THE USE OF OUTLIERS IN THE DETECTION OF SUSPICIOUS EXAMINATION MALPRACTICES IN THE SCORES OF STUDENTS IN A NIGERIAN UNIVERSITY

Sep 14, 2026

Uploaded by Gloria

Detecting outliers helps identify observations that deviate from the dominant pattern in student test or examination score datasets. The study applies both graphical and analytical outlier detection methods to flag points that may reflect either legitimate but unexpected results or suspicious examination malpractice. Outliers can distort regression-based statistical analyses by producing erroneous parameter estimates and misleading inferences. The work examines masking and swamping effects and recommends further investigation of flagged data points with appropriate anti-malpractice measures.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages10

Reading Time25 min

AccessFree to download

Abstract

Introduction

Definition and implications of outliers

Outliers and regression influence

Examination malpractice and contributing factors

Data Collection

Study population and sampling

Disciplinary committee context

Detection Outliers in Simple Linear Regression

THE USE OF OUTLIERS IN THE DETECTION OF SUSPICIOUS EXAMINATION MALPRACTICES IN THE SCORES OF STUDENTS IN A NIGERIAN UNIVERSITY Consul, Juliana Iworikumo ([iwori2001@yahoo.com](iwori2001@yahoo.com)) and Ndiwari, Godgift ([gifttola@yahoo.com](gifttola@yahoo.com)) Department of Mathematics and Computer Science, Niger Delta University, Bayelsa State, Nigeria ABSTRACT It is very important that we detect outliers as it deviates from the majority of the data. In our study, we have used both graphical and analytical methods to detect outliers which may be mistakes or else accurate but unexpected suspicious examination or test scores. These observations may be due to malpractices. In cases where we might want to perform statistical analysis like simple linear regression models, outliers may lead to erroneous parameter estimate and inferences from the model. We recommend that suspicious data points should be further investigated and in the case of examination malpractices, appropriate measures should betaken. The masking and swamping effect of outliers appeared in this study. Keywords: Outlier, influential, examination malpractice, residual, masking and swamping effect. 1.0 INTRODUCTION An outlier in a set of data is defined to be an observation or a subset of observations which appear tobe inconsistent with the remainder of the set of data. Outliers may not be genuine members of the main population (Barnett and Lewis (1994) and Cook (1977)) . Outliers can also be defined as observations which are found to differ so much from other observations. These observations are often seen as contaminating the data. They either reduce or distort the information about the source of the data. We might want to seek means of interpreting, categorising or handling outliers in other to reduce the effect on statistical analysis if required. Outliers may frustrate attempts to draw inferences about the basic population. Barnett and Lewis (1994) showed that extreme observations and contaminants may or may not be outliers. Outliers may arise from any of measurement, recording, execution or calculating error in the data. Barnett and Lewis (1994) gave an example of which a recording error may appear as an outlier in a set of data. Outliers can be handled either by being retained or rejected. A very important part of statistical practice is considering the study of regression models. Suppose that we have a linear regression model ݕ = ߚ଴+ ߚଵx + ∈ . We have that the expectation of ∈ , E ( ∈)= 0 and ܸܽݎ ( ∈) = ߪ ଶ which is a constant. An observation that seems to deviate so much from the linear model than other observations is said to be an outlier. The outlier breaks the expected pattern of results for the regression model. One very interesting effect of an outlier is its influence on the estimation of the parameters in the regression model. Barnett and Lewis (1994) records that outlier inflates the estimate of ߪ ଶ . They also showed that robust methods of inference can be employed to the data to minimise the influence of outliers. Onyibe et al. (2015) mentioned that Education is a process of teaching and learning which is evaluated through examination at the end of the learning period. It serves as feedback for the examiners as to ascertain the level of knowledge acquisition of the student. This knowledge is meant to be retained by the student. Examination is part of evaluation in an educational system that is aimed at determining a students’level of intellectual competence or understanding after learning. Examination malpractice can therefore be defined as a deliberate wrong doing, misconduct or improper practice during or after examination as contrary to official examination rules in view to obtaining good results by fraudulent means (Onyibe et al. (2015)) . The causes of examination malpractice may be due to inadequate teaching or learning facilities, an intension of someone to help a friend, poor parental upbringing, ineff

FAQs

What is the main purpose of using outliers in this study?

To detect suspicious examination malpractice by identifying score observations that deviate from the majority pattern, using graphical and analytical outlier detection approaches.

How do outliers affect statistical analysis such as linear regression?

Outliers can cause erroneous parameter estimates and incorrect inferences in simple linear regression by breaking the expected model pattern and influencing estimation of parameters.

What effects of outliers are discussed in the study?

The masking and swamping effects of outliers are reported as relevant to how outliers may influence detection and interpretation.

Sep 14, 2026

Uploaded by Gloria