ILJTER.ORG Volume 22 Number 06 June 2023

International Journal
of
Learning, Teaching
And
Educational Research
p-ISSN:
1694-2493
e-ISSN:
1694-2116
IJLTER.ORG
Vol.22 No.6

International Journal of Learning, Teaching and Educational Research
(IJLTER)
Vol. 22, No. 6 (June 2023)
Print version: 1694-2493
Online version: 1694-2116
IJLTER
International Journal of Learning, Teaching and Educational Research (IJLTER)
Vol. 22, No. 6
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically those of translation, reprinting, re-use of illustrations,
broadcasting, reproduction by photocopying machines or similar means, and storage in data banks.
Society for Research and Knowledge Management

The International Journal of Learning, Teaching and Educational
Research is a peer-reviewed open-access journal which has been
established for the dissemination of state-of-the-art knowledge in the
fields of learning, teaching and educational research.
Aims and Objectives
The main objective of this journal is to provide a platform for educators,
teachers, trainers, academicians, scientists and researchers from over the
world to present the results of their research activities in the following
fields: innovative methodologies in learning, teaching and assessment;
multimedia in digital learning; e-learning; m-learning; e-education;
knowledge management; infrastructure support for online learning;
virtual learning environments; open education; ICT and education;
digital classrooms; blended learning; social networks and education; e-
tutoring: learning management systems; educational portals, classroom
management issues, educational case studies, etc.
Indexing and Abstracting
Research is indexed in Scopus since 2018. The Journal is also indexed in
Google Scholar and CNKI. All articles published in IJLTER are assigned
a unique DOI number.

Foreword
We are very happy to publish this issue of the International Journal of
Learning, Teaching and Educational Research.
Research is a peer-reviewed open-access journal committed to
publishing high-quality articles in the field of education. Submissions
may include full-length articles, case studies and innovative solutions to
problems faced by students, educators and directors of educational
organisations. To learn more about this journal, please visit the website
http://www.ijlter.org.
We are grateful to the editor-in-chief, members of the Editorial Board
and the reviewers for accepting only high quality articles in this issue.
We seize this opportunity to thank them for their great collaboration.
The Editorial Board is composed of renowned people from across the
world. Each paper is reviewed by at least two blind reviewers.
We will endeavour to ensure the reputation and quality of this journal
with this issue.
Editors of the June 2023 Issue

VOLUME 22 NUMBER 6 June 2023
Table of Contents
Generative AI Chatbots - ChatGPT versus YouChat versus Chatsonic: Use Cases of Selected Areas of Applied
English Language Studies......................................................................................................................................................1
Chaka Chaka
Influence of Teacher-Targeted Bullying Behaviour on Teachers in Selected Schools in South Africa ...................... 20
Japsile Sambo, Sumeshni Govender
Validating and Testing the Teacher Self-Efficacy (TSE) Scale in Drug Education among Secondary School
Teachers ................................................................................................................................................................................. 45
Ciptro Handrianto, Ahmad Jazimin Jusoh, Nazre Abdul Rashid, Muh Khairul Wajedi Imami, Suzaily Wahab, M Arinal
Rahman, Ary Kiswanto Kenedi
Voices from the Field: Pre-Service Teachers’ First Time Experiences of Teaching Physical Sciences during School-
Based Experience ..................................................................................................................................................................59
Sakyiwaa Boateng, Benjamin Tatira
The Development of an Online Study Adaptability Scale for Chinese College Students During the Global COVID-
19 Pandemic........................................................................................................................................................................... 78
Guo Jun Tan, Jia Qi Wei, Chia Ching Tu
The Influence of Information and Communication Technology in the Teaching and Learning of Physics ............. 98
Abdussamad Ringim Hussaini, Shehu Ibrahim, Kingsley Eghonghon Ukhurebor, Grace Jokthan, Juliana Ngozi Ndunagu,
Adeyinka Oluwabusayo Abiodun, Fwa Enoch Leonard, Benjamin Maxwell Eneche, Dorothy Nalwadda
Examining Philology Teachers’ Lesson Planning Competencies in Vietnam............................................................. 121
Hien Thu Thi Pham, Nguyet Minh Thi Le, Huyen Thanh Thi Doan, Hien Thi Luong
Promoting Critical Thinking through Socratic Questions in Health Sciences Work-Integrated Learning ............. 137
Zijing Hu
Mobile Learning in Higher Education: Insights from a Bibliometric Analysis of the Body of Knowledge............ 152
Godwin Kaisara, Kelvin Joseph Bwalya
Entrepreneurship Education: Encouraging Entrepreneurial Intentions for Equality Education Students in
Semarang ............................................................................................................................................................................. 175
Imam Shofwan, Sunardi Sunardi, Gunarhadi Gunarhadi, Abdul Rahman
English Language Skills and Becoming a Global Entrepreneur: Lessons for Entrepreneurship Education........... 195
Ismail Sheikh Ahmad, Zarinah Jan Yusof Khan
Motivational Factors that Influence the Course Completion Rate of Massive Open Online Courses in South Africa
............................................................................................................................................................................................... 212
Liezel Cilliers, Hossana Twinomurinzi, Obrain Murire

