increasing the number of courses
may affect performance because students may spend less time on each course, it also indicates that
this is subject to the time allowable for students to work on these courses in the timetables and other
scheduling documents [10]. So as stated, increasing responsibilities for students leads to increased
performance. And one way to increase performance based on increasing course load is giving students
more options in course scheduling [10]. This then concurs with the increase in flexibility of academic
schedules as expressed in one of the performance objectives above for timetable design.

2.5 Time in the timetable design

In addition to the number of courses as stated above, in [10] it is also indicated that time investment
also affects performance. The paper asserts that there is a strong association between students’ study
time and their performance [10]. According to the reference, students’ time and effort respectively
improve student’s participation and performance [10]. So, the number of hours students work in a day
might affect students’ performance, hence an optimal timetable needs to take this into consideration.

8
In [12] it is indicated that an ideal class length is supposed to be shorter than the conventional class
which lasts for approximately fifty to Ninety minutes. According to the paper, lecture durations need
to be kept as short as possible, it indicates that as much as 30 minutes allows students to maximize on
their learning time and experience [12]. The article argues that the average attention span for students
is approximately 10 minutes, this prediction applies to the normal classroom setting [12]. Classes or
lectures which end beyond this, result in students missing a great margin of instructional material or
course content [12]. The article suggests that breaking up the material into smaller periods of time
improves students’ attention.

Disadvantages of longer classes discussed in the reference include that with more time lecturers or
teachers tend to spend more time on non-value adding activities or classroom maintenance such as
stretching out time for filling attendance registers and taking up papers [12]. Other disadvantages
include that the course material covered may even go beyond what is supposed to be covered [12],
meaning more information is given to students even when it is not needed for the course. Another
disadvantage included in the paper is that lecturers or teachers tend to allocate more time to activities
such as question time even when students have no questions which results in them socializing in a
time that could have been used to perform useful work [12]. The advantages then discussed in the
paper for shorter classes include that constantly shifting classes break up the day and prevent boredom
amongst students [12]. Learning time is maximized with shorter classes, also lectures are changed on
a regular basis allowing for new information to be constantly taught [12]. The time for non-value-
adding activities is shortened and lecturers are able to focus on the core activities.[12].

Reference [13] also concurs with the argument on shorter classes and indicates that optimal focus by
students is around 10- 18 minutes no matter how good the lecturer is. It also speaks against the value
of long lectures and states that after 10- 18 minutes students tend to lose focus [13]. According to the
study, students are able to recall facts contained at the beginning of the lecture, and their focus
declines as the lecture progresses [13].

2.6 Class size and space

The other factors which need to be considered in terms of timetable scheduling are the class size and
space which is mentioned as one of the common factors in the constraints section. Space in terms of
timetable scheduling refers to the venues available for use by institutions for lectures. Timetabling
depends on the availability of these spaces/venues and so it is crucial to investigate the effect that this
aspect might have on the scheduling process. In a study conducted by Laiqa et al. [14], it was
discovered that quality space positively affects students’ academic achievement. In the study groups
of students were exposed to two different space conditions. The control group of students was
exposed to an ordinary classroom environment while the experimental group was provided with
quality space [14]. A test was conducted before the experiment under normal conditions for all

9
students, then another during the experimentation [14]. It was found that there was a significant
difference between the two groups of students [14]. The experimental group had a significant
development in performance compared to the control group which had no significant development in
performance [14]. This then indicates that quality space is essential in improved students’
performance. In a study conducted in [15], it was also discovered that reduction in class size, which is
the number of students in a class, increased student achievement. This class size reduction is
associated with space and accessibility in that with a decrease in class size the greater the accessibility
to lecturers that each student has and in a physical environment the increase in space available. This
decrease in class size also means more students are denied access to courses in the curriculum.

2.7 Assessments timetable

After consideration of class timetabling and the factors which affect this process, it is then essential to
understand the applicability of this in assessments timetable which may include tests, examination,
and assignment timetables.

2.7.1 Time for assessments

In a study conducted by Onwuegbuzie et al. [16] where students were exposed to two time-varying
examinations which included a timed examination and one which was not timed. It was discovered
that both low and highly anxious students performed better on the course examinations under the
untimed condition than under timed conditions [16]. It was also discovered that highly anxious
students performed better than low anxious students during the untimed examinations [16]. This
shows that during assessments student’s performance is affected by time constraints given and that
ease in the time constraints can improve the performance of students thus showing their true learning
potential. Another aspect of this study is that the difference between high and low anxious students in
assessments is caused by the difference in motivational disposition and attentional focus [16]. The
paper indicates that anxiety badly affects students’ test performance, self-efficacy, and attitude
towards the course [16]. In relation to the study conducted in this paper, this shows that time also
plays a role in the anxiety of students during assessments. Students tend to be more anxious under
time constraints which results in them performing badly on the assessments given resulting in a
negative attitude towards the course.

According to Khan [13], for assessments or tests, the length of time in which these should run is
dependent on factors such as the number of questions and the difficulty. Reference [17] expands on
this further and indicates that to determine the length of an assessment lecturers need to take the test
they have set and record the time to complete the assessment themselves. Then they need to multiply
this time by three which will give them the time students are expected to take the assessment [17].

10
In [8] it is also indicated that performance is positively associated with the time between
examinations. Time for preparation and recuperation is also considered in the paper and it is asserted
that the effect on performance that the time length between exams has, may be as a result of the
impact this aspect has on preparations [8]. The effect on performance of the time between the first
exam and the exam that is about to be taken can indicate students’ fatigue from exam completion [8].
The paper states that for examinations taken later in the assessment schedule, students tend to perform
higher in them than examinations taken earlier [8]. The study also indicates that exam performance
decreases with an increase in the number of exam days, so additional days in the examination period
increase exam fatigue thus poor performance in some exams [8]. The number of days between
examinations is also associated with a decrease in exam performance.

2.7.2 Number and arrangement of assessments

In [8] the other factors that influence performance are indicated which include the order in which
examinations are taken which has been slightly discussed above and the number of examinations to be
completed in a specific period of time. The paper asserts that productivity increases with exams,
meaning that the learning effect is positively associated with taking an additional examination [8].
The paper indicates that for an optimal exam schedule the higher the number of exams taken, the
higher the potential benefit of optimizing on exam scheduling, thus improving the overall
performance [8].

In reference [18], it is indicated that for assessments written later in the day, students tend to have
lower performance in them even though it gives time for more preparation. It is revealed that for
every hour that students take from 8 AM in the morning there is an average decline in the test scores
thus a decline in performance [18]. Students are said to perform better in the morning as they are not
tired from attending class or from doing school-related tasks [18]. The ability to focus and make
decisions and react is affected by cognitive fatigue [18]. This cognitive fatigue is a result of expanding
energy on mentally demanding tasks [18]. So, with examinations written later in the day, students
might be cognitively fatigued to participate in them, due to activities that might have been done earlier
in the day [18]. Another discovery in relation to the arrangement of assessments was that taking
breaks also improves the performance of students in these assessments [18].

2.8. Timetable scheduling during Online Learning

As stated in the introduction it is very important to take note of the transition to remote learning that
has been adopted by most institutions in the wake of the covid19 pandemic and its influence on
schedules. It is clearly notable that with online learning, scheduling has become more flexible as there
are fewer constraints to be considered. Online learning comprises of two learning approaches which
include synchronous learning where there is real-time interaction between students and lecturers. In
this approach students and lecturers are expected to be on the same platform and at the same time for

11
teaching and learning [19], while for the other approach which is asynchronous learning this is not the
case [19]. In the asynchronous approach, learning can occur in different places and times [19]. Both
methods eliminate the need for a designated space where a class of students would need to meet, thus
making it easier to schedule lectures. This then can cater to larger groups of students and lecturers as
space is no longer a constraint.

Though space is eliminated as a constraint, for synchronous learning time still exists as a constraint
that majorly affects timetable scheduling. Asynchronous learning eliminates even the aspect of time
meaning course material can be accessed at any time anywhere. Virtual classrooms in this form of
learning are accessible 24 hours a day, seven days a week [20]. Another advantage is that learning is
even more time efficient as individuals are able to perform more activities and are able to plan their
days more easily [20]. The elimination of the constraints as mentioned above make it easier to
distribute course material and schedule academic activities such as lectures, tests or assessments, and
consultations at any time suitable for the parties involved [21]. Thus, increasing the ability to expose
as many students as possible to an increased number of courses, though it still has constraints such as
lecturers’ availability and the amount of workload that can be given to students.

12
Chapter 3

Research Method

This study explored different factors considered in the literature review to see if there is a correlation
between previously performed experiments or studies and this research study. Other factors specific to
the school of Mechanical, Industrial, and Aeronautical Engineering were also evaluated.

3.1 Research Design

In this research both qualitative and quantitative research methods were used which were structured to
provide responses to the critical research question and objectives stated above. For both qualitative
and quantitative research, surveys and semi-structured interviews were utilized. These tools were used
for primary data collection and were administered through a common medium that is accessible to all
the targeted participants in the form of an invitation. Once the participant accepted the invitation they
were interviewed and for surveys, participants had to complete an online form with questions related
to the research topic. Simio was also used for quantitative research. Further information on these tools
is given in the subsequent subsections.

3.2 Instrumentation

The instrumentation or industrial engineering tools that were used in this study includes:

A Survey: which is a primary data collection tool that was used to collect data from students in the
form of responses to a fixed set of questions in relation to the research topic, research problem, critical
research question, and objectives. This made use of google forms which is another tool that was used
for the design of the survey which includes setting the research questions. The form also allows for
the researcher to download responses from participants and indicates the percentage number of
respondents within a given period of time. The survey was designed to lead the participants in
addressing the research questions and objectives. It starts with students’ demographic questions which
are designed for demographic segmentation of the sample. Then progresses to questions that need
more thinking and recalling which includes questions in regards to their experience of previous and
current timetables, their preferences with regards to the academic schedules, their experiences of the
tests timetable, and on the factors that affect their participation in academic activities. Their thoughts
on a possible future model were also explored. The survey had 27 questions which were distributed in
the sections mentioned above. Some of the questions used a 5-point Likert scale to gauge students’
experience of a particular factor. On this scale, 1 and 5 represented the two extremes of the gauge,
where 1 represented a strong negative reaction and 5 represented a strong positive reaction. A
schematic summary of the survey structure is shown below with sections included in the survey and
the questions they comprise of.

13
 Future schedules
Question Number:
24,25,26,27 Demographics
Previous and Current timetable
Question Number: 1,2,3
Question Number:
4,5,6,7,8,9,10

Survey
p Students Preferences
Tests timetable
Question
Question Number: 18,1920,21 Number:11,12,13,14,15,16,17
Factors affecting student participation
in academic activities
Question Number: 22,23
FigureFigure
1: Survey
SEQStructure
Figure \*
ARABIC
Refer to Appendix A for more details1:on
Survey
the survey developed.
Structure
Semi-structured interviews: this is also a primary data collection tool that was used to collect data
from lecturers in the form of responses to interview questions to try and understand their views on the
research topic. These interviews were conducted through Microsoft teams which is another tool that
acts as an online meeting platform, and it included a private session between the researcher and the
interviewee which in this case are lecturers. The interviews also made use of an interview protocol
that outlines the ground rules and procedures to be followed during the interviews. The questions
were designed to allow for responses such as lecturers’ demographics, preference, their view of the
current scheduling system, and suggestions on a possible model suitable for them. The interview
questions were also separated to include questions for the member responsible for the scheduling of
yearly academic schedules. They comprised of 19 questions in total that were prepared to be asked,
where 7 of which are for the person responsible for the scheduling of yearly academic timetables. A
schematic summary of the general makeup of the interview questions is as follows.

