Design and Implementation of a Time Series Analysis System for Historical Students' Graduation Results

1.1 Introduction

Time series analysis is the process of examining data collected at successive points in time to identify patterns, trends, and changes and to use these patterns to estimate future outcomes. A time series may consist of annual, quarterly, monthly, or other regularly recorded observations. In forecasting, historical data provide the basis for understanding how a variable has behaved and estimating how it may behave in the future (Hyndman & Athanasopoulos, 2021). In an educational institution, students' graduation results recorded across different academic sessions form a time series that can be analyzed to determine changes in the number of graduating students and other relevant graduation patterns.

The analysis of historical students' graduation results is important because educational institutions generate large amounts of academic records over the years. When these records are properly organized and analyzed, they provide useful information for academic planning, resource allocation, monitoring of student outcomes, and forecasting. Time series methods help to identify whether graduation outcomes are increasing, decreasing, or following a particular pattern over time. Such analysis is also applicable to educational data, where historical academic records may be examined to support prediction and informed decision-making (Sinharay, 2010).

As a prelude to other parts of this study, this chapter will discuss the background upon which this study was initiated, the statement of problems that led to this study, the aim and objectives of the study. Others are significance of the study, scope of work, limitation of the study and definition of technical terms.

1.2 Background of Study

Time series analysis has a long history in statistics and developed from the need to understand observations recorded over time. The formal development of time series analysis according to Mills (2011) can be traced to the late nineteenth century, when researchers began developing statistical methods for studying patterns in observations that changed over time. Early applications were concerned with areas such as economics, astronomy, physical sciences, and population studies. Over time, researchers developed methods for identifying trends, cycles, seasonal movements, and relationships between observations. The development of probability theory and statistical modelling later provided stronger foundations for analysing time-dependent data.

The development of modern time series forecasting became more significant during the twentieth century as researchers sought more reliable ways of predicting future events from historical observations. As cited by Chatfield (2005), forecasting methods developed from relatively simple techniques such as moving averages and exponential smoothing to more advanced statistical models. Similarly, Box and Jenkins made a major contribution through their work on autoregressive integrated moving average models. Their influential book, Time Series Analysis: Forecasting and Control, first published in 1970, provided a systematic framework for identifying, estimating, and checking time series models (Ljung et al., 2014).

Time series analysis refers to the systematic study of observations collected over time to identify patterns, trends, seasonal movements, and changes that may assist in forecasting future outcomes. Hyndman and Athanasopoulos (2021) stated that forecasting involves using available historical information to estimate future values, while proper understanding of the characteristics and patterns within the data is necessary before selecting a forecasting method. In the educational sector, historical records such as students' enrolment, academic performance, completion rates, and graduation results provide valuable information for institutional planning. Similarly, Box et al. (2015) reported that time series methods provide statistical approaches for understanding observations that occur sequentially over time.

Higher education institutions generate large volumes of student records through academic activities carried out over several years. These records contain information that may reveal changes in student performance and graduation outcomes when properly examined. Romero and Ventura (2020) stated that educational data mining provides techniques for discovering useful patterns and knowledge from educational data. These approaches show that academic records are not merely administrative documents but also sources of information that may support institutional planning. Furthermore, graduation results recorded over successive academic sessions may provide a basis for determining whether the number of students graduating is increasing, declining, or changing irregularly.

The use of historical academic data for prediction has received increasing attention because institutions need timely information to support student management and academic planning. As described by Livieris et al. (2019), educational data mining has become important for predicting students' progress and graduation time, particularly in identifying students who may experience difficulties completing their studies. Their study demonstrated that predictive techniques may be used to classify students according to their expected graduation time. Likewise, Co and Casillano (2022) reported that student academic performance could be used to predict whether students would graduate on time. These findings indicate that academic data has predictive value when suitable analytical techniques are applied.

In Nigeria, the application of data-driven methods to students' academic outcomes is also relevant because higher institutions manage large numbers of students across different programmes and academic sessions. According to Adebayo et al. (2019), predictive analysis of students' academic performance provides an opportunity to identify patterns that may be associated with eventual graduation outcomes. Their study involving engineering students in a Nigerian university found that students' earlier academic performance, programme of study, and year of entry could be used to predict final graduation performance. Supporting this position, Bako et al. (2023) examined timely graduation prediction among postgraduate students in a Nigerian university and affirmed that predictive models could support the identification of students based on expected graduation outcomes.