The Relationship Between Academic Self-Efficacy and Undergraduate Students’ Perceptions of Electronic
Assessment: A Mediation Analysis .................................................................................................................................. 226
Ahmed M. Asfahani
Problem-based Learning (PBL) with Reading Questioning and Answering (RQA) of Preservice Elementary
School Teachers................................................................................................................................................................... 245
Marleny Leasa, Abednego Abednego, John Rafafy Batlolona
Using the ADDIE Model to Teach Creativity in the Synthesis of Raw Materials ...................................................... 262
Hussein Ahmed Shahat, Sherif Adel Gaber, Hussam Khalifah Aldawsari
A Systematic Review of the Practicum Experience in Preservice Teacher Education During the COVID-19
Pandemic.............................................................................................................................................................................. 282
Taghreed Abdulaziz Almuqayteeb, Dalal Alzahrani
Gamification in Engineering Education during COVID-19: A Systematic Review on Design Considerations and
Success Factors in its Implementation.............................................................................................................................. 301
Omar Chamorro-Atalaya, Guillermo Morales-Romero, Nicéforo Trinidad-Loli, Beatriz Caycho-Salas, Teresa Guía-
Altamirano, Elizabeth Auqui-Ramos, Yadit Rocca-Carvajal, Maritza Arones, José Antonio Arévalo-Tuesta, Roxana
Gonzales-Huaytahuilca
The Role of Metacognitive Strategies in Academic Writing Skills in Higher Education........................................... 328
Lilis Amaliah Rosdiana, Vismaia S. Damaianti, Yeti Mulyati, Andoyo Sastromiharjo
Parental Occupation, Social Class, and School Choice in Southern Philippines: Their Implications to Educational
Public-Private Partnership vis-à-vis the K-12 SHS Voucher Program......................................................................... 345
Fernigil L. Colicol, Fauzia K. Sali-Latif
Learning Moral Values Through Cartoons for Malaysian Preschool-aged Children................................................ 370
Muhammad Alif Redzuan Abdullah
An Investigation into Communication between Teachers and Parents of Students with Autism Spectrum
Disorder ............................................................................................................................................................................... 395
Abdulaziz Hamad Al-Hamad, Sherif Adel Gaber, Sayed Ibrahim Ali
The Effects of an MMORPG on Thai EFL University Students’ Reading for Main Ideas ......................................... 415
Wanwisa Changkwian, Suksan Suppasetseree
Developing Elementary School Teacher’s Professional Competence in Composing Traditional Songs: An Action
Research in Indonesia......................................................................................................................................................... 435
J. Julia, Tedi Supriyadi, Enjang Yusup Ali, Egi Agustian, Afi Fadlilah
Emotional Competency in Teaching: A Qualitative Study of Practices among Preschool and Elementary School
Teachers ............................................................................................................................................................................... 459
Laila Ouchen, Lahcen Tifroute, Khadija El Hariri
Piloting Supplementary Materials Aimed at Developing Students’ Problem-Solving and Self-Regulated Learning
Skills...................................................................................................................................................................................... 475
Liena Hacatrjana, Inga Linde
Axiological Study of Educational Projects in Schools.................................................................................................... 494
Rodrigo Arellano Saavedra, Andrew Philominraj, Ranjeeva Ranjan, Claudio Andrés Ceron Urzua
The Impact of Combination of Natural Sciences and the Humanities on the Quality of Modern Education ........ 515
Kateryna Kyrylenko, Mykhailo Martyniuk, Tetiana Mahomet, Volodymyr Mykolaiko, Iryna Tiahai, Olesia Beniuk

The Relationship between Malaysian Students’ Socio-Economic Status and their Academic Achievement in STEM
education.............................................................................................................................................................................. 533
Saras Krishnan, Enriqueta Reston, Sheila Devi Sukumaran
Value of clinical observational learning in work-integrated learning in health sciences education: Students’ views
and experiences................................................................................................................................................................... 479
Darren Carpenter, Zijing Hu
Representation of National Identity and Culture in the Saudi EFL Textbook Series Mega Goal: A Critical
Discourse Analysis ............................................................................................................................................................. 568
Ali Abbas Falah Alzubi, Khaled Nasser Ali Al-Mwzaiji, Mohd Nazim
Assessing the Effectiveness of Computer-Aided Instructional Techniques in Enhancing Students’ 3D Geometry
Spatial Visualization Skills Among Secondary School Students in Tanzania............................................................. 613
Marcellina Andrea Mjenda, Vedaste Mutarutinya, Owiti Dickson
Demo Lessons and Peer Observation to Enhance Student Teachers’ Competencies and Exit Profiles................... 638
Agnes Orosz, Uvaldo Recino, Maria Caridad Ochoa
Gender Equality in Science Classrooms: Examining the Implementation of Gender-responsive Approach and its
Impact on Science Education............................................................................................................................................. 659
Peter Paul Canuto, Felina Espique
Teacher Recruitment and the Right Career Choice: Parents’ Perceptions of the Teaching Profession in Oman ... 679
Khalaf M. Al‘ Abri, Omer H. Ismail, Aieman A. Al-Omari
Role of Executive Functions in Improving Students' Narrative Text Writing Ability............................................... 694
Rastya Mutiarani Zahra, Sumiyadi Sumiyadi, Isah Cahyani, Andoyo Sastromiharjo

1
©Authors
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0
International License (CC BY-NC-ND 4.0).
Vol. 22, No. 6, pp. 1-19, June 2023
https://doi.org/10.26803/ijlter.22.6.1
Received Mar 24, 2023; Revised May 16, 2023; Accepted May 22, 2023
Generative AI Chatbots - ChatGPT versus
YouChat versus Chatsonic: Use Cases of Selected
Areas of Applied English Language Studies
Chaka Chaka
University of South Africa, Pretoria, South Africa
Abstract. This paper reports on the comparison of the accuracy and
quality of the responses produced by the three artificial intelligence (AI)
chatbots, ChatGPT, YouChat, and Chatsonic, based on the prompts (use
cases) related to selected areas of applied English language studies
(AELS). An exploratory research design was employed and we utilised
purposive sampling. The three aforementioned AI chatbots were used to
collect data sets. Of the three chatbots, YouChat was technically unstable
and unreliable, and had some inconsistency in generating responses. The
other two chatbots, ChatGPT and Chatsonic, consistently exhibited a
tendency to plagiarise responses from internet information without
acknowledging the sources. In certain cases, the three chatbots all
generated almost similar responses for different and unrelated prompts.
This made their responses look like run-of-the-mill responses that lacked
credibility, accuracy, and quality. One chatbot (ChatGPT) could not
recognise a scholar mentioned in one of the prompts, while the other one
(Chatsonic) misrecognised this scholar, and ended up rambling parts of
its response. Additionally, the three chatbots all mechanically and
superficially generated phrases and ideas in their responses without
detecting the related critical nuances in the original sources in which they
were used. This made the knowledge communicated by those responses
appearing too fluffy. In this paper the educational and knowledge
implications of the generated responses for AELS were educed. Based on
the shortcomings the three AI chatbots displayed, I concluded that these
three chatbots are not yet credible and reliable generators of knowledge
for the aspects of AELS discussed.
Keywords: ChatGPT; YouChat; Chatsonic; AI chatbots; responses;
applied English language studies
1. Introduction
Shortly after ChatGPT, an OpenAI-owned artificial intelligence (AI) powered,
generative chatbot, was launched in late November 2022, followed by an AI and
Internet search engine race, similar, maybe, to the space race. Of course, prior to