Future schedules
Demographics
Previous and Current timetable
Question Number: 11
Question Number: 1,2
Question Number: 3,4,5,6,7,8

Interview
Lecturers Preference
Question Number: 9,10
For Person Responsible for timetable scheduling at the
school
Question Number: 12,13,14,15,16,17,18,19

Figure 2: Interview Structure

14
Refer to appendix B and C respectively for more details on the interview questions and protocol.

Email: This is the medium that was used to distribute the survey and interview invitations to all
participants. In this research, this included the use of the Wits University email which is capable of
reaching a larger student participant group. It allows for attachments and links to be embedded in the
emails which were used to include documents such as consent forms in each invitation and also for
the survey invitations a link to the online google form was embedded in the emails. So, this enabled
participants to simply click on the link which then led them to the google form. The email invitations
also included information that informed participants of the research study, the objectives, and what
their participation will entail. Information on confidentiality was also given with directives on how to
participate.

WhatsApp Messenger: This was the second medium used to distribute the survey so that it can be
seen by all students. This tool allows communication between two people through direct messaging,
video calls, and normal calls. It also allows for groups to be formed, which has been done by students
for different years of study. Each class in all years of study has its own group where class matters can
be discussed and important information can be shared such as documents, links, videos, etc.

Excel spreadsheet: This was used to record all students’ responses to the research questions for the
survey and the interviews. This was also used for processing the responses through features of the
software which includes graphing, comparisons, data analysis, and other features for statistical
measures.

Sample Size Calculator: This is a tool that was used to obtain the sample size that is statistically
significant relative to the population or for a given sample size and population the confidence level
and margin of error can be obtained.

Coding Tree: This was used to sort interview questions, it comprises of attribute codes and
substantive codes for themes related to the research study. Attribute codes in this case were used to
sort questions related to participants’ demographics or questions that define the participant, and
substantive codes were used to sort questions more directed to the research study and topic.

Simio: This tool was used to identify clashes in the School of Mechanical, Industrial, and
Aeronautical Engineering class and tests timetables. It was used to create simulation models to mimic
real existing situations, which in this case were the class and tests timetables. This tool allows for
different resources to be represented which includes the use of sources, servers, paths/connectors,
sinks, and other features as shown in figure 3 to represent different parts of a system. Experimentation
can also be conducted in Simio for models created which cannot be done using excel.

15
Figure 3: Simio Object Resources

The features of most importance in this study include:

The source which is used to generate entities of different types for a particular simulation.

Output Node

Figure 4: Source Object

Servers which are used as processing stations where different operations can be performed.

Input Node Output Node

Figure 5: Server Object

Paths which are routes where entities can travel on, between two locations.

Figure 6: Simio Path

Model entity which is the entity that is processed in the system.

16
And finally, the sink which acts as the endpoint for processed entities, this can represent the exit to a
particular system.

Input node

Figure 7: Sink Object

A simulation model is created by simply dragging any of these features to the empty plane/facility
window where different models can be created. This may also include defining aspects such as states
which are variables that are defined with an initial value asserted to them but can change as the
simulation progresses. Then properties which are variables that are defined at the start of the
simulation and do not change throughout the simulation. Processes may also need to be defined with
the use of predefined steps given in Simio which includes the development of a process that a
resource or entity needs to follow. The steps are indicated below.

All these Industrial Engineering tools above were used in a Personal Computer, with an Intel Pentium
processor and 4GB ram.

17
3.3. Sampling

The targeted population includes students and lecturers in the School of Mechanical, Industrial, and
Aeronautical Engineering. The potential number of student participants considered in this study is
equal to 838 and the potential number of lecturer participants is equal to 29. Student participants
considered are students from each year of study from second year to fourth year and lecturer
participants are lecturers which come from each year of study from second year to fourth year. This
includes the lecturer responsible for timetable scheduling. Simple random sampling was used for
student participants who participated in the survey. This included a random selection of students in
the years of study considered which is from second to fourth year. For the interview, lecturer
participants were selected using stratified random sampling, which included dividing lecturers into
groups based on the year of study they lecture. Then one lecturer was selected from each year of study
group to represent that year of study in the interview. So, the sample size of lecturers is four, and for
students, it was dependent on the number of students who completed the survey. Participants
considered are of ages from 18 and above and of all sexes and ethnic groups. This was done to get
greater representation in all the social groups and to prevent bias in the study.

3.4 Data collection procedures

Simio

The following procedure was used in the creation of the Simio models for the class timetables

Required Objects

The following table shows the Simio model objects and what they represent in real life for the models
created in this study.

Table 1: Real-life representation of Simio objects

Simio model Object Real-life

Source Students entering the System
Server Classes/courses
Paths Progression from one course to another
Sink Students leaving the system
Model Entities Students

18
 Class Timetable models

Three models were made for each stream of engineering in the School of Mechanical, Industrial, and
Aeronautical Engineering, these models compare courses of different years of study to assess if there
are clashes or not. Courses for each year of study will be represented by servers as indicated in the
table above. The general form of the models will include Four rows of objects for each day of the
week in the timetable. The first row includes the sources from which the model entities(students)
come from, the second row shows the servers representing classes for the first year of study
considered which are connected to the source by a path that leads the students from the source to the
servers or classes. The third row comprises of servers which represent classes for the year of study
proceeding the first considered in the second row, these are also connected through a path and finally,
the fourth row contains sinks which students can use to exit the system. These objects are also
arranged in columns that represent the timeslots as in the timetables which might have clashes. These
groups of rows and columns for weeks in the timetables are then arranged in semesters. It is also
important to note in this case that for comparison between third year and fourth year only the first
semester is considered since in fourth year the whole second semester is used for the research and
design projects. The following is a representation of one column of the comparisons between a second
year course(Math2011) and a third year course(Math3033) to see if there is a clash in the timeslot
since both are in the same timeslot.

Column 1(10:15-12:00)

Row 1
Monday

Row 2

Row 3

Row 4

Figure 8: Simio model comparison of courses for one timeslot

19
For a semester the model is as shown below

Figure 9: Simio course comparison model for a semester

For the full models for each stream please refer to appendix E. And the simulation files can be
accessed in the following link:
https://drive.google.com/drive/folders/15rypVVAYSNQU_5zilsqtXhVlevqmhUkJ?usp=sharing.

The models recognize clashes by comparing the two courses in the same timeslot and if according to
the rules it is allowable to enrol in both courses in the same timeslot then there is a clash. This is
because though a student is allowed to do both courses according to the rules, they are unable to take
both courses since they are in the same timeslot. This then limits the number of courses done by cross-
registered students.

Model building steps

⮚ Open the Simio User interface

⮚ Drag the source object from the standard library of the Simio model to the workspace
⮚ Drag the server object from the standard library of the Simio model to the workspace

20
 ⮚ Drag the second server object from the standard library of the Simio model to the workspace
⮚ Drag the sink object from the standard library of the Simio model to the workspace
⮚ Connect the output node of the source object to the input node of the first server object using
a path, then connect the output node of the first server object to the input node of the second
server object this is also done using a path. Then finally connect the output node of the second
server to the input node of the sink object.
⮚ For a particular day of the week in the class timetables and for a specific timeslot indicate the
courses for the timeslot by changing the names of the two server objects to the names of the
courses that are attended by students for that timeslot for the two years of study considered.
⮚ Apply progression rules by changing the path selection, through imposing a constraint that
must be met in order for model entities or students to go through the path. This includes
defining state variables where letters are used which in this case X is used, and letters from A
to H, then for a specific course a number is allocated. For example, for MATH2011 the state
variable may be named X1, and as it is repeated throughout the model it uses different letters
such as A1, B1, C1, and so on. These state variables increase in value as the model entities
enter the input node of the server representing a specific course. Then based on whether the
student can do both courses in the same timeslot or not, a process is defined which uses the
scan action which halts a process until a particular condition is true. This process is used to
halt the model entity at the processing stage of the first row of servers. So, in this case, the
condition is defined as the state variable >=1, which allows model entities to go through the
path connecting the two servers if the state variable is greater or equal to 1. Another condition
is defined as the state variable <=0, which allows model entities to pass only if the state
variable is less than or equal to zero. These are defined interchangeably based on whether
students are allowed to do both courses in the same timeslot or not according to the rules.
⮚ To count the number of clashes for each timeslot in the timetables, another set of state
variables need to be defined which increases as the model entities exit the system through the
sinks. This includes representing these state variables by a letter different for each timeslot
and a number that is different or increases for each day of the week considered. For example,
K1 where K is for the 8:00 AM to 9:45 AM timeslot and 1 is for Monday. These are set to
increase every time the model entity enters the input node of the sink.
⮚ These steps need to be repeated for each timeslot, day, week, and semester.

Tests Timetable

For the tests timetable, one model was created which includes all tests written in the year. This model
was used to assess if there are clashes by comparing tests written across the years of study in the
school. In this model comparisons were made for tests that were scheduled to be written in a day, to

21
see if there are clashes or not. So, in contrast to the models for the class timetable, this included only
one row which represented a day, with courses written in the days arranged within the row. The
servers arranged in the row line up in ascending order according to the years of study. It begins with
tests written in second year then to third year and finally fourth year. If any two courses of different
years of study that a cross-registered student qualifies to register for, are written on the same day then
there is a clash, and the student might need to drop one of the courses due to this clash. The model
considers only the days where tests are written. The figure below is a model representation of one day
where tests were written.

Figure 10: Simio Course Comparison model for a single Test day

Refer to Appendix E for the full-year model. In the full model, different days are represented by
different colours. The simulation file can be accessed in the following link:
https://drive.google.com/drive/folders/15rypVVAYSNQU_5zilsqtXhVlevqmhUkJ?usp=sharing.

Model building steps

The steps in building the model are similar to the procedure outlined above for class timetables. The
exception, in this case, is that the model is arranged in rows which represent test days, also three
courses, in this case, can be compared on the same day. This comparison includes comparing the first
two courses written, which are a second-year course and third year course. Then comparing the last
two which are the third-year course and fourth year course. State variables that count clashes are
defined in the input node of the Third-year course server and the Fourth-year course server.

Interview and survey

The following defines the procedure used in collecting data through interviews and surveys.

1. Identify and select a target population with participants suitable for the study, in this research, it
included students and lecturers, so the population needs to be selected based on the needs of the study.

2. Since this study involves human beings as participants, an application for ethics clearance needs to
be done which involves filling an ethics application form that provides information about the type of
study you are conducting and the level of risks it will put participants in. The application also involves
developing documents such as participant’s information sheet which informs participants of the study
and what their participation entails. A consent form is also needed which allows participants to give
consent to participate in the study, and lastly, a permission letter needs to be applied for since research
is conducted in a particular organization’s premises using its stakeholders.

22
3. Develop the sampling plan to determine the types of respondents to include in the research, this
includes selecting a sample size that is statistically significant relative to the targeted or given
population. It includes stating or calculating the confidence level that the study will consider thus the
level /margin of error allowable for the study. Make use of the sample size calculator to obtain these
variables.

Survey

4. Select the platform for developing the survey such as Google forms, survey monkey, or other
platforms. The selected platform needs to be suitable for the research study and comfortable to use for
the researcher. Select a platform that will be able to display participants’ data such as the number of
participants who have responded, the response data, and must allow to a minimal extent some level of
analysis such as the graphical representation of respondents information.

5. Develop questions for the survey as explained in the instrumentation subsection under the survey
tool. While developing the survey the researcher needs to understand the kind of information that is
needed from the participants for this study, and also the type of participants involved in the study.