The proposed study is therefore concerned with the design and implementation of a time series analysis system capable of examining historical students' graduation results at IMT Enugu. This study is set against the backdrop of the growing need for higher institutions to use historical academic data, computerized information systems, and time-based analytical techniques to know graduation trends and support informed decisions about future students' graduation outcomes.

1.3 Statement of Problems

The existing system for managing and analysing historical students' graduation results at IMT Enugu has several problems that make effective analysis and forecasting difficult. The major problems include:

1. The traditional system relies heavily on manual examination of historical graduation records, which makes the analysis of large amounts of data slow and stressful.
2. The existing system does not provide a dedicated platform for identifying graduation trends across different academic sessions.
3. The old existing system makes it difficult to quickly compare graduation results from different years, departments, or programmes.
4. The present system does not provide an effective forecasting feature for estimating future graduation outcomes from historical results.
5. The previously used system makes retrieval and review of historical graduation information more difficult when records are stored across different files or documents.
6. Lastly, system does not adequately transform historical graduation records into useful information for academic planning and decision-making.

1.4 Aim and Objectives of the Study

The aim of this study is to design and implement a time series analysis system for historical students' graduation results at the Institute of Management and Technology (IMT), Enugu. In achieving this aim, the following specific objectives were laid out as follows:

1. To examine the existing method of managing and analysing historical students' graduation results at IMT Enugu.
2. To develop a time series analysis system for identifying trends in students' graduation results.
3. To design a database for storing and managing historical students' graduation results.
4. To implement a forecasting feature for estimating future graduation outcomes from historical data.
5. To evaluate the performance and usability of the developed system for analysing students' graduation results.

1.5 Significance of Study

The deployment of the proposed system will hold significant relevance in the following ways:

1. The proposed system will provide management with organized graduation trends and forecast information for academic planning.
2. The new system will make it easier to review graduation results across different academic sessions and identify changes in student outcomes.
3. The system will simplify the storage, retrieval, and processing of historical graduation information.
4. The project will provide a practical example of applying time series analysis and computer-based data processing to educational records.
5. The developed system will serve as a useful reference for developing related educational data analysis and forecasting systems.

1.6 Scope of Study

This study focuses on the design and implementation of a time series analysis system for historical students' graduation results at the Institute of Management and Technology (IMT), Enugu State, Nigeria. The study covers the collection and storage of historical graduation records, analysis of graduation trends across academic sessions, graphical presentation of results, and forecasting of possible future graduation outcomes.

1.7 Limitations of the Study

During the course of this study, many things militated against its completion, some of which are:

1. Financial Constraints: Limited funds affected transportation, data collection, printing, internet access, and other research expenses.
2. Time Constraints: The period available for the study was limited, which affected the amount of time available for collecting, validating, developing, and testing the system.

1.8 Definition of Terms

Time Series:

A sequence of observations recorded at different points in time. In this study, it refers to graduation results recorded across different academic sessions and used to identify patterns and trends (Hyndman & Athanasopoulos, 2021).

Time Series Analysis:

The process of examining data collected over time to identify patterns, trends, and changes that may assist in understanding or forecasting future values (Box et al., 2015).

Graduation Results:

The academic records showing students who successfully completed their programmes and graduated during particular academic sessions.

Historical Data:

Previously recorded information collected from past academic periods. In this study, it refers mainly to students' graduation records from previous academic sessions.

Forecasting:

Forecasting is the process of using historical information and an appropriate analytical method to estimate possible future values (Hyndman & Athanasopoulos, 2021).

Trend:

Trend is the general direction in which a set of observations moves over time, such as an increase or decrease in the number of graduating students.

Academic Session:

It is a defined period during which academic activities are carried out in an educational institution and students' academic records are maintained.

Graduation Rate:

Graduation rate is the proportion or number of students who successfully complete their programme within a specified period.

Historical Graduation Result:

Historical graduation result is a recorded graduation outcome from a previous academic session that is used as input for analysis and forecasting.

Database:

Database is an organized collection of related data that allows information to be stored, retrieved, and managed efficiently (Laudon & Laudon, 2020).

System:

System is a combination of related components that work together to collect, process, store, and produce useful information.

Information System:

A computerized arrangement used to collect, process, store, and provide information for organizational activities and decision-making (Stair & Reynolds, 2018).