2
http://ijlter.org/index.php/ijlter
this, AI has been making its presence visible in, for example, areas such as drones,
self-driving cars, mobile phones, and robotic personal assistants (Chaka, 2020). If
so, then, what has changed now? The sudden and almost unannounced arrival of
ChatGPT seems to have changed and rattled the generative AI world. Existing
Internet search engines such as Bing and Google instantly started incorporating
AI chatbots like Bing AI and Bard AI into their search engine ecosystems,
respectively. Bard AI seems to be Google’s answer to, or its intended killer of
ChatGPT and Bing AI (Eliaçık, 2023a; Kamran, 2023; Knight, 2023; ul Haq, 2023).
Similarly, other new AI chatbots suddenly emerged. These include Caktus,
Chatsonic, Chinchilla, Jasper Chat, Perplexity, and YouChat. At the moment, the
general view is that ChatGPT rules the roost on a first come, first served basis
(Eliaçık, 2023b), even though the final determinant of the ultimate ruler will be the
best large language model each of these AI-powered chatbots will have when
compared to one another. On this score alone, Chinchilla, which is still in its beta
stage, is likely to be the winner as it has 1.4 trillion training tokens vis-à-vis
ChatGPT’s 300 billion training tokens (Eliaçık, 2023c). These are staggering and
intimidating numbers. But in the realm of generative AI spheres, an entity is not
a winner until it has won the contest. Be that as it may, a human brain comprises,
as Nawrocki (2011) points out, about 100 billion neurons and almost 1,000 trillion
synaptic connections organised into many and varied areas that perform different
brain functions, which include, among others, complex cognitive functions
(Ackerman, 1992; Atallah et al., 2004; Deacon, 1997; cf. Adesso, 2023). While this
analogy is way too far-fetched as it does not represent a like-like comparison, it,
nonetheless, brings home the vast difference between the capacity (and the depth
and breadth) of the current generative AI chatbots and that of a human brain.
The emerging impact of AI chatbots is felt in various spheres of human lives and
in different sectors of life. This is because, by their very nature, these chatbots
serve multiple purposes in each sector. In the education sector, these purposes are
many and varied. For example, they can operate as online search engines, respond
to written prompts, write essays on topics (Anders, 2023; Kumar, 2023; Pittalwala,
2023), summarise and translate text, and correct grammar errors (Eliaçık, 2023c;
SGA Knowledge Team, 2023). They can also define concepts/terms, remix, edit
and improve writing, and generate lesson plans (Cutcliffe, 2022; Harris, 2022).
Moreover, they can offer advice on conducting research in the digital age, create
a structure for a research proposal, offer advice about given aspects of a research
proposal, and provide sources of citations (at least some of them) (Chaka, 2023a).
Given the multiple purposes that AI chatbots can serve within the education
sector as outlined above, assertions have been made that these chatbots are a big
deal for education (Anders, 2023), may challenge disciplinary specialisation (they
can generate responses across a range of academic disciplinary boundaries
[Chaka, 2023b]), could be game-changers (Harris, 2022) and disruptors (Fraser,
2023), possess essay-writing skills that can stun teachers (Bowman, 2023; Hern,
2022), and can do homework for students (Pittalwala, 2023). Taking into account
the multiple functions the AI chatbots can perform and considering the foregoing
assertions made about them within the education sector, it appears that these

3
chatbots can do almost anything. Therefore, there are concerns that AI chatbots
will churn out plagiarised information (Dilmegani, 2023; Pittalwala, 2023),
generate responses containing factual inaccuracies, and invent fictitious names
(Browne, 2023), hallucinate about things (ul Haq, 2023), waffle facts and
misattribute work (Ceres, 2023). These may be used by students in their academic
tasks without them noticing all these drawbacks.
However, regardless of the afore-mentioned issues, some suggested positive
educational applications of ChatGPT exist. Dilmegani (2023), for example,
suggests the following for teachers and for students:
• For teachers: Content creation; grammar and writing corrections (e.g., proof-
reading and editing, offering student feedback, and writing-skills teaching);
grading; designing course outlines (e.g., course goals and objectives,
generating course topics, lesson plans, and locating and identifying course
materials and resources).
• For students: Assisting with homework (e.g., answering questions,
reinforcing concepts, improving writing skills, and solving problems);
research (e.g., selecting a topic, topic background information, locating and
identifying suitable resources, organising research, and locating citations or
sources of reference); and learning language (Dilmegani, 2023; cf. Chaka,
2023a).
A case that employed a different AI chatbot is the one used by Wiggers (2023).
The used case entailed generating samples of writing covering diverse genres.
These diverse writing genres (applications) were:
• An application letter for a paralegal position
• A curriculum vitae for a software engineer
• An email message to market shoe polish
• An online news article covering the 2020 U.S. presidential elections
• An essay outline focusing on the merits of gun control (in the U.S.)
• A college-level essay on the fall of Rome
• An encyclopaedia entry for Mesoamerica (Wiggers, 2023).
Taking cognisance of the diverse applications of chatbots, the current paper
argues that, thus far, there have not yet been enough documented-use cases of AI
chatbots that focus on given academic disciplines in the higher education (HE)
sector, particularly, on the specific aspects of such academic disciplines. Mostly,
AI chatbots have been used to respond to generic prompts that are not related to
specific academic disciplines. Even in instances where they have been used to
respond to generic prompts as in Wiggers’s (2023) case, they have not been
employed in a sustained and robust manner to interrogate the types of responses
they generate in respect of specific aspects of given academic disciplines in the HE
sector. Against this background, in this paper an attempt is made to fill the gap
that has not yet been explored by using three generative AI chatbots, namely
ChatGPT, YouChat, and Chatsonic, to generate responses related to selected areas
of applied English language studies (AELS). The aim was to compare the accuracy

4
and quality of the responses these three AI chatbots generated about the selected
areas of AELS as informed by the specific prompts provided as input.
In this regard, in this paper it was strived to find answers to the following two
research questions (RQs):
• How accurate are the responses generated by ChatGPT, YouChat, and
Chatsonic to selected areas of applied English language studies such as
decolonial applied linguistics, critical southern decoloniality, and
translanguaging, multilanguaging, and languaging as based on the prompts
inputted to them?
• What is the quality of the responses of these three AI chatbots?
2. ChatGPT, YouChat, and Chatsonic: A Brief Overview
As, at the time of writing this paper, there were not yet scholarly papers published
on the use cases of AI chatbots in AELS in the HE sector, as highlighted above, the
paper rather offers a brief overview of the three AI chatbots it employed for its
use case.
2.1 ChatGPT
Since its release in late November 2022, ChatGPT has had several comments,
reports, descriptive analyses, and reviews (Bowman, 2023; Cutcliffe, 2022; Harris,
2022; Hern, 2022; Meghmala, 2023; Ofgang, 2023; Solé, 2023). At the time of
writing this paper, the number of such comments, reports, descriptive analyses,
and reviews was increasing exponentially. ChatGPT is an AI chatbot, whose
parent company is OpenAI. On its website, OpenAI says that it is “an AI research
and deployment company”, whose mission “is to ensure that artificial intelligence
benefits all of humanity”. It defines artificial general intelligence as “AI systems
that are generally smarter than humans” (OpenAI, 2015-2023). The GPT in
ChatGPT stands for Generative Pre-trained Transformer. As an AI-powered
chatbot, ChatGPT is one of the new-generation AIs that employ large language
models (LLMs). As Eliaçık (2023c) points out, LLMs utilise deep learning, which
relies on multi-layered neural networks for collecting, processing, and analysing
complex datasets with a view to making predictions and generating natural
language responses (OpenAI, 2022; SGA Knowledge Team, 2023; Stiennon et al.,
2020).
Moreover, as a third generation of the Generative Pre-trained Transformer (GPT-
3) based chatbot, ChatGPT also utilises an autoregressive language model that
helps it to generate text that cannot be distinguished from human-written text
(Eliaçık, 2023c). Importantly, ChatGPT has a reinforcement learning from human
feedback (RLHF) enhancement, which is a form of machine learning that enables a
tool to learn through trial-and-error experimentation (Aleem, 2023; SGA
Knowledge Team, 2023; ul Haq, 2023). As SGA Knowledge Team (2023) puts it,
RLHR offers an added layer of input training, which helps the chatbot to have the
ability to learn from the input and follow prompts so that it can generate
satisfactory responses (Kumar, 2023).