6. Once the development of the survey is completed, send the survey to the targeted participants using
the internet university email, this includes embedding a survey link in the emails sent to participants.
This also includes written participants information which informs the participants of the research
study, the importance of their contribution, and contact details for any concerns. Consent for
participation in the survey is assumed through student’s submission of the completed survey. And
through a simple click of the link embedded in the emails, the students should be directed to the
survey platform with the questions to be answered. Also contact the class representatives of each class
in the school through email, for them to distribute the invitation in their class WhatsApp groups.

7. Collect the responses and record them on an Excel spreadsheet for processing, when using google
forms as your survey platform this includes downloading an excel spreadsheet with the students’
responses.

8. Process students’ responses using statistical measures and comparison methods as described in the
data processing section to follow.

9. After the responses have been processed, analyze the results thoroughly to understand and get any
trends or correlations on the processed responses, this includes graphical methods to display
processed results for easier analysis of the data.

10. Discuss the analysis of the processed results in relation to the research question, problem, and
objectives. This also includes comparing the results to the literature review to see how the results
obtained from the research are related to previously conducted studies, whether there is any

23
correlation or consolidation in terms of the information covered. This includes a comparison of key
topics covered in this study and whether the information is sufficient to support the motivation for the
study.

Interviews

11. Develop Interview questions suitable for the study and that allow lecturers to address issues
related to the critical research question and objectives. This must be designed and developed in such a
way as to allow the interviewer some form of control over the interview session and the interviewee
enough freedom to express their views on all the aspects of the research study. This includes
following the standards of good interview questions.

12. Develop an interview protocol that will be used to facilitate the interview procedure. This includes
developing the ground rules to be abided to in the interview and the procedure that will be followed
which includes the allocation of time to be spent on each section and the order to be followed in the
session.

13. After developing the interview questions and protocol, send an invitation to targeted participants
to participate in the study through the university email. This email should include participant’s
information such as details on the research study and what their participation entails. The email
should also include an attached consent form which participants need to fill and complete before any
interview can be conducted.

13. Choose a media which will be suitable for conducting the interview or hosting the interview
session. This includes selecting an interview instrument based on the nature of the study.

14. Set up the interview sessions in the platform considered, which in this research study Microsoft
teams was selected. This will inform participants to join the interview session, once it is set up.

15. Then conduct the interviews in the selected platform and ensure that participants are well aware
of the study, the rules and procedures that the sessions will follow.

16. Before the participant answers any of the interview questions, make sure that the session is being
recorded for future use. This should also be communicated with the participants beforehand.

17. After the session, transfer the responses to an excel spreadsheet by typing the responses of the
participants from the recorded sessions. This also includes noting down even the reaction of the
participants with the change in the questions asked.

18. Process and analyze the responses from the participants which includes developing a coding tree
for sorting and to better understand participants’ responses.

24
19. Discuss the processed data in relation to the research question, problem, and objectives. Similar to
the survey this also includes comparing the results to the literature review to evaluate if there is no
concurrence in the data obtained to previous experimental studies performed.

3.5 Data Processing and Analysis Procedures

The following are the steps used to process the data collected or investigated.

Simio

Calculate the number of clashes in each timeslot for each stream of engineering. This includes
counting the number of state variables defined in the input node of the sink in the fourth row of the
timeslot. If the value of the state variable is greater than 1, then this represents a clash. These clashes
are then counted based on the number of state variables greater than 1.

For the tests timetable, calculate the number of clashes in each day where tests are written across the
years of study. This includes counting the number of state variables defined in the input node of the
Third and Fourth-year server nodes. If these are greater than one, then this represents clashes in the
timetable.

Then draw a bar graph that depicts the number of clashes that occur in the class timetables for each
stream and in the tests timetable of the school. This includes having bar graphs for each stream on the
same axis for easy comparison.

Interviews and survey

Develop a coding tree to sort the interview responses and for graphs to be drawn. This includes
developing attribute and substantive codes for the different questions that participants will be asked.
These codes should accommodate participants’ responses.

Then for questions in the interviews and survey select the responses to questions more suitable to the
study. The percent number of respondents for each of these questions should be calculated. After the
percent number of students has been calculated, draw a graph to represent the number of respondents
for a particular question, the type of graph or chart should be based on the question that was
answered.

To compare means of the groups of datasets use a t-test, which in this research was used to compare
students’ and lecturers’ ratings of the timetables. And to also assess if there is a difference in the
preferred lecture durations by lecturers and students.

Then for comparisons between three means or more for a specific factor, perform an ANOVA
analysis in excel. In this research, a single factor ANOVA was used to find out whether there was a
difference in how students with different course loads have rated the class timetable or tests timetable.

25
It consisted of identifying the factor to be analyzed, the levels of data to the factor, then for each level
the response data is written. Then using the data analysis feature in excel calculate the F which is the
factor level mean, compare this mean to the critical factor obtained through the F tables or through
excel.

To check the correlation of data points, regression analysis was used which included plotting data into
a scatter plot and using the trendline function in excel to draw the line of best fit. In this research, the
linear and polynomial correlation was assessed through the trendline option. And the regression
coefficients were obtained through this option which measured the correlation of the data.

3.6 Validity and reliability

Internal Consistency is ensured through the strong relation amongst students’ responses and the
relation of these results to the results obtained from the interviews conducted for lecturers. This is also
strengthened by the concurrence of the research study to previous experimental studies conducted. All
these ensure consistencies in the research study and thus reliability.

Construct has been ensured through correlation in themes of the research study such as the correlation
of the response data for class timetables and for the tests timetable. This then ensures high construct
validity.

The research study covers all aspects related to timetable scheduling such as time, space,
environment, and people. The study takes all this into consideration and thus ensures that all aspects
that are intended for the study are all measured thus ensuring validity through content.

The survey was also tested by a random student in the School of Mechanical, Industrial, and
Aeronautical Engineering prior to officially sending invitations to the larger student body, to test if the
survey correctly measures what it is intended to measure. The research instruments were also sent to
the supervisor who checked whether the questions align with the subject matter.

3.7 Ethical Considerations

● Participants’ information gathered from the survey and interview will always remain
confidential and can only be shared with relevant parties such as the supervisor.
● Participants of the research were asked for their consent before the surveys were distributed
and interviews were conducted.
● Participants of the research were briefed about the purpose and the nature of the research.
● The use of offensive and other unacceptable language was avoided when formulating the
survey and conducting the interview.

26
 ● Ethics clearance was applied for and obtained through the school of MIA Ethics Committee;
the ethics clearance number is MIAEC 049/21. Refer to Appendix I and J for Ethics
Clearance Certificate and Registrar Approval Letter.

3.8 Risks Mitigation and Contingency Plans

● Interviews were conducted through an online platform to practice social distancing and to
mitigate the possible risk of transmission of covid19.
● Surveys were also conducted through an online form which was distributed using emails.
Thus, reducing the need to physically distribute paper surveys to participants.
● Soft Copy Documents were used for secondary research and no hardcopy document was
utilized; this was to reduce the risk of transmission.
● Interviews were recorded and saved on multiple devices to prevent the loss of information.

27
 Chapter 4

Data Analysis and Results

4.1. Data

Simio Model

As mentioned above the Simio models make use of the class timetables in the School of Mechanical,
Industrial, and Aeronautical Engineering for all streams and all years of study. Refer to reference [22]
for the class timetables. Below is the summary of the courses the models compare in each year of
study.

Table 2: Second to Fourth year courses

2nd Year All 3rd Year 3rd Year 3rd Year 4th Year 4th Year 4th Year
streams Mechanical Industrial Aeronautical Mechanical Industrial Aeronautical
ELEN2000A MATH3026A MECN3004A MATH3026A MECN4020A MECN4009A MECN4020A
Electrical Mathematical Industrial Mathematical Systems Manufacturing Systems
Engineering Methods Engineering Methods Management Technology: Management
Design and Integration Systems and Integration
MATH2011A MECN3041A/ME MECN3042A/ MECN3040A/ MECN4023A MECN4015A MECN4024A
Mathematics II CN3007A MECN3006A MECN3003A Mechanics of Business Gas Dynamics
Mechanical Industrial Aeronautical Solids II Studies and Propulsion
Engineering Engineering Engineering
Investigation Investigation Investigation
MECN2010A MECN3010A MECN3012A MECN3038A MECN4029A MECN4020A MECN4025A
Introduction to Mechanics of Mechatronics I Aeronautical Mechatronics Systems Aerodynamics
Materials Science Solids I Engineering II Management
and Engineering Design and Integration
MECN2011A MECN3012A MECN3013A MECN3043A MECN4031A MECN4028A MECN4026A
Applied Mechatronics I Business Manufacturing Compressible Decision Flight Dynamics
Mechanics A Management Processes Flows Support and
Intelligence
Systems
MECN2012A MECN3013A MECN3014A MECN3008A MECN4032A MECN4030A MECN4027A
Computing Skills Business Operations Introduction to Energy Operations Aircraft
and Software Management Management: Aeronautics Conversion Management: Structures II
Development Techniques

28
 and Utilisation Systems
Systems Integration
MECN2013A MECN3043A MECN3025A MECN3012A MECN1004A MECN1004A MECN4029A
Applied Manufacturing Manufacturing Mechatronics I Selected Selected Mechatronics II
Mechanics B Processes Technology: Topics in Topics in
Processes Social Science Social Science
MECN2022A MECN3039A MECN3026A MECN3013A MECN1004A
Fluid Mechanics Mechanical Principles of Business Selected Topics
Engineering Organisational Management in Social
Design Behaviour Science
MECN2024A MECN3027A MECN3028A MECN3027A
Engineering Mechanical Engineering in Mechanical
Thermodynamics Vibrations its Social Vibrations
Context
MECN2025 MECN3028A MECN3030A MECN3028A
Mechanical Engineering in its Operations Engineering in
Engineering Social Context Research its Social
Design I Context
MECN2005A MECN3032A MECN3031A MECN3032A
Mechanical Numerical Mathematical Numerical
Engineering Methods and Topics Methods and
Laboratory I Statistics (Industrial) Statistics
MECN2026A MECN3036A MATH3033A MECN3035A
Machine Incompressible Mathematical Aircraft
Elements Flows Methods Structures
(Industrial)
MECN3037A MECN3036A
Fundamentals of Incompressibl
Heat Transfer e Flows

The above courses were compared for different years of study using the method outlined above in the
procedure subsection, this included the use of rules such as prerequisites and corequisites within the
faculty rule and syllabus book. These prerequisites and corequisites for each course are presented in
reference [23].

29
Survey responses

There was an overall response of 25 students for the survey. The following is a summary of the
responses recorded in the survey. The tables show the characteristics of the response data based on
students’ demography.

The following table shows the number of responses based on streams.

Table 3: Responses Per stream

Degree Number of responses

Mechanical 6
Industrial 14
Aeronautical 5

The table that follows shows the responses of students based on their year of study.

Table 4: Responses per year of study

Year of study Number of responses

Second 3
Third 5
Fourth 12
3rd -4th year cross-registered 5

The table that follows shows students course load

Table 5: Students responses per course load

Number of courses Number of responses

2 1
3 1
4 1
5 5
6 2
7 3
8 5
10 4
11 2
13 1

30
This last table in the data section shows the number of responses per question in the survey, it is for
questions after the demographic section.

Table 6: Responses per question

Questions Number of responses for each question

4,5,11,12,13,14,15,16,18,19,20,21,22,23,24,25, 25
6,7,8,10 2
9 9
17,26 20
27 21

Uncertainty analysis

The survey was sent out to a population of 270 students in the school of Mechanical, Industrial, and
Aeronautical Engineering, and as stated only 25 students completed the survey. Using the sample size
calculator as shown in appendix K the margin of error was obtained at a 95% confidence level.

The margin of error or confidence interval obtained is 18.71%, this then shows how much uncertainty
exists in the students’ responses. So, we are 95% sure that the true value which is reflective of the
population falls within ±18,71% of the confidence interval.