5
2.1.1 Capabilities
ChatGPT, as an AI chatbot has many and diverse capabilities. For example, it is
pre-trained on large amounts of data that enable it to predict an accurate sequence
of words in a sentence. It does this in an autocompletion form in generating
sentences and paragraphs (Kumar, 2023). To this end, Aleem (2023) states that
ChatGPT possesses a hyper-sophisticated autocomplete function. In this sense, it
is an autoregressive model that uses past behaviour (data) to predict future
behaviour (data) (Eliaçık, 2023c). As a GPT 3.5 system, ChatGPT also was trained
on massive databases sourced from the internet, reddit discussions (Kumar, 2023);
Wikipedia, web texts, online articles, books, and other internet-related
information. Together, these databases amount to 570GB (Fraser, 2023; Hughes,
2023; Sharma, 2023).
Additionally, ChatGPT possesses a natural language comprehension because of
its ability to figure out various levels of abstraction from text input. This allows it,
among other things, to answer questions, summarise text, and analyse sentiments,
meaning that it has generative capabilities. Therefore, as a generative AI
employing LLM, ChatGPT can produce large chunks of human-like sentences and
paragraphs, and massive human-like conversational responses. It is able to
remember what was said to it in previous conversations, and allows for follow-
up corrections, including regenerating responses (OpenAI, 2015-2023). Crucially,
ChatGPT possesses contextual language embeddings that help it have a better
semantic understanding through linking words and phrases within their
provided context (Eliaçık, 2023c).
2.1.2 Uses
ChatGPT has different uses or applications. For example, it can respond to written
queries and can write poems, short stories, and songs (in line with an author’s
style) in addition to being able to write essays on nearly any topic. It is able to
create structures for articles (Anders, 2023; Kumar, 2023). Its other application is
to summarise different types of articles or documents, translate text (Eliaçık,
2023c; SGA Knowledge Team, 2023), rectify grammar mistakes, and make
customised recommendations (Eliaçık, 2023c).
What is more, it can edit, remix, and mend writing, as well as define concepts or
terms and simple or complex explanations. Moreover, it can write reports and
cover letters, and produce lesson plans, reports, and email drafts (Cutcliffe, 2022;
Harris, 2022; Hetler, 2023).
2.1.3 Limitations
ChatGPT’s limitations are well known. Even its parent company, OpenAI, openly
flags and acknowledges them on its website. First, some of the information in its
training data lacks recency because the cut-off date for its training data was
September 2021. It has been programmed not to provide harmful or toxic
information (Kumar, 2023; OpenAI, 2015-2023; SGA Knowledge Team, 2023).
Second, there are times when it provides inaccurate or wildly incorrect responses
or answers (Kumar, 2023; OpenAI, 2015-2023; SGA Knowledge Team, 2023; ul
Haq, 2023), or plausible-sounding answers that are nonsensical (OpenAI, 2022).
Called artificial hallucination, this is a propensity in which ChatGPT unexpectedly

6
deviates from its training data output (ul Haq, 2023). The chatbot is also sensitive
to input phrasal tweaks, especially when feeding it the same prompt several times.
For example, it may claim to not know the answer when a prompt is phrased one
way, but will provide the answer when the prompt is tweaked in another way.
Third, the chatbot is sometimes prone to verbosity and overusing certain phrases
due to training data bias. It even has the tendency to guess the intent of the user’s
prompts rather than asking for clarification when the prompt is ambiguous. At
times it displays biased responses or responds to harmful prompts,
notwithstanding a disclaimer that it cannot do so (OpenAI, 2022). Fourth, the
quality of its output depends on the quality of the input it receives (Kumar, 2023;
SGA Knowledge Team, 2023).
Another limitation, but which also may be an advantage, depending on how it is
perceived, is that ChatGPT generates different text responses at different instants
(Aleem, 2023). Moreover, ChatGPT does not understand the sentences it churns
out, nor does it possess the capacity to reason like humans. Instead, all it does is
mimic and reorder human language based on vast numbers of datasets it has been
trained in, and make probabilistic calculations concerning words related to an
answer without even comprehending the underlying concepts for those words.
So, it is reasonable to say that it operates more in the realm of the plausibility of
words than in the truth or moral value of words. To suggest otherwise would be
to commit an anthropomorphic error (Aleem, 2023).
Beyond its limitations, two of its major criticisms have been plagiarism and
copyright laundering. The latter refers to a practice in which information is derived
from existing sources, especially from internet sources, without breaching
copyright (Chaka, 2023b; Hern, 2022). Something worth noting is that as at the
time of writing this paper, OpenAI announced the launch of GPT-4, a ChatGPT
successor. It is said that GPT-4 can respond to images, and caption and describe
them, and process 25,000 words, which is eight times as many as ChatGPT can
(Derico & Kleinman, 2023).
2.2 YouChat
YouChat, which is owned by You.com and was released on 30 December 2022, is
a free-to-use, alternative generative AI to ChatGPT. At the time of writing this
paper, it was still in a beta stage (Ortiz, 2023), and there were not yet many
comments, reports, reviews, analyses, and use cases written about it. However, it
was already functional with no waitlist requirement such as currently
characterising Bing AI (Eliaçık, 2023d). It combines both a generative AI tool and
a search engine (Conroy, 2023; Eliaçık, 2023b, 2023d), and has a conversational or
natural language offering (Eliaçık (2023d). In terms of its architecture, it uses
OpenAI’s GPT-3 model that has been slightly refined. On its website, it states that
it can reply to general queries, suggest ideas, explain things, summarise text,
translate, write code, and compose emails, among other things (Conroy, 2023;
Eliaçık, 2023b, 2023d; Ortiz, 2023; YouChat, 2023). YouChat can also create images,
send letters (Eliaçık, 2023b, 2023d), and respond to math prompts (Ortiz, 2023).
Besides its being a free AI chatbot, two of YouChat’s major differentiating features

7
are that it provides citations to its responses and offers sources from which its
citations have been derived. Sources it cites are from Google. In this case, it has
access to the latest internet sources, which is something that ChatGPT does not
have (Conroy, 2023; Ortiz, 2023).
Moreover, YouChat is capable of generating charts, photos, videos, tables, graphs,
text, or code through its YouChat 2.0’s C-A-L (Chat, Apps, and Links) LLM. All
of this is enabled by YouChat’s integrated YouChat, YouCode, YouWrite, and
YouImagine features (Eliaçık, 2023d). Nevertheless, like any other AI tool, it has
drawbacks, one of which is that it, too, at times, generates incorrect answers or
responses. This is something it acknowledges on its website (YouChat, 2023).
2.3 Chatsonic AI
Chatsonic AI is owned by Writesonic. Like YouChat, at the time of writing this
paper not many comments, reports, reviews, analyses, and use cases had been
written about it. It is based on ChatGPT’s foundational structure and leverages its
capabilities. However, unlike ChatGPT, it has access to the internet as is the case
with YouChat. Four of its differentiating features are: different personas, real-time
data access, a web browser extension, and up-to-datedness. It has a free trial
version and a premium version. The former has a 2,500-word limit per month
(Eliaçık, 2023b), which can be consumed in a large, single response, or, which can
be staggered in smaller responses over a month. This free trial version has been
used for this paper. Chatsonic is supported by Google, has an AI image generator,
and offers voice dictation. The voice dictation feature allows the user to initiate
voice-powered prompts (Ortiz, 2023). As is the case with any AI tool, Chatsonic,
too, is prone to generating incorrect answers or responses.
3. Research Methodology
3.1 Study Design
This study was exploratory in nature. Exploratory research studies are employed
for exploring new areas, or for investigating areas that have not been studied
much (Leavy, 2017; Nkhobo & Chaka, 2021, 2023). The use cases of AI chatbots in
relation to applied English language studies (AELS) in higher education (HE) are
new areas that have not yet been investigated much as the AI chatbots under
study in this paper only came into existence after 30 November 2022.
3.2 Sampling
The study utilised purposive sampling to collect its data sets. Two of the salient
features of purposive sampling are: approaching the sample with a specific
purpose in mind, and predetermining the criteria of what is to be included in the
sample (Alvi, 2016). For this study, the data comprised the responses generated
by ChatGPT, YouChat, and Chatsonic on the selected areas of AELS as informed
by the four prompts stated below. These selected areas were: decolonial applied
linguistics, critical southern decoloniality, and translanguaging, multilanguaging,
and languaging. The purpose was to find out the accuracy and the quality of the
responses these three AI chatbots would generate in these selected areas, based
on the four prompts. AELS is one of research interests of the writer of this paper

8
and the selected areas are some of the areas in which the writer has published
journal articles.
3.3 Data Collection Procedure and Data Analysis
As mentioned above, four prompts related to selected areas of AELS were used as
input to each of the three AI chatbots to yield responses from each of them. These
four prompts, which were in the form of queries, were phrased as follows:
• What is decolonial applied linguistics?
• What is critical southern decoloniality?
• What does Chaka say about critical southern decoloniality?
• What is the difference between translanguaging, multilanguaging, and
languaging?
All the queries were used as input to ChatGPT on 29 January 2023, while all were
used as input to both YouChat and Chatsonic on 07 March 2023. The reason for
this temporal difference is that before 07 March 2023, I did not know about nor
was I aware of the existence of YouChat and Chatsonic including the other
generative AI chatbots mentioned earlier.
All the responses generated by each AI chatbot, as per their respective prompts,
were copied and stored in MS Word files (Appendices A, B and C). The accuracy
and quality of the generated responses were verified and benchmarked against
the relevant sources that were cited by YouChat. In some instances, it failed to cite
sources. Both ChatGPT and Chatsonic generated responses that did not provide
cited sources. This, then, constituted the manner in which these AI chatbot-
generated responses, as data sets for this paper, were analysed.
4. Findings and Discussion
4.1 What Is Decolonial Applied Linguistics?
ChatGPT generated a definitional response to this prompt as depicted in
Appendix A. It viewed decolonial applied linguistics as a theoretical and
methodological framework for studying language and power, the aim of which is
to question and disrupt colonial representations and legacies in applied linguistic
research and practice, and which focuses on colonial impacts on language and
society. It did not cite any source for its generated response. On its landing page,
ChatGPT mentions its capabilities and limitations, and displays its disclaimer
statement (Appendix D). Initially, when YouChat was fed the same prompt, it had
a technical glitch, and exhibited a message that read, “😕 Sorry, too many people
have been asking me questions at once. Give me a moment and try again”, which
was prefaced by a sadness emoji. It also depicted a disclaimer that read as follows:
“This product is in beta and its accuracy may be limited. You.com is not liable for
content generated” (Appendix B). Nonetheless, it was able to display three
possible sources that dealt with some aspects of decolonial applied linguistics, one
of which was from Rosa and Flores’ (2021) Decolonization, language, and race in
applied linguistics and social justice (Appendix B). When, finally, YouChat was able
to generate its response to the prompt under the spotlight, it identified decolonial
applied linguistics as a field of study the purpose of which is to decolonise
language and linguistics through challenging the power dynamics and

9
assumptions informing traditional language studies. It also contended that this
type of linguistics attempts to disrupt and problematise Eurocentric biases and
colonial legacies underpinning language and discourse, and advocates just,
inclusive, and equitable linguistic environment in which all languages and
language varieties have equal respect and value (Appendix B). However, no
sources were cited.
The same prompt was used as input to Chatsonic. The chatbot, which has a
Regenerate feature that ChatGPT also has, identified decolonial applied
linguistics as a field of study aiming at challenging colonial legacies and power
imbalances inherent in language and linguistics, and Western, Eurocentric views
informing traditional approaches. It stated that this linguistics decentres these
views so as to create an inclusive, equitable approach to language, which
accommodates diverse and marginalised linguistic practices. Additionally, it
pointed out that this type of linguistics is aligned with postcolonial studies, critical
race theory, and interdisciplinary approaches that question hegemonic power
structures (Appendix C). Chatsonic, too, did not cite sources for its generated
response.
At a mechanical, academic level, the three AI chatbots managed to respond to this
prompt, even though YouChat initially had a technical glitch. However, not one
of them cited or referenced the sources of their responses, even when it was
manifestly evident that they stitched together their responses from some currently
published online sources. In this case, this highlights their proclivity to plagiarise
others’ ideas in generating their responses to a prompt. The odd thing is that
YouChat provided titles of its reference sources in an instance in which it did not
give an actual response to this prompt due to its technical glitch. This points to
some inconsistency on its part. Both YouChat and Chatsonic identified decolonial
applied linguistics as a field of study, while ChatGPT recognised it as a theoretical
and methodological framework for studying language and power.
At a substantive, qualitative level, the three chatbots responses shared certain core
aspects. Two of these aspects are inclusivity and equitability, two notions which
have to do with social justice, or, in this case, with linguistic justice (Piller, 2021;
Randolph Jr. & Johnson, 2017). One of the sources listed by YouChat in an instance
in which it could not provide a response as mentioned above, which deals with
inclusion and equality, is Rosa and Flores’ (2021) work (journal article). This work
critiques the notions of inclusion, equity and affirmation as they relate to
marginalised languages, and as advocated and promoted by mainstream applied
linguistics. It maintains that these notions are a deceptive trap that feeds into
normative sociolinguistic and applied linguistic logics and notions (e.g., linguistic
deficiency frequently associated with “raciolinguistic Others”) (Rosa & Flores,
2021, p. 1164), which are grounded on global colonialism, racism, and whiteness.
Additionally, it foregrounds a raciolinguistic perspective in which racism and
colonialism, and not conventional language use, are centred in engaging with
applied linguistics. While this work has a decolonial angle, it explicitly states that
it has less interest in associating itself with any specific decolonial or anticolonial