Interviews

There were 4 lecturers interviewed and there was 100% participation by lecturers in all the interview
questions.

Uncertainty analysis

The total population for lecturers is 29 as expressed in the sampling section. Using the sample size
calculator as shown in appendix K the margin of error was obtained at a 95% confidence level. The
margin of error or confidence interval obtained is 46.30%, this then shows how much uncertainty
exists in the lecturers’ responses. So, we are 95% sure that the true value which is reflective of the
population falls within ±46.30% of the confidence interval.

31
4.2 Data Processing

Simio

As stated in the procedure sections to count the number of clashes a set of state variables are defined
which increases as model entities enter the input node of the sink object or for the tests timetable this
occurs when the entity enters the input node of the third year and fourth year course server. To count
the number of these state variables with values greater than 10 at the end of the simulation, a state
variable is defined in the processing stage of the sink object. And at the end of the simulation, it
counts the number of state variables with values greater than 10, which then represents clashes in the
timetables. In the case of the tests timetable, this is defined in the third year and fourth year course
server. The state variable in this research was defined to be ‘L’. This increases if the state variables
defined at the input node are greater than 10 which is the number of allowable entities for each
timeslot or day in the case of the tests timetable. These state variables in this research are defined as
‘K, W, Y, Z’. These are used 14 times in the simulation models. Where in the tests timetable model
only the ‘W’ and ‘Y’ variables were used. For every K1-14, W1-14, Y1-14, and Z1-14 greater or
equal to 10(>=10), ‘L’ increases by 1 or L+1.

Experiments were conducted in Simio using the experiment feature in the top ribbon, this then
showed the value of L at the end of each simulation in a response named clashes. This is shown in the
figures below.

Figure 11: Experiment for the Industrial Engineering stream

32
Figure 12: Experiment for the Mechanical Engineering stream

Figure 13: Experiment for Aeronautical Engineering stream

Figure 14: Experiment for the tests timetable

33
Survey and Interview

As indicated in the procedure section, for the processing of the interview questions a coding tree was
developed. This includes developing attribute and substantive codes. From the interview questions in
appendix B, the attribute codes include sorting responses from the demographic section. Then the
substantive codes sort other sections more related to timetable scheduling. Below is a summary of the
questions covered in each set of codes.

Attribute code
Substantive Code
Questions: 1, 2 Interview
Questions: 3-19

Figure 15: Coding tree Structure

The coding tree can be seen in appendix D, this includes the substantive and attribute codes that were
used to transform interview responses into graphs displayed in the results for interviews.

Graphs

For the graphs or figures which show the percentage number of respondents to a question. The
following equation was used which calculates the percentage number of students that responded to a
question in the survey. These were then plotted using the insert graph tool on Microsoft excel.

𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑒𝑛𝑡𝑠 𝑓𝑜𝑟 𝑡ℎ𝑒 𝑞𝑢𝑒𝑠𝑡𝑖𝑜𝑛

Percentage number of students= 𝑡𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑝𝑎𝑟𝑡𝑖𝑐𝑖𝑝𝑎𝑛𝑡𝑠
×100

Equation 1: Percentage number of respondents for each question

Pie Charts

To represent the percentages in pie charts the following equation was used to determine the angle
distribution in the pie chart with the use of a percentage number of students.

𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑎𝑔𝑒 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑠𝑡𝑢𝑑𝑒𝑛𝑡𝑠

Angle= 100
×360

Equation 2: Angles calculation for the pie chart.

Mean

To calculate the average for a particular dataset the following equation was used.

Equation 3: Mean value Calculation

34
Where xi is for an individual value within a dataset and n is the total number of values in the dataset.

Standard Deviation

And for the standard deviation, the following equation was used.

Equation 4: Standard Deviation Calculation

ANOVA

For comparisons between means when there are more than two levels of data to a single factor, the
single factor ANOVA is used. After identifying the factor to be measured and the data for each level,
the data analysis feature is used as follows.

Figure 16: Anova Excel

The rows and columns of data to be compared are written in the input section in the image above and
based on how they are arranged in the excel spreadsheet the rows and columns options are chosen.
The level of risk is applied in the alpha section based on the confidence level selected for the study,
where in most cases the default 5% risk is used as seen in the image. This means the analysis is at
95% confidence level. So alpha(α)= 100%-95%.

Alternatively, this can be performed through the following process.

The total number of data responses is calculated using the following equation.

35
𝑁 = ∑𝑎𝑖=1 𝑛𝑖

Equation 5: Number of response data

Where n is the number of responses for each level and 𝑎 is for the total number of levels.

Then for the sum of data response values in each level of data the following equation is used

𝑖𝑛
𝐴𝑖 = ∑𝑗=1 𝑦𝑖𝑗

Equation 6: Sum of response data values for each level

Where yij is the value of each response in a specific level and Ai is the summed values for each level
of data.

Then the following equation is for the total summed value for all the levels. This uses the summed
values for each level obtained from the use of equation 6.

𝑇 = ∑𝑎𝑖=1 𝐴𝑖

Equation 7: Total Sum of response data values for a factor

Where Ai is the summed value for the responses in each level, then T is the total value obtained from
the summation of these values.

The total sum of squares is obtained using the following equation. It uses the calculations from the
above equations.

𝑎
𝑛𝑖
2 𝑇2
𝑆𝑆𝑇 = ∑ ∑ 𝑦𝑖𝑗 − 𝑁
𝑗=1
𝑖=1

Equation 8: Total Sum of squares

Then the sum of squares due to the factor is obtained using the following equation. This is calculated
using values obtained from the equations above.

𝑎
𝐴2𝑖 𝑇2
𝑠𝑠𝐹 = ∑ −
𝑖=1 𝑛𝑖 𝑁

Equation 9: Total Sum of squares due to factor

Then finally the sum of squares due to error is calculated through the difference of the sum of squares
obtained from the above, this is shown in the equation below.

𝑆𝑆𝐸 = 𝑆𝑆𝑇 − 𝑆𝑆𝐹

Equation 10: Total Sum of squares due to error

36
After the sum of squares, the mean square of the factor is calculated using the following equation.

𝑆𝑆𝐹
𝑀𝑆𝐹 = 𝑣𝐹

Equation 11: Mean squares for factor

Where VF is the degree of freedom for the factor and is obtained using the following

VF = a-1, where a is the number of levels of data.

The mean square of the error is also obtained from the below equation

𝑆𝑆𝐸
𝑀𝑆𝐸 = 𝑣𝐸

Equation 12: Mean squares for error

Where VE is the degree of freedom for the error and is calculated using the following

VE= N-a

To calculate the factor level mean, the equation below is used.

𝑀𝑆𝐹
𝐹=
𝑀𝑆𝐸

Equation 13: Factor level mean

This is then compared with the critical factor obtained from the F- tables in appendix G at a 95%
confidence level.

T-Test

To compare the means of two groups of datasets a T-test was used. This includes defining the null and
alternative hypotheses which indicate what you are trying to find out from the test. It also includes
calculating the mean and standard deviation of the two groups as shown in Equations 3 and 4
respectively. And using the following equation to obtain the t-statistic.

̅𝑥̅̅1̅−𝑥̅ 2
𝑡=
1 1 [(𝑛1 −1)𝑆2 2
1 +(𝑛2 −1)𝑆2 ]
√𝑛 +𝑛 √ 𝑛1 +𝑛2 −2
1 2

Equation 14: t-statistic Calculation

The equation above is used assuming that there is equal variance in the two groups of data.

For a two-tailed test, the following hypotheses were applied.

37
(Null Hypothesis) H0: 𝜇1=𝜇2 There is no statistically significant difference in the means of the two
groups compared.

(Alternative Hypothesis) H1: 𝜇1≠𝜇2 There is a statistically significant difference in the means of the
two groups compared.

The two-tailed tests were used to compare two means as seen from the hypotheses to see if there is
any significant difference between the two groups and not to see if one group is greater than the other.
For such a one-tailed test is used. And the following hypotheses were applied.

(Null Hypothesis) H0: 𝜇1=𝜇2 There is no statistically significant difference in the means of the two
groups compared.

(Alternative Hypothesis) H1: 𝜇1>𝜇2 Group 1 or variable 1 mean is greater than the mean of variable 2

OR

(Alternative Hypothesis) H1: 𝜇1<𝜇2 Group 1 or variable 1 mean is less than the mean of variable 2.

After the t statistic has been calculated the t-critical is obtained through t-tables as shown in appendix
H, the t-critical value is obtained based on the confidence level considered for the study which in this
study a 95% confidence level was considered. This is referred to as t.95 in the table for a one-tailed test
or t.975 for a two-tailed test. Another consideration is the degrees of freedom(df) which is obtained
from adding the sample sizes of the two groups considered and subtracting by two (n1+n2-2). The
obtained t-critical value was then compared to the t-statistic. For a t-statistic greater than t-critical, the
null hypothesis(H0) is considered to be false meaning we accept the alternative hypothesis(H1), and
for a t-statistic less than the t-critical then the null hypothesis(H0) is true meaning we reject the
alternative hypothesis(H1).

Alternatively, through excel this can be obtained by using the data analysis feature by selecting t-test.
The input box displayed when t-test is selected is shown below.

Figure 17: T-test on excel

38
Regression analysis

The regression analysis was done with the use of an excel trend line as explained in the procedure
section above. After a scatter plot is developed the trendline option is used to obtain the regression
coefficient as shown below in red.

Figure 18: Correlation Coefficient

After the regression coefficient has been obtained, for a coefficient greater than 0, there is a positive
correlation while for a regression coefficient less than 0 there is a negative correlation. For a case
where the correlation is 0 then there is no correlation in the data.

39
4.3 Results

Clashes

Simio

The following figure displays the number of clashes identified through experimentations in Simio.
The x-axis indicates the stream of engineering within the school and the y-axis represents the number
of clashes for each stream. Bar graphs for all the streams within the school are within the same axes
for easy analysis and comparison. It can be seen that there are clashes in all streams with the highest
being the Industrial Engineering stream with 33 clashes, followed by the Aeronautical Engineering
stream with 17 clashes, and finally the Mechanical engineering stream with 16 clashes.

35

30
Number of Clashes

25

20

15

10

0
Mechanical Engineering Industrial Engineering Aeronautical Engineering

Stream

Figure 19: Clashes in each stream of engineering

The following bar graph is a graphical representation of the clashes identified in the tests timetable
based on the experiments conducted in Simio. The x-axis indicates the number of clashes that occur in
the tests timetable. There are 11 clashes identified in the tests timetable which the graph displays.

Tests Timetable

0 2 4 6 8 10 12
Number of Clashes

Figure 20: Clashes in the tests timetable

Refer to appendix F for more details on the clashes and where they appear in the timeslots and dates
of the timetables.

40
Survey

The pie chart below shows the percentage number of students who have had clashes in their
undergraduate years and those who have not had clashes. Those who have had clashes responded yes
and those who have not experienced any clashes responded no. Most students have not had clashes in
their undergraduate years, with 64% of them saying No and 36% responding yes.

36%

64%

YES NO

Figure 21: Students with clashes in undergraduate years

The pie chart that follows indicates the percentage number of students who are either cross-registered
or not cross-registered from the students who indicated that they have had clashes in their
undergraduate years. 67% of the students indicated that they are not currently cross-registered which
is most of the students who responded while 33% of them are cross-registered.

33%

67%

Cross Registered Not Cross Registered

Figure 22: Cross Registered students with clashes

41
The figure below shows the percentage number of students who have had clashes for each stream. The
industrial engineering stream has the highest number of students who have had clashes with 87%
while the mechanical engineering stream has the lowest with 0%. Then 13% of these students are
from the aeronautical engineering stream.