10
perspective, but rather, that it is more inclined to centring and critiquing global
colonialisms.
These points of departure in this particular work are some of the nuances that the
three AI chatbots could not pick up in their responses. Their responses lacked
qualitative richness (essential qualitative nuances) and the accuracy of detail.
They all tended to uncritically associate decolonial applied linguistics with
inclusivity and equity. The same applies to the other two reference sources, Motha
(2020) and Canagarajah (2022), listed by YouChat. For example, the former
reference source is focused on antiracist and decolonising applied linguistics, and
not just on decolonial linguistics, while the latter reference source foregrounds
disability studies and (crip) linguistics. Needless to say, there are other online
sources dealing with some aspects of decolonial applied linguistics, aspects of
which comprise the three responses, but which have not been acknowledged in
any of these responses (Appendix E). Two of such online sources are Chaka (2021)
and Makoni and Severo (2022).
4.2 What Is Critical Southern Decoloniality?
Concerning this prompt, the three chatbots responded and reacted in a manner
similar to how they responded to the first prompt. For example, YouChat
exhibited its system’s instability and unreliability: it had the same technical glitch
and only generated a response in the second attempt, except that it did not list any
reference source related to the prompt in the first attempt. Rather, it did so with
the second attempt when it was able to generate a response. It listed six sources,
which it had not cited within its response. Two of these listed reference sources
were Chaka (2022a) and Chaka (2022b) (Appendix B). Pertaining to the three
chatbots’ responses, two of them, a ChatGPT’s response and a YouChat’s
response, mimicked, verbatim, some of the phrases and clauses they had
generated for responding to the first prompt, What is decolonial applied linguistics?
That is, of ChatGPT’s response comprising 131 words, 37 words were the same
as those it used in its first response. Similarly, 69 words of the 114 words that
YouChat used to respond to the second prompt were exactly the same as those it
generated in responding to the first prompt (Appendices A and B). Chatsonic’s
response to the second prompt sparingly replicated the words it had used in its
first response to the first prompt. Both ChatGPT and Chatsonic had few identical
phrases in their responses, two examples of which were passive recipient and active
resistance.
This second prompt, again, reflects the inconsistency in which YouChat generated
its response: listing sources it had failed to cite or acknowledge. It also
demonstrates the propensity for both ChatGPT and Chatsonic to generate
responses without citing their reference sources and without providing any
reference list for them. This practice, which they displayed in their responses to
the first prompt, is tantamount to plagiarism, as the responses they generated are
scholarly published information available online. Of the six sources listed by
YouChat for its unacknowledged response, only two were directly related to
critical southern decoloniality (CSD). The rest were not. In fact, they have little to
do with this notion in its current conceptualisation. The two reference sources that

11
have a direct relation to CSD are Chaka (2022a) and Chaka (2022b). Both of these
reference sources use the acronym, CSD, which YouChat also uses right at the
beginning of its response, and twice in this response. However, they are employed
in two different and unrelated contexts: CSD as an approach to datafication,
algorithms, and digital citizenship; and CSD as a two-eyed framing to critique,
problematise, and challenge knowledge production practices (the geopolitics of
knowledge production) in applied English language studies (AELS). By contrast,
the responses generated by the three AI chatbots referred to CSD generically in
relation to traditional language studies, non-Western languages (cf. YouChat’s
response in Appendix B), colonialism, imperialism, and the Global South (cf.
ChatGPT’s response and Chatsonic’s response in Appendices A and C,
respectively). They could not detect these finer nuances and their accompanying
differential usage contexts. It should, nonetheless, be mentioned that ChatGPT’s
response made reference to the fact that CSD critiques (dominant) research and
knowledge production. But that was all it could say. It is also worth mentioning,
as pointed out earlier on, that of the three chatbots, ChatGPT’s training data cut-
off date is 2021. Overall, then, the three AI chatbots’ responses lacked the accuracy
of detail and were devoid of fundamental subtle differences inherent in the use of
CSD by Chaka (2022a) and Chaka (2022b).
4.3 What Does Chaka Say About Critical Southern Decoloniality?
Regarding this prompt, only YouChat generated, at face value, a rather
convincing response, that had in-text citations and references for the cited sources.
The other two chatbots responses were not up to scratch. For example, ChatGPT
said that “I’m not aware of any specific quotes or writings from an individual
named ‘Chaka’ on the topic of critical southern decoloniality” (Appendix A). Then
it went on to assert that CSD is a relatively recent (new and emerging) framework.
Yet, in its response to the second prompt above, it never made such a claim. It,
thereafter, offered completely different and new information about CSD as a
complex and multidisciplinary framework. Chatsonic started its response by
making up a surname for Chaka, and continued to assert that this person had
written much on both CSD and decolonisation. Besides, mimicking some of the
terms and phrases it used in its response to the second prompt, most of the views
it attributed to Chaka, barring knowledge production, had nothing to do with
Chaka’s views of CSD as highlighted under the second prompt above.
With reference to YouChat, it correctly identified Chaka’s professional title, his
academic department, and his affiliation, and referenced this information using
Chaka’s Academia.edu’s online profile. Then, it regurgitated the phrases and
clauses it used in its response to the second prompt by providing two in-text
citations for one part of its response, but not by offering any citation for the
remaining part. The two citations it referenced were Chaka’s ResearchGate’s
online profile and Chaka (2022a). The three sources of reference it listed at the end
of its response were Chaka (2022a), Chaka (2022b), and Ndlangamandla and
Chaka (2022). As the focus of the first two reference sources and their use of CSD
were mentioned under the second prompt above, only the third reference source
is worth contextualising. This source of reference employs CSD specifically for
challenging Eurocentric scholarship of teaching and learning (SoTL) practices and