100
90
Percent Number of students

80
70
60
50
40
30
20
10
0
Mechanical Industrial Aeronautical
Stream

Figure 23: Students with clashes for each stream

The following chart is for students’ willingness to work extra time to do work that causes clashes in
the original timetable. Most students are willing to do extra work with 56% of them responding yes
while 44% of students responded no.

44%

56%

NO YES

Figure 24: Students willingness to do extra work to reduce clashes

42
The figure that follows indicates the timeslots that students who said yes in the above chart are willing
to use in order to do extra work and also shows the year of study for the respondents. Students were
able to choose from more than one option. It can be seen from the figure that most students are willing
to use their free afternoons to perform extra work with 78% of the students selecting this option in
which 50% are 4th years, then 21% are 3rd years and lastly, 7% are 3rd-4th year cross-registered
students. This is followed by a significant 56% of students who are willing to use free lectures to do
more extra work, of which 21% are 4th years, then 14% are 2nd-year students, another 14% are cross-
registered 3rd -4th year students and the last 7% are 3rd year students. Then both lunchtime and
consultation time are at 21% each, with a similar distribution of 3rd year students, 3rd -4th cross-
registered students, and 4th year students which are at 7% each.

Consultations time 7 7 7

Lunch time 7 7 7

Free Afternoons 21 7 50

Free Lectures 14 7 14 21

0 10 20 30 40 50 60 70 80 90
Percent Number of students

2nd 3rd 3rd-4th 4th

Figure 25: Timeslots for extra work to be done

The following diagram indicates whether more work/courses could affect students’ performance.
Most students indicated that increasing the number of courses can affect their performance. This is
seen by 64% of students who responded yes to the question of which 12% are cross-registered
students and 52% are not cross-registered. And 36% of students do not think it can affect their
performance which includes 8% of students who are cross-registered and 28% who are not cross-
registered.

43
 70
Percent Number of students
60
12
50

40

30 8
52
20
28
10

0
YES NO

Not Cross Registered Cross registered

Figure 26: Effect of course load on students’ performance

Class Timetable

The following diagram displays the students’ ratings of the current class timetable. A 5-point Likert
scale was used, where a scale of 1 indicates strongly dislike and a scale of 5 indicates strongly like.
The figure shows that 44% of students rated/scored the class timetable with a 3 which indicates that
students are just ok with the timetables, the second highest rating is 4 where 36% of students rated,
followed by 12% of students who strongly like the timetable. The lowest being 8% of students who
strongly dislike the class timetable.

50
45
Percentage Number of

40
35
students

30
25
20
15
10
5
0
1 2 3 4 5
Score Rating

Figure 27: Students rating of the current class timetable

44
The following table shows the ratings of the class timetable matched with the average number of
courses that students who rated the timetable are enrolled in.

Table 7: Students rating of class timetable matched with course load

Rating 1 2 3 4 5
Average Number 7,5 0 6,6 7,6 8,7
of courses done

For further understanding of the difference in the averages above, the below table shows results from
an ANOVA calculation conducted at a 95% confidence interval. The table shows F which is the factor
level mean and F crit, which is the critical factor. F has a value of 0,46 while F crit has a value of 3,07
which is greater than F which means that there is no significant difference in the factor level means.

Table 8: Anova analysis on students’ ratings of the class timetable

ANOVA
Source of SS df MS F P-value F crit
Variation
Between Groups 11,10566 3 3,701886 0,462917 0,711173 3,072467
Within Groups 167,9343 21 7,996873

Total 179,04 24

45
In the figure below students’ preferred number of courses per day is plotted. The figure shows that
most students prefer 4 courses per day with 44% of students choosing this option, followed by 40% of
the students preferring 3 courses and 12% opting for 2 courses per day. The least preferred number of
courses per day is 1 with 0% of students opting for this option, which is slightly lower than the second
lowest choice of 5 courses per day at 4%.

50
Percentage number of students

45
40
35
30
25
20
15
10
5
0
1 2 3 4 5
Number of courses

Figure 28: Students preferred number of courses per day

The number of hours per day preferred by students to perform their scheduled timetable activities is
recorded in the figure below. 40% of the students prefer to work for more than 5 hours which is most
of the students, followed by 32% of students opting for 5 hours a day. The lowest hours preferred
being 3 and 4 hours per day which are both at 8%.

45

40
Percentage number of students

35

30

25

20

15

10

0
less than 1 1 2 3 4 5 more than 5
Number of hours

Figure 29: Students preferred number of hours per day

46
The following figure shows the constraints that students consider to affect their class attendance in
both online and contact learning. Students were able to select from more than one option. From the
figure, most students consider time to be the greatest constraint that affects their class attendance with
64% of students choosing this option. This is followed by distance which is at 32% then lecture
hall/classroom space at 12% with internet connection and the lowest selected being the ability to
engage which is at 4%.

None

Ability to engage

Lecture Hall/classroom space

Internet connection

Distance

Time

0 10 20 30 40 50 60 70
Percent Number of students

Figure 30: Constraints that affect students’ attendance of classes

And the pie chart that follows indicates students preferred lecture duration. The chart indicates that
most of the students have no problem with the conventional lecture length of 45 minutes with 40% of
them choosing this option. Though most students prefer 45 minutes, a considerable 36% of students
prefer 35 minutes for each lecture and about 12% of the students prefer more than 45 minutes for each
lecture, another 12% prefer 25 minutes.

12% 12%

40% 36%

25 minutes 35 minutes 45 minutes More than 45 minutes

Figure 31: Students preferred lecture duration

47
Tests Timetable

The following diagram displays the students’ ratings of the current tests timetable. A 5-point Likert
scale was used, where a scale of 1 indicates strongly dislike and a scale of 5 indicates strongly like.
The figure shows that 32% of students rated the tests timetable with a score of 3 which indicates that
students are just ok with the tests timetable, the second highest rating is 2 where 28% of students
rated, followed by 16% of students who strongly dislike the timetable. Both the score ratings of 4 and
5, had 12% of students who chose these options.

35
Percent Number of students

30

25

20

15

10

0
1 2 3 4 5

Score Rating

Figure 32: Students rating of the tests timetable

The following table shows the ratings of the tests timetable matched with the average number of
courses that students who rated the timetable are enrolled in.

Table 9: Students rating of the tests timetable matched with course load

rating 1 2 3 4 5
Average Number 9 6,428571429 6,5 7 9,33333333
of courses done

48
For further understanding of the difference in the averages above, the below table shows results from
an ANOVA calculation conducted at a 95% confidence level. The table shows F which is the factor
level mean and F crit, which is the critical factor. F has a value of 1.2 while F crit has a value of 2.9
which is greater than F which means that there is no significant difference in the factor level means.

Table 10: Anova analysis on students’ ratings of the tests timetable

ANOVA
Source of SS df MS F P-value F crit
Variation
Between Groups 34,65905 4 8,664762 1,200264 0,341451 2,866081
Within Groups 144,381 20 7,219048

Total 179,04 24

To understand the correlation of students’ ratings of the tests timetable and class timetable the
following scatter plot was developed which also shows the squared regression. Where for a
polynomial fitted line the squared regression coefficient is 0,1419 and for a linear fitted line the
squared regression coefficient is 0,0396

4
Rating

R² = 0,0396 R² = 0,1419
3 Series1
Poly. (Series1)
2
Linear (Series1)
1

0
0 10 20 30 40 50 60
Number of responses

Figure 33: Regression Analysis

When taking the square of the squared regression coefficient shown in the plot the resulting regression
coefficient is shown in the table below. This includes a 0.2 linear regression coefficient which
indicates that there is a positive linear correlation between the two ratings. For the polynomial
regression coefficient, the value is much higher at 0.4 which also indicates a positive polynomial
correlation between the two datasets.

49
Table 11: Regression Coefficient

Regression coefficient
Linear Regression 0.2
Polynomial Regression 0.4

The following pie chart shows the number of tests students are willing to write in a day. In the chart,
most students are willing to write only one course per day with 96% of them choosing this option and
about 4% of students are willing to write tests for 3 courses in a day.

4%

96%

1 2 3 More than 3

Figure 34: Students preferred number of tests per day

And students preferred timeslots to write these tests are shown in the pie chart below. Most students
prefer the morning slot with 76% of them selecting this option while 24% of them prefer the afternoon
timeslot to write their tests/ exams.

24%

76%

between 8:00 AM to 12:00 PM morning slot

between 1:00 PM to 4:00 PM Afternoon slot

Figure 35: Students preferred timeslot to write tests and exams

50
The following figure shows the constraints that students consider to affect their test/exam writing
conditions in both online and contact learning. Students were able to select from more than one
option. From the figure, most students consider time to be the greatest constraint that affects their
test/exam writing conditions with 68% of students choosing this option. This is followed by distance
which is at 20% then exam hall space at 16% and the lowest selected being Internet connection and
emotional and physical stresses which are at 12% and 4% respectively.

None

Emotional and physical stressors

Time

Network/ Internet connection

Exam room/hall space

Distance

0 10 20 30 40 50 60 70 80
Percent Number of students

Figure 36: Constraints affecting students tests/exams writing conditions

Interviews

Lecturers workload

The following pie chart was obtained from the interviews conducted and shows the number of courses
each lecturer interviewed lectures in the school. All of the lecturers interviewed are responsible for
two courses in the school, as seen with the 100% response in the chart.

0%

100%

One Two three

Figure 37: Number of Courses for each lecturer

51
The figure to follow shows the percent of lecturers who have other duties in the school besides
lecturing and those who do not have other duties. 100% of the lecturers have other duties besides
lecturing in the school as seen in the figure.

0%

100%

No Other Duties except lecturing Have Other duties

Figure 38: Amount of work performed by lecturers

From those who have answered that they have other duties in the school, the figure below shows the
duties that these lecturers are responsible for. Most of the lecturers interviewed are involved in school
committees as seen with 50% of them in the figure. Some have admin roles such as undergraduate
affairs which are tied with the percent number of lecturers who are doing research at 25% each.

School Committees

Admin Role

Research

0 10 20 30 40 50 60
Percent Number of lecturers

Figure 39: Other Duties performed by lecturers

52
Lecturers Timetable experience

The following figure shows how lecturers feel about the scheduling system or the timetables. Where
100% of them indicated that it is good enough, well planned and there is sufficient time given.

0%

100%

Really good, easy to work with

Good enough, sufficient time given
It is ok, just enough time given, manageable
Not good, there time is not sufficient hard to follow
Really Bad, its really complex and not well arranged, not enough time for lectures.

Figure 40: Lecturer's thoughts on the scheduling process

Then a t-test was conducted as shown in the table below to assess if there is a difference between
students’ ratings of the academic schedule and lecturers’ ratings of the schedules. This was done by
turning the legend in figure 40 above into a numeric Likert scale, so 1 represents ‘Really Bad, it is
really complex and not well arranged, not enough time for lectures’ and the extreme of 5 represents
‘Really good, easy to work with’. From the table, it can be seen that the t statistic is lower than the t
critical value for both two-tailed and one-tailed tests, which means there is no significant difference
between the two means.

Table 12: T-test for ratings on Timetables

Variable Variable 2
1
Mean 4 3,44
Variance 0 1,006667
df 27
t Stat 1,099316

t Critical one-tail 1,703288

t Critical two-tail 2,051831

53
The following figure shows the constraints that affect the class and test timetable based on lecturers’
responses. As seen in the figure all the constraints have equal percentages of lecturers which is at
25%.

30
Percent Number of lecturers

25

20

15

10

0
Coding that comes with Time-Preparation and Number of tests and Arrangement of tests
change in arrangement Marking Assignments and assignment
and number of courses
Contraints

Figure 41: perceived constraints on test and class timetables by lecturers

Proceeding from the figure above, the following bar graph shows the constraints that lecturers
consider to affect their lecturing schedule. The figure shows that most lecturers are constrained by the
extra activities in the school outside lecturing.