12
colonialist English language paradigms. It does not appropriate it in a broad-
stroke manner suggested by YouChat’s response.
What emerges from the three chatbots’ responses to this prompt is that YouChat
correctly identified the personal, professional, and affiliation detail of the scholar
whose name was mentioned in the prompt. It provided in-text citations for one
part of its response, but did not do so for the other part, something which lends
itself well to plagiarism. It offered a correct reference list for its cited sources, even
for the other two sources it had not cited. In this sense, it was consistent in one
instance, but inconsistent in another instance. ChatGPT could not recognise the
scholar mentioned in the prompt, but went on to provide the new information
about CSD, which it did not provide in its response to the second prompt above.
In addition, it plagiarised its response as it did not acknowledge it. For its part,
Chatsonic invented the surname of the scholar mentioned in the prompt, moved
on to regurgitate parts of its response to the second prompt, and started
hallucinating (Anders, 2023; Browne, 2023; Knight, 2023; Metz, 2022; ul Haq, 2023)
certain parts of its response, which it misattributed to the scholar in question.
Again, the three AI chatbots responses lacked the accuracy of substantive details,
except for the correct mechanical/personal details that YouChat generated.
Additionally, these three responses did not reflect the fundamental subtle
differences Chaka (2022a), Chaka (2022b), and Ndlangamandla and Chaka (2022)
make in their use of CSD.
4.4 What Is the Difference Between Translanguaging, Multilanguaging, and
Languaging?
The three AI chatbots responses to these fourth prompts resembled their
responses to the first two prompts. For instance, YouChat was unstable and
unreliable in the first prompt that it was fed. It only generated the actual response
in subsequent attempts. It listed its three sources of reference at the end of its
response for which it had not provided in-text citations. And, its response was
way shorter (96 words) compared to ChatGPT’s 292-word-long response and to
Chatsonic’s 209-word-long response. The three responses exhibited a similarity of
words and phrases, and none of them had any in-text citations. With its
telegraphic definition and differentiation of the three concepts mentioned in the
prompt, YouChat regards translanguaging as a process in which multiple
languages and varieties of language are used for effective and accurate
communication. It states that multilanguaging is similar to translanguaging, even
though it is about using multiple languages in a single interaction. Then, it says
languaging is a process in which multiple languages and varieties of language are
used in an interaction. To this end, it points out that the major difference between
these three concepts is that translanguaging focuses on the fluidity of language
use, while both multilanguaging and languaging are about using multiple
languages in an interaction (Appendix B).
ChatGPT starts off its response by stating that the three concepts are both related
but different. It sees translanguaging as a framework involving the dynamic and
fluid use of language by speakers possessing wide-ranging linguistic resources,
which they use simultaneously in a given context. Its view of multilanguaging is
that it entails the use of multiple languages by speakers in different domains or

13
contexts (e.g., home, school, or workplace). Concerning languaging, it points out
that it is an ongoing and dynamic process (the act of making language itself) in
which speakers use language creatively to construct and communicate meaning.
It, then, maintains that the three concepts involve multilingualism and language
use (Appendix A).
For its part, Chatsonic begins by saying that the three concepts are related in that
they focus on fluid and dynamic language use, even though their meanings are
slightly different. It regards translanguaging as a meaning-making and effective
communication process in which speakers draw on all available linguistic
resources, including different languages, dialects, and registers. This process
entails moving between languages fluidly/fluently. In addition, it asserts that
multilanguaging has similarity to translanguaging, but that it foregrounds the use
of multiple languages (multilingualism) in a single interaction or conversation. It
states that languaging is a broader concept referring to the process in which
meaning is made by means of any kind of communication that includes language,
facial expressions, gestures, and other types of non-verbal communication.
Finally, Chatsonic contends that the common point shared by the three concepts
is their emphasis on the fluid and dynamic use of language and their valorising
of linguistic and cultural diversity. However, it says their difference lies in their
focus and scope, with both translanguaging and multilanguaging foregrounding
the use of multiple languages, whereas languaging is more generic as it entails all
communication forms.
YouChat regards the three concepts mentioned in the fourth prompt as processes
involving the use of multiple languages for communication (translanguaging) in
a single interaction (multilanguaging and languaging). This is more of an
alternation between various languages. It sees translanguaging’s focus on the
fluidity of language use as its differentiating factor. The view of translanguaging,
multilanguaging, and languaging as processes is too limiting and superficial. The
same applies to fluidity as a differentiating factor between translanguaging and
multilanguaging and languaging, and to reducing the three concepts to
communication alone. ChatGPT sees the three concepts as simultaneously related
and unrelated, and maintains that translanguaging is a framework (unnamed) for
the dynamic and fluid use of language. Its view of multilanguaging is not
dissimilar to the perspective it attaches to translanguaging. Its characterisation of
translanguaging is actually an alternation among multiple languages, and among
varieties of language. Thus, its view of these three concepts and its
characterisation of them are too shallow and mechanical. Chatsonic, like
ChatGPT, asserts that the three terms are both related and unrelated (their
meanings slightly differ), with their relatedness being the fluid and dynamic
language use. It says the slight difference in the meanings of the three terms is
their focus and scope. This is very vague and unhelpful as one does not know
what both focus and scope in this case entail. Its reference to languaging as a more
generic term for communication is equally vague and shallow.
This, then, takes us to the three reference sources YouChat listed at the end of its
response, but which it did not cite within its response. These were Li (2018a), Li

14
(2018b), and Mora et al. (2022). I will briefly use the first source as a case in point,
and highlight only its key relevant aspects related to the three responses. Li
(2018a), who uses translanguaging with a capital “T” and whose article’s major
objective is to explicate the theoretical reasons for translanguaging, responds to
some of the questions asked about it, and clarifies some of the confusion related
to its increasing usage, talks about translanguaging as a theory of language
(theoretical concept) and as a pedagogical practice. He also focuses on a
translanguaging space, the translanguaging instinct, and translanguaging and
multimodality. Additionally, he argues that translanguaging is transformative
and re-envisions language as a multilingual, multimodal, multi-semiotic, and
multisensory resource for meaning- and sense-making. Importantly, he contends
that translanguaging challenges and breaks border between named languages,
and between language varieties. Concerning languaging, which Li (2018a)
varyingly writes in lower-case “l” and in capital “L”, and whose origins he traces
to multidisciplinary fields of study, it might do to sum up his view of it as a
heterogeneous human linguistic performance that challenges named and
nameable languages, formalistic language structures, and the divide between
linguistic, paralinguistic, and extralinguistic properties of human communication.
All of these truncated nuances of both translanguaging and languaging are what
the three AI chatbots’ responses could not pick up. Rather, their responses strung
together some of the words and phrases used in Li’s work (2018a) without
matching them to their related and underlying finer nuances.
5. Implications for Applied English Language Studies (AELS)
Of the three AI chatbots tested and discussed in this paper, YouChat appears to
be an AI chatbot dogged by technical glitches and instability. It also displayed
inconsistency in generating responses: in some instances, it never provided in-text
citations for its responses, but in one instance it did. This inconsistency is a
drawback for AELS undergraduate and graduate students looking for generated
responses related to their discipline, which are always acknowledged through in-
text citations. Even in instances where it provided lists of references for its
responses, some of the sources listed in those reference lists were not entirely
relevant to the generated information. This is another pitfall. The other two AI
chatbots, ChatGPT and Chatsonic, exhibited a proclivity to generate uncited
responses. As such, they seem to be prone to generating plagiarised information
from their training data (ChatGPT) and from the internet (Chatsonic). This is one
of the major shortcomings these two chatbots currently have. All these
shortcomings displayed by the three chatbots manifestly imply that AELS
undergraduate and graduate students need to consult the relevant sources of
information (e.g., journal articles, books, and monographs), many of which are
now available online, and for their teachers/professors to know that others’ views
are always acknowledged, and for students to master citation and referencing
skills.
Additionally, the three chatbots displayed a tendency to generate almost similar
responses for different and unrelated prompts. Not only that, but in one instance,
one of them (ChatGPT) could not recognise a scholar mentioned in the prompt,
while the other one (Chatsonic) misrecognised the scholar in question and