Activities in committees such as meetings

Constraints

Administrative activities

0 10 20 30 40 50 60
Percent Number of lecturers

Figure 42: Constraints on the lecturing schedule

54
The figure below shows the ideal class size that lecturers prefer. It can be seen that there is an equal
split in that half of the lecturers prefer less than 50 students while the other 50% can handle more than
50 students.

50% 50%

As big as possible, more than 50 students A small as possible, less than 50 students

Figure 43: Preferred Class Size by lecturers

The following bar graph is for the number of hours that lecturers spend doing academic activities
which also includes other duties in the school outside lecturing. From the graph, it can be seen that
most lecturers spend more than 7-8 hours doing academic activities.

120
Percent Number of Lecturers

100

80

60

40

20

0
More than 7/8 hours Normal work hours, 7-8 less than 7 hours
hours
Working Hours

Figure 44: Number of hours spent on academic activities by lecturers

55
Lecturers Preference

This next chart shows lecturers’ preferred lecture timeslot. There is an equal split with 50% of
lecturers preferring the morning timeslot while the other 50% prefer the afternoon slot.

50% 50%

Morning slot Afternoon slot

Figure 45: Lecturers Preferred timeslot

In terms of the lecture duration, the following chart specifies lecturers’ preference in lecture duration.
As seen from the chart most of the lecturers interviewed prefer more than 45 minutes for their
lecturers.

0%

100%

More than the conventional 45 minutes The conventional 45 minutes

Less than 45 minutes

Figure 46: Lecturers Preferred lecture duration

56
A t-test was then conducted to assess if there is a difference in the average time for lectures preferred
by students to the average time preferred by lecturers. This is also assuming there is equal variance.
For cases where both parties indicated that they preferred more than 45 minutes, this was then
allocated 90 minutes for two lecture periods. From the table, it can be seen that the t statistic is greater
than the t critical value for both two-tailed and one-tailed tests which means there is a significant
difference between the two means. And variable 1 mean which is for lecturers is greater than variable
2 mean which is for students.

Table 13: T-test on lecture duration

Variable Variable
1 2
Mean 90 44,4
Variance 0 340,25
df 27
t Stat 4,869037
t Critical one-tail 1,703288

t Critical two-tail 2,051831

Future Schedules

The following chart shows students’ willingness to work on an integrated learning system that would
offer students courses that they have clashes for online, even when physical lectures continue. 80% of
students are willing to work in this system while 20% are not willing.

20%

80%

YES NO

Figure 47: Students willingness to work on an integrated system for clashes

57
The figure below then shows the method of online learning that students are willing to use for the
integrated system mentioned above for students who said yes. A significant 65% prefer a combination
of both synchronous and asynchronous online learning methods while 20% prefer asynchronous
learning and the last 15% prefer synchronous learning.

15%

65% 20%

Synchronous Asynchronous A combination of both

Figure 48: Students preferred learning method to reduce clashes

Since clashes are caused by students failing certain courses which results in them having to do two or
more courses of different years of study in the same timeslot. Students were then asked if they would
accept a timetable that proactively changes based on the failure rate and types of courses failed. The
following chart shows students’ responses. 92% of students indicated that they would accept the
change while 8% reject the proposed change in the timetables.

8%

92%

YES NO

Figure 49: Students response to a timetable that changes with the failure rate

58
Chapter 5

Discussion

Clashes

Improving timetables in this research is majorly dependent on reducing the clashes that exist in the
timetables, as stated in the research question. So, it is imperative to understand the existence of
clashes in the timetables and where these clashes appear. Through the use of Simio modelling as
explained in the research method clashes were identified, along with where they appear. This is
displayed in figure 19 where clashes for each year of study are plotted. In the bar graph, the most
clashes are apparent in the industrial engineering stream as stated and the lowest clashes are in the
Mechanical Engineering stream. This difference in the number of clashes is due to the separation of
streams in the third year of study, where for streams such as industrial engineering the curriculum
becomes more specialized in the field. And courses are less dependent on the courses done in the
second year of study which is common for all streams. This then causes an increase in clashes because
cross-registered students might qualify to do courses of two different years of study because the
courses are not interdependent, meaning second year courses are not prerequisites to higher year
courses such as third year courses.

These clashes are not only experienced in the class timetables but also in the tests timetable as shown
in figure 20, where 11 clashes were identified that could be a possible reason for students' omission of
certain courses in their studies. As shown in figure 10 which is a Simio model for a day of tests,
multiple tests across all years of study can be written in the same day which mostly occurs in the same
timeslot which is the Monday morning timeslot, that is designated for tests. And based on figure 34
where most students indicated that they prefer only one test per day, this then makes it impossible for
students who qualify for courses in two different years of study to write tests for these courses if they
are scheduled on the same day. Thus, causing a clash.

Though most students who participated in the survey do not have clashes as shown in figure 21, a
significant 36% of them have experienced clashes in their undergraduate years, of which 33% are
currently cross registered as shown in figure 22. Figure 23 further supports the existence of clashes,
where 87% of industrial engineering students have experienced clashes, followed by aeronautical
engineering students at 13% and mechanical engineering students with 0% of students who have
experienced clashes. These differences in clashes across streams is reflective of the clashes discovered
using the Simio models and displayed in figure 19.

59
These clashes discovered in the Simio experimentation and through surveys support the problem
statement and motivation in chapter 1 above. These show the importance of this study and how it
affects students in the School of Mechanical, Industrial, and Aeronautical Engineering.

Students willingness to reduce clashes

In figure 24, all students who participated in the survey were asked whether they would be willing to
do extra work to reduce clashes, and most students who responded indicated that they would be
willing. This then concurs with the literature review, in [10] which indicated that students would be
willing to substitute some of their leisure time to do more academic activities if given the opportunity.
In the case of respondents to the survey conducted in this research, most students in figure 25
indicated that they are willing to use their free afternoons from 1:15 pm to 4 pm to perform more
academic activities of which 50% are fourth year students. The second timeslot students are willing to
use is free lectures, so instead of having free periods, students are willing to use these slots to perform
more academic activities.

Though more students are willing to perform more academic activities, this aspect is also subject to
lecturers’ availability as expressed in the literature review as one of the constraints. And in figure 37 it
is shown that 100% of lecturers interviewed lecture more than one course. In figure 38 it is also
shown that these lecturers have other duties within the school they are responsible for, which include
being part of school committees which most lecturers are part of, also administrative roles, and their
own research studies, as shown in figure 39. All this workload reduces lecturers’ availability making
it difficult to allocate more work or lecture sessions. This can also be seen with the number of hours
lecturers work in figure 44, which exceeds the normal working hours if classes start at 8 am and end
at 4 pm.

Another constraint indicated in the literature review and investigated in this study is the effect of
increasing workload on student’s performance where in figure 26 a significant 64% of students
believe that it could affect their performance. About 12% of these students are cross registered which
is still greater than the percent for those who think it would not affect their performance. This then
shows that an increase in workload could badly affect the performance of all groups of students. This
is in contrast to the previous studies performed and indicated in the literature review which argued
that increasing the number of courses does not affect performance but helps align students with their
studies.

Class Timetable

Though the class timetables have clashes, most students in figure 27 have rated the timetable to be ok
or manageable and they neither dislike it nor like it. And in table 8 through Anova, it was found that
there is no significant difference in how students of different course loads have rated the timetables.

60
Meaning that the students’ ratings are not subject to coarse load. In figure 40 the same procedure was
applied for lecturers and most lecturers rated the class timetable to be good enough and that sufficient
time is given. In table 12 the difference between students and lecturers ratings of the class timetable
were investigated and it was found that there is no significant difference between the scores of the two
groups meaning that both students and lecturers have no problem with the current class timetable.

To assess the optimal number of courses suitable for students, figure 28 was developed which shows
that most students prefer 4 courses or less a day in the timetable and the preferred number of hours per
day to work on academic activities is more than 5 hours for most students in figure 29. So taking into
account the lecture duration preferred by students in figure 31 which is 45 minutes for each lecture
and considering the number of courses preferred per day it means that students are willing to work for
3 hours in these courses, leaving more than 2 hours for other academic activities in the preferred
number of hours per day.

As stated in the discussion above most students in figure 31 prefer the conventional 45 minutes of
lecture duration, which is in contrast with the argument presented in the literature review. The
literature indicates that a student's attention span is between 10 to 18 minutes on average. Though
40% of students prefer 45 minutes, a significant 36% prefer 35 minutes which is 10 minutes lower
and may support the previous literature which indicates that the smaller the lecture duration the
greater the ability of students to recall and understand the content presented in the lecture. And even
though this is the case for students, most lecturers in figure 46 prefer more than 45 minutes for their
lectures which differs from both literature and students’ preference. Lecturers indicate that 45 minutes
is often not enough to cover all the content and that they are sometimes subject to cutting the content
down in order to present it in 45 minutes. Table 13 then shows a t-test performed to assess if there is a
difference in the preferred lecture duration between students and lecturers and it was discovered that
there is a significant difference between the two. Using the one-tailed test it is clear in the table that
the average lecture duration preferred by lecturers is significantly greater than the students’
preference. This then further shows that lecturers prefer more time to lecture and students are less
willing to attend long lectures. This conflict in preferences may be due to the contrast in lecturers’
desire to cover as much content as possible and students’ attention span.

In relation to the constraints which affect the students’ attendance of lectures scheduled in the class
timetable most students in figure 30 feel that time is a major constraint in both online and contact
learning. This is followed by distance in the case of contact learning. Time as a major constraint is
also correlating with the literature review which indicated that a major constraint in timetable
scheduling is the time available for teaching and learning which includes the starting time and the
ending time for lectures. Another constraint that 12% of students considered in figure 30 and which is
indicated in the literature review is the classroom space, which is followed by 4% of students who

61
said the ability to engage which relates to student engagement as a performance objective considered
in the literature review. This then means that for an optimal timetable suitable for students there has to
be consideration of the starting time and ending time, level of engagement allowed for students, and
enough space or an accommodative meeting platform.

Tests Timetable

Similar to the class timetable, most students in figure 32 rated the tests timetable to be okay or
manageable, they neither like it nor dislike it. Though in contrast to the class timetable where most of
the students’ responses are concentrated from a rating of 3 and above, for the tests timetable it is the
opposite. Most of the students’ responses for the tests timetable are mostly concentrated from 3 and
below. This can be seen with a significant 28% of students in the figure who dislike the timetable. So,
a significant proportion of students do not like the current tests timetable. ANOVA was again used in
this case to assess if students’ ratings are subject to students’ workload as shown in table 10. This
further showed that there is no significant difference in how students with different course loads rated
the tests timetable, meaning the tests timetable ratings were not subject to students’ workloads.
Though there is no significant difference in students’ ratings based on workload, figure 33 and table
11 show that there is a positive correlation in students’ ratings of both the tests timetable and class
timetable. This means that though there might be slight differences in the ratings for both timetables,
these are closely related, a decrease in the rating of one is indicative of a likely decrease in the rating
of the other.

For the number of tests that students are willing to write in a single day, as stated before figure 34
indicates that a dominating 96% of students prefer to write one test a day and most students in figure
35 prefer to take these tests in the morning timeslot. This then concurs with the literature which
indicates that students tend to perform better in the morning timeslot rather than in the afternoon due
to cognitive fatigue. This cognitive fatigue may also be the reason why most students prefer only one
course per day. And about 50% of lecturers interviewed also prefer the morning timeslot for academic
activities as shown in figure 44 while the other 50% prefer the afternoon timeslot.

In consideration of the constraints that affect students’ tests/exams writing conditions in figure 36, it
is shown that most students consider time as a major constraint in this regard, which is similar to
constraints considered for the class timetable. This constraint of time in tests timetable concurs with a
study explained in the literature review which shows that students perform better when they are not
subjected to time limits. The study also associates time as a factor contributing to increased anxiety by
students during tests and examinations, which can affect their performance.

62
Class size

Another important aspect that is considered in the study is the class size as addressed in figure 43. In
the figure, there is an equal split with 50% percent of lecturers preferring smaller classes which are
lower than 50 students and the other 50% of lecturers indicating that they can handle bigger classes
which are more than 50 students. For lecturers with smaller classes, their reason was that their courses
require engagement with students on an individual basis and not in groups. This then concurs with the
literature review on student engagement and class size. Where for student engagement, the more
engagement with the lecturers’ students get, the more their performance improves. For the class size,
the literature review indicates that the reduction of class size increases student’s achievement. This
may be due to the fact that the aspect of class size affects lecturers’ availability and thus the academic
support. So, for a large class size, the less available the lecturers become and thus the poorer the
academic support provided to students. So, in this case, there should be a trade-off between an
increased class size thus exposing as many students as possible to courses and the amount of
academic support these students will be offered which is linked to performance.

Future Schedules

In relation to students’ willingness to do more work to reduce clashes in the discussion above, in
figure 47, 80% of students also indicated that they would be willing to enrol in an integrated online
system that would offer students courses they have clashes for. The majority of these students in
figure 48 prefer a combination of both synchronous and asynchronous online learning methods at
65%, which is followed by asynchronous learning at 20%. All this further supports the fact that
students would be willing to do extra work to reduce clashes. Another change to the schedules was
also proposed in figure 49 which includes a timetable that proactively changes based on failure rate
and an overwhelming majority of 92% of students accept this change which could reduce clashes in
the timetables.

63
Chapter 6

Conclusions and recommendations

It is evident from the analysis above that indeed the academic schedules consist of clashes which
could result in students’ omission of courses in their undergraduate years. These clashes through the
examination of the academic schedules using Simio models occur when two courses from different
years of study which are allowed for cross-registered students, are in the same timeslot. Therefore, it
is possible to reduce these clashes through the rearrangement of courses in the timetables which
include the tests timetable. The analysis also shows that students are willing to perform extra work to
reduce these clashes. And also, through the analysis, it is clear to see that free afternoons and free
lectures can also be used to reduce clashes in the timetables. And since this has a constraint of lecturer
availability which can decrease with the increase in lecture sessions and class size, asynchronous
online learning methods can be used to provide extra courses for which students have clashes for. This
is a method which most students prefer from the analysis which includes those who prefer a
combination of both asynchronous and synchronous learning. This would then reduce the need for
lecturers to always be available for synchronous meetings with students and students would have
more control over their schoolwork.

The optimal number of courses suitable for students to attend was identified as 4 lectures a day which
includes a 45 minutes or lower lecture duration. In consideration of the optimal number of hours
which were identified in the analysis as more than 5 hours, this therefore, means that even though
students prefer attending 4 lectures a day the remaining time can be used for asynchronous learning or
other academic activities. In relation to the first objective achieved of identifying the optimal number
of courses, for the tests timetable, the optimal number of tests for students to write in a day without
compromising on their performance is one, which needs to be written in the 8 am to 12 pm morning
slot. This number also makes it easier to rearrange the tests timetable to avoid clashes.

All this, therefore, proves that clashes in the timetables can be reduced, and below are the more
detailed ways or recommendations on how these clashes can be reduced.

Recommendations

● Through the use of Excel solver, a new timetable arrangement can be determined. This
includes using binary numbers of 0 and 1 to differentiate between courses that cause clashes
in the timetables and those which do not cause clashes. Once this differentiation has been
done, constraints can be applied where for a particular timeslot a number is used as a
constraint on the number of clashes allowed for that particular timeslot. Then using the solver
feature in excel the optimal solution in terms of the reduced number of clashes can be

64
 obtained. And though this has a constraint of coding where a change in the class timetable
means making changes in other departments in the university. This can be overcome by just
changing the arrangement of courses in the timetable for a week of lecturers and not the
course names and where they appear in the semesters. This then means that students still
register for the same courses each semester. Students will be given these new timetables after
registration, so there is no need to change the coded timetables in other departments. Students
will just be informed of the changes made in the timetables compared to the ones they have
seen in other departments or on the wits self-service portal.
● Using MATLAB courses most failed in each year of study within the school can be identified
using students’ marks at the end of the year. So, when courses that are most failed and likely
to cause clashes are recognized then a program changes the arrangement of courses in the
timetables to cater for the failure rate. This solution uses some machine learning aspects and
proactively changes the timetable to suit the failure rate. Meaning that for courses most failed
they are compared with other courses in the timetable and if a course could cause a clash
with the failed course if placed in the same timeslot this course is then moved to a different
timeslot. This could reduce clashes in the timetables and can allow students to register for the
maximum number of courses possible. According to the analyses above about 92% of
students accepted this possible solution.
● In this solution, students are offered an integrated online asynchronous learning system based
on their workload. A system similar to the above is used, where students’ workload is
determined using a MATLAB program that identifies the number of courses cross-registered
students would be doing if they failed a particular course. Then based on the optimal number
of courses and hours described above students are enrolled in the system for courses they
have clashes for in the original timetable. This then will ensure that students are enrolled in
the maximum number of courses possible and that they have a workload that they can
manage.
● In the case of the tests timetable, students should write a single test in a day. The possible
solution for the duration is as proposed in the literature review, that lecturers should first take
the test or exam themselves then multiply the time it took them to finish by three. Courses
that have the greatest failure rate should always use the Monday morning timeslot which as
stated in the literature that for tests written in the morning students tend to perform better.
● In order to counter the constraint of lecturer’s unavailability on certain dates for tests, a
Kanban system can be introduced using software such as Trello. For this system, lecturers
will indicate days where they are unavailable on Kanban cards and which days they have
things to do but can be rescheduled. This information will be written on Kanban cards on the
system, which is accessible to all lecturers, lecturers can then know when they can be able to
schedule tests and assignments easily and clashes can also be avoided. To counter the time

65
constraint expressed above, the timetables can be linked to the outlook calendar which will
then alert students and staff of any academic activity which they need to attend.

66
References
[1] Definitions.net, STANDS4 LLC. Timetable.2021. Available from:
https://www.definitions.net/definition/timetable. [Accessed July 28, 2021].

[2] The Analytics and Institutional Research Unit. Facts & Figures. University of the Witwatersrand,
Johannesburg, 2020. Available from: https://www.wits.ac.za/about-wits/facts-and-
figures/#:~:text=Wits%20University%20has%20five%20faculties,over%20400%20acres%20in%20Jo
hannesburg. [Accessed July 28, 2021].

[3] Oude Vrielink RA, Jansen EA, Hans EW, Van Hillegersberg J. Practices in timetabling in higher
education institutions: a systematic review. Department of Industrial Engineering and Business
Information Management, University of Twente, Netherlands, 2017.

[4] Almond M. An algorithm for constructing University timetables. Mathematics Department, Queen
Marry College, London.

[5] Daniel PG, Dr Maruf AO, Dr Modi B. Paperless Master Timetable Scheduling System. Vol. 8,
No.2, International Journal of Applied Science and Technology, 2018.

[6] UNESCO. School timetable. International Bureau of Education. Geneva, Switzerland. Available
from: http://www.ibe.unesco.org/en/glossary-curriculum-terminology/s/school-
timetable#:~:text=Some%20factors%20that%20influence%20timetabling,and%20the%20availability
%20of%20facilities. [Accessed July 29, 2021].

[7] Larabi-Marie-Sainte S, Jan R, Al-Matouq A, Alabduhadi S. The impact of timetable on students’

absences and performance. PLoS ONE 16(6):e0253256, 2021. Available from:
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0253256. [Accessed July 29, 2021].

[8] Goulas S, Megalokonomou R. The effects of exam scheduling on academic performance. VOXEU,
2020. Available from: https://voxeu.org/article/effects-exam-scheduling-academic-performance.
[Accessed July 30, 2021].

[9] Smith WE. Efficiency of a university timetable an application of entropy of choice, Vol. 30, Pg.
19-26, 1984. Available from: https://doi.org/10.1017/S0004972700001672. [Accessed: July 30, 2021].

[10] Huntington-klein N, Gill A. Semester Course Load and Student Performance. SringerNature,
2020. Available from: https://doi.org/10.1007/s11162-020-09614-8. [Accessed: July 30, 2021].

[11] Bhargana D. Balanced Timetable Key to Student Engagement. 2015. Available from:
https://www.carsonst.wa.edu.au/wp-content/uploads/2016/08/Balanced-Timetable-Manual.pdf.
[Accessed: July 30, 2021].

67
[12] Tenney School. What is the Real Ideal Class Length. 2016. Available from:
https://tenneyschool.com/what-is-the-real-ideal-class-
length/#:~:text=Typically%2C%20classes%20last%20between%20fifty,the%20style%20of%20the%
20classes. [Accessed: July 30, 2021].

[13] Khan S. Why Long Lectures Are Ineffective. 2012. Available from:
https://ideas.time.com/2012/10/02/why-lectures-are-ineffective/. [Accessed: July 30, 2021].

[14] Laiqa R, Shah RU, Khan SM. Impact of quality space on students’ academic achievement, Vol.
3, No.2, International Journal of academic research, 2011.

[15] Whitehurst GJ, Chingos MM. Class Size: What Research says and what it means for state policy.
brookings.edu, 2011.

[16] Onwuegbuzie AJ. The effects of time constraints and statistics test anxiety on test performance in
a statistics course. Journal of Experimental Education, 63(20), pg. 115-124.

[17] Tookoian J. What is the right number of test questions. Edulastic, 2016. Available from:
https://edulastic.com/blog/the-right-number-of-test-questions/. [Accessed: July 30, 2021].

[18] Gregoire C. Scientists reveal the best time of day to take an exam. HUFFPOST, 2016. Available
from: https://www.huffpost.com/entry/best-time-of-day-for-exam_n_56cb0b68e4b041136f177a11.
[Accessed: July 30, 2021].

[19] Finol MO. Asynchronous vs Synchronous Learning: A Quick Overview. BRYN MAWR College,
2020. Available from: https://www.brynmawr.edu/blendedlearning/asynchronous-vs-synchronous-
learning-quick-overview. [Accessed: July 30, 2021].

[20] ION Professional learning programs. Strengths and Weaknesses of Online Learning. University
of Illinois Springfield.

[21] Hammond HG, Coplan MJ, Mandernach BJ. Administrative Considerations Impacting the
Quality of Online Teaching. Available from:
https://www.westga.edu/~distance/ojdla/winter214/hammond_coplan_mandernach214.html.
[Accessed: July 30, 2021.07].

[22] School of Mechanical, Industrial and Aeronautical Engineering. Documents, Timetable, Maps.
University of the Witwatersrand, 2021. Available from: https://www.wits.ac.za/mia/undergraduate-
study/current-and-prospective-students/. [Accessed: August 17, 2021].

[23] The faculty register. University Rules and Syllabuses. pg.83-87, Faculty of Engineering and the
Built Environment, University of the Witwatersrand.

68
Available from:
https://www.google.com/url?sa=t&source=web&rct=j&url=https://www.wits.ac.za/media/wits-
university/students/academic-
matters/documents/2021%2520Engineering%2520and%2520the%2520Built%2520Environment%25
20Rules%2520and%2520Syllabuses.pdf&ved=2ahUKEwip64XY87fyAhWNSsAKHR0yCocQFnoE
CAYQAQ&usg=AOvVaw1u4fDuuCrWTmCZXiLxnN86. [Accessed: August 17, 2021].

69
Appendices

A: Survey

Timetable Scheduling
This survey or questionnaire forms part of research which investigates how timetable scheduling
can be improved in the school of mechanical, industrial and aeronautical engineering thus ensuring
that there are minimal clashes and that students can register to the maximum number of courses
possible. The purpose of this survey is to get an understanding of students' experience and
knowledge of the scheduling system used in your school.
*Required

Demographics

1. Which branch of engineering are you currently studying *

Mark only one oval.

Mechanical

Industrial

Aeronautical

2. What year of study are you in? *

Mark only one oval.

Second

Third

Fourth

2nd – 3rd Year cross registered

3rd – 4th Year Cross-registered

3. Please indicate the number of courses you are currently enrolled in? *

70
Previous and Current Class Timetable

4. On a scale of 1 to 5(where 1 is strongly dislike and 5 is strongly like) please indicate

how you feel about your current work schedule/Timetable *

Mark only one oval.

1 2 3 4 5

strongly dislike strongly like

5. Have you experienced any clashes in your timetables before in your undergraduate
years which caused you to omit some courses? * Mark only one oval.

YES

No

6. If yes, please indicate which courses (If cannot recall Indicate by 'I don't Remember')

7. If yes to question 5 please indicate the year of study

71
8. If yes, Please indicate how not progressing with a certain course due to clashes
affected your learning process

9. Do you think these courses could have been incorporated in your timetable at the
time?

Mark only one oval.

YES

No

10. If yes, please specify how you think this could have been done

72
Students Preferences

11. Please indicate the number of courses suitable for you to participate in on a given day
*

Mark only one oval.

More than 5

12. Please indicate below the number of hours suitable for you to do scheduled timetable
activities in a day * Mark only one oval.

More than 5

13. Please indicate the ideal/Preferred length of time for your lectures *

Mark only one oval.

73
 Less than 10 minutes

10 minutes

15 minutes

25 minutes

35 minutes

45 minutes

More than 45 minutes

14. If you had clashes would you be willing to work extra time to do work that causes
clashes in the original timetable. *

Mark only one oval.

YES

NO

15. If yes, Which of the following time slots would you be willing to use in order to do
more work? *
Tick all that apply.

Consultation times
Free lectures
Lunch time
Free Afternoons(between 1:15-4:00 PM)
Other:

16. Would you be willing to work on an integrated learning system which would offer
students courses which they have clashes for, online when physical lectures
continue? * Mark only one oval.

YES

NO

17. If yes select a method you think would be most suitable from the below [Synchronous
Learning: Live lectures/interactions at set times on platforms such as BBB, MS Teams

74
 and Ulwazi Chatroom/Discussions. Asynchronous Learning: Lecture material
(Slides/Videos/Notes) is distributed available for access at any time]

Mark only one oval.

Synchronous

Asynchronous

A combination of both

Test Timetable

18. On a scale of 1 to 5(where 1 is strongly dislike and 5 is strongly like) please indicate

how you feel about the MIA Test timetables * Mark only one oval.

1 2 3 4 5

Strongly dislike Strongly like

19. Please indicate the number of courses you would be willing to write on a day
*

Mark only one oval.

More than 3

20. Please indicate the time slots most suitable for you to write your tests or
assessments *

Tick all that apply.

between 8:00 AM to 12:00 PM morning slot

… between1:00 PM to 4:00 PM Afternoon slot
Other:

75
21. Please indicate your ideal/preferred length of time that you feel tests or
assessment should run for. * Mark only one oval.

Less than 1 hour

1 hour

2 hours

3 hours

More than 3 hours

Other:

Factors affecting student’s participation in academic activities

22. From the below, please indicate which factors affect your class attendance (during
online or contact learning) *
Tick all that apply.

Time
Distance
Lecture hall/Classroom space
Other:
23. What factors badly affects your assessment/ test writing conditions(during online or
contact learning) *
Tick all that apply.

Time
Distance
Exam room/hall space
Other:

76
Future Schedules

24. Would you accept a timetable that proactively changes based on the failure rate and
types of courses failed? * Mark only one oval.

YES

NO

25. Do you think increasing the number of courses done/course load can affect your
performance * Mark only one oval.

YES

NO

26. If yes, Please indicate how? *

27. What suggestions do you have that could improve the MIA scheduling system? *

77
B: Interview questions

Interview Questions
Demographics
1. What years of study are you responsible for lecturing and which courses do you lecture?
2. What other activities are you responsible for in the school?
Previous and Current Class Timetable
3. What do you think about the scheduling system from a lecturer’s perspective?
Probe: Do you think enough time is given for teaching and learning
5.Do you have consultation sessions?
6. How are they facilitated
Probe: Do students actively participate in the consultation sessions
Probe: Is the time allocated for these sessions always exhaustively used
7. What constraints do you think affect your lecturing schedule and your lecturers?
Probe: How are they affecting them
8. What constraints affect your way of scheduling tests and assignments
Probe: How are they affecting this process
Lecturers Preference
9. How many students are ideal for you in a course to handle (as small as possible or as big as
possible)
10. Which time of the day is most suitable for you, and why
11.How many working hours are ideal for you in a day?

Future Schedule
11. Any suggestions you have for improving the timetable to be more suitable for lecturers

For Person Responsible for timetable scheduling at the school

12. How is the school of MIA timetables scheduled?
13. What factors do you take into account when scheduling the timetable
14. How are the times and arrangement of courses decided upon?
15. Are students and lecturers involved in the scheduling of these documents
Probe: If not, why has this not being the case
16. Were there any specific improvements to be incorporated but could not due to certain
constraints
Probe: If there were, what were the constraints
17.During this time of online learning how has timetable scheduling changed
Probe: What are the new constraints and advantages
18.What are the ideal times for activities to be worked on in the timetable?
19.What are the ideal number of courses you think could be included in a day?

78
C. Interview Protocol
Introduction:
Introduce myself to the interviewee
Brief them about the research we are conducting
Spend 5 minutes for this introduction and for explaining the ground rules below.
Ground Rules:
Inform the interviewees that the session is being recorded but assure that it will not be shared with
anyone but the supervisor.
Interviewees can stop the session anytime if they are starting to feel uncomfortable
Interviewees and interviewer can stop each other at any time during the session, if one needs
clarification or has a question on certain topics of the interview
Interviewee can ask for more time to answer any questions they find difficult
Questions and Probes:
Start with the demographic section
Move to their current experience of their academic schedules and preference section
Finish with the Future Schedules section
Start with the more general, grand tour and easy to answer questions and move to the more focussed,
structured and specific questions
Prioritise the more primary or most important questions, then after these move to the more secondary
or least important questions
Accommodate any questions that the interviewees ask during the session.
Spend 5 minutes on the demographic and 20 minutes on the academic schedules, preference, Future
Schedules sections
End:
Show appreciation to the interviewee on their participation and ask if they would like to receive the
analysed report on their inputs, and exchange any important contact details.

79
D. Coding Tree

1. Attribute Codes

1.1 Year of Study lecturing

1.1.1. Second Year

1.1.2. Third year

1.1.3. Fourth Year

1.1.4. Both Second and Third year

1.1.5 Both Third and Fourth year

1.1.6 Both Second and Fourth

1.1.7. All years of study

1.2. Courses Lecturing

1.2.1. One

1.2.2. two

1.2.3. three

1.2.4 More than three

1.3. Name of courses lecturing

1.3.1. Any from table [2]

1.4. Other Activities in the school

1. 4.1. Research

1.4.2. Admin Role

1.4.3. School Committees

1.4.4. Both 4.1 and 4.2

1.4.5. Both 4.2 and 4.3

1.4.6. Both 4.1 and 4.3

1.4.7. All of the Above

1.4.8. Other

80
2. Substantive Codes

2.1. Scheduling system from a lecturer’s perspective

2.1.1. Really good, easy to work with

2.1.2. Good enough, sufficient time given

2.1.3. It is ok, just enough time given, manageable

2.1.4. Not good, there time is not sufficient hard to follow

2.1.5. Really Bad, it’s really complex and not well arranged, not enough time for lectures.

2.2. Consultation Sessions

2.2.1. Yes

2.2.2. No

2.3. How they are Facilitated

2.3.1. Students book an appointment

2.3.2. After every lecture

2.3.3. Fixed timeslot

2.3.4. Meet anytime, anywhere

2.3.5. Both 2.3.1 and 2.3.2

2.3.6. Both 2.3.2 and 2.3.3

2.3.7. Both 2.3.1 and 2.3.3.

2.4. Is the time for Consultations exhaustively used?

2.4.1. Yes, Very much so

2.4.2. Yes, Sometimes

2.4.3. No

2.5. Constraints/factors that affect lecture schedule

2.6. How do these constraints affect the lecture schedule?

2.6.1. Positively, they work in my favour

2.6.2. Do not affect me much, I still manage

81
 2.6.3. Negatively, cannot do my job effectively

2.7. Constraints that affect scheduling of tests and assignments

2.8. How do these constraints affect the test and assignment schedule?

2.6.1. Positively, they work in my favour

2.6.2. Do not affect schedules much

2.6.3. Negatively, cannot schedule these assessments effectively

2.9. Number of students ideal for you to handle

2.9.1. As big as possible, more than 50 students

2.9.2. A small as possible, less than 50 students

2.10. Working Hours ideal in a day

2.10.1. More than 7/8 hours

2.10.1. Normal work hours, 7-8 hours

2.10.2. less than 7 hours

2.11. Suggestion on improving the timetable.

For Person responsible for Scheduling

2.12. How the school timetable is scheduled.

2.13. Factors considered in scheduling

2.14. How times and arrangements of courses are decided upon.

2.15. Are students involved in the scheduling process?

2.15.1. Yes

2.15.2. Sometimes/ somewhat

2.15.3 No

2.16. Has timetable scheduling changed during online learning?

2.16.1. It has completely changed

2.16.2. It has somewhat changed, not by much though

2.16.3. It has not changed at all

82
2.17. If it has changed, how has it changed?

2.18. Ideal times for times for activities in the time timetables

2.18.1. Morning

2.18.2. Afternoon

2.18.3. Other

2.19. Ideal time duration for timetable e.g. lectures

2.19.1. More than the conventional 45 minutes

2.19.2 The conventional 45 minutes

2.19.3. Less than 45 minutes

83
E. Simio Timetable models

Figure 50: Mechanical Engineering Model

84
Figure 51: Industrial Engineering Model

85
Figure 52: Aeronautical Engineering Model

86
Figure 53: Tests timetable Model

87
F. Clashes in Timetables

Table 14: Clashes in the timetables

Mechanical Stream Clashes

Timeslot 8:00-9:45 AM 10:15-12:00pm 12:30-13:15 pm 14:15-16:00pm
Monday 0 0 2 0
Tuesday 1 1 1 1
Wednesday 1 0 1 0
Thursday 1 2 0 0
Friday 1 2 2 0

Industrial Stream Clashes

Monday 0 0 2 2
Tuesday 3 2 1 1
Wednesday 1 3 1 1
Thursday 1 3 1 3
Friday 0 3 3 2

Aeronautical Stream Clashes

Monday 0 0 3 0
Tuesday 1 1 1 2
Wednesday 1 0 1 0
Thursday 2 2 0 0
Friday 0 1 2 0

Tests Timetable
Date Clashes
30 March 2021 0
1 April 2021 0
6 April 2021 5
12 April 2021 2
15 April 2021 0
19 April 2021 1
3 May 2021 1
10 May 2021 2
31 May 2021 0

88
G. F tables

Table 15: F tables

89
90
H. t-tables

Table 16: T-tables

91
I. Ethics Certificate

92
J. Registrar Approval Letter

93
K. Margin of error calculations

Figure 54: Margin of error calculation for students’ responses

94
Figure 55: Margin of error calculation for lecturers’ responses

95