15
misappropriated the views it generated in its response to him. Thereafter, it
reproduced parts of its response to the second prompt, and hallucinated the other
parts of its response. Again, in this case, AELS undergraduate and graduate
students have to rely on relevant original sources and on their teachers/professors
to get the credible and reliable type of scholarly information.
Moreover, the three AI chatbots simplistically and superficially parsed phrases
and ideas from uncited sources without detecting the nuances inherent in the
ideas with which those sources deal. Importantly, the responses of the three AI
chatbots lacked the accuracy of substantive details. All of this is tantamount to
generating a fluffy form of knowledge, which flies in the face of the deep, credible,
nuanced form of knowledge that AELS undergraduate and graduate students are
eagerly looking for in their discipline.
All of the above-mentioned shortcomings mean that AELS undergraduate and
graduate students and scholars need always to double-check the authenticity,
credibility, and depth of the responses generated by these three AI chatbots. These
shortcomings also mean that only the uninitiated undergraduate and graduate
students might be persuaded to believe and blindly accept the responses
(answers) generated by these chatbots to the prompts they were fed in this paper
as the correct and credible responses. Undergraduate and graduate students who
are well-grounded in the AELS aspects discussed in this paper will not be
persuaded to do so. In view of how the three chatbots performed pertaining to the
prompts they were required to respond to in this paper, it is plausible to say that
they do not yet signal the end of nor a threat to human-generated or classroom-
based knowledge. Neither do they spell the end of original thinking or original
ideas (Careen, 2023; Coleman, 2023). Maybe, in this case, the role these chatbots
can play is that of primers and catalysts for discussing and debating the types of
AELS information generated by AELS undergraduate and graduate students and
scholars.
6. Conclusion and Future Research
This paper was aimed at comparing the accuracy and quality of the responses
produced by the three AI chatbots, ChatGPT, YouChat, and Chatsonic, based on
the prompts related to selected areas of applied English language studies (AELS).
It also provided the educational and knowledge implications of the generated
responses for AELS. YouChat stood out as a technically unstable and unreliable
chatbot with a degree of inconsistency in generating responses. The other two
chatbots, ChatGPT and Chatsonic, consistently displayed a propensity to
plagiarise responses from the information available on the internet without
acknowledging sources. In certain instances, the three chatbots generated nearly
similar responses for different and unrelated prompts, something which made
their responses look like run-of-the-mill responses that lacked credibility,
accuracy, and quality. One chatbot (ChatGPT) failed to recognise a scholar
mentioned in a prompt, while the other one (Chatsonic) misrecognised this
scholar, and ended up hallucinating parts of its response. Again, the three
chatbots mechanically and superficially strung together phrases and ideas in their
responses without detecting the subtleties associated with them in the original

16
sources that used them. This caused the knowledge embedded in those responses
to appear too flossy and to lack nuances. Given all these shortcomings, these three
AI chatbots are not yet credible and reliable generators of knowledge for the
aspects of AELS discussed in this paper.
As generative AI chatbots are emerging technologies, the presence of which has
been heralded by ChatGPT, more research is needed to study the accuracy and
quality of the responses these AI technologies generate in AELS as well as in other
academic subject areas offered at the higher education (HE) level. This is crucial
as there is an ever-increasing overload of information across academic disciplines
in HE. In the midst of incremental information overload and in the era of AI
chatbots, there is a need to verify and authenticate the credibility and integrity of
the information provided by AI chatbots like ChatGPT, YouChat, and Chatsonic,
and many others, by both faculty and students. Failure to do so, will result in a
shoddy and fluffy form of knowledge being accepted as credible and sound. This
is what the current study has attempted to avoid by investigating the types of
academic responses the three AI chatbots generated in respect of selected areas of
AELS.
7. References
Ackerman, S. (1992). Discovering the brain. National Academy Press.
Adesso, G. (2023). PT4: Towards the ultimate brain: Exploring scientific discovery with
ChatGPT AI. Authorea. https://doi.org/10.22541/au.167052124.41804127/v2
Aleem, Z. (2023). No, ChatGPT isn’t willing to destroy humanity out of ‘wokeness’.
https://www.msnbc.com/opinion/msnbc-opinion/chatgpt-slur-conservatives-
woke-elon-rcna69724
Alvi, M. (2016). A manual for selecting sampling techniques in research.
https://mpra.ub.uni-muenchen.de/70218/
Anders, B.A. (2023). Why ChatGPT is such a big deal for education.
https://scalar.usc.edu/works/c2c-digital-magazine-fall-2022---winter-
2023/why-chatgpt-is-bigdeal-education
Atallah, H.E., Frank, M.J. & O’Reilly, R.C. (2004). Hippocampus, cortex, and basal ganglia:
Insights from computational models of complementary learning systems.
Neurobiology of Learning and Memory, 82, 253-267.
Bowman, E. (2023). A college student created an app that can tell whether AI wrote an
essay. https://www.npr.org/2023/01/09/1147549845/gptzero-ai-chatgpt-
edward-tian-plagiarism
Browne, R. (2023). All you need to know about ChatGPT, the A.I. chatbot that’s got the
world talking and tech giants clashing.
https://www.cnbc.com/2023/02/08/what-is-chatgpt-viral-ai-chatbot-at-heart-
of-microsoft-google-fight.html
Canagarajah, S. (2022). A decolonial crip linguistics. Applied Linguistics, XX(XX), 1-22.
https://doi.org/10.1093/applin/amac042
Ceres, P. (2023). ChatGPT is coming for classrooms. Don't panic.
https://www.wired.com/story/chatgpt-is-coming-for-classrooms-dont-
panic/#intcid=_wired-bottom-recirc_9c0d2ac5-941b-45c7-b9ac-
7ac221fc2e33_wired-content-attribution-evergreen
Chaka, C. (2020). Skills, competencies and literacies attributed to 4IR/Industry 4.0:
Scoping review. IFLA Journal, 46(4), 369-399.
https://doi.org/10.1177/0340035219896376

17
Chaka, C. (2021). Alan Davies: Ostensive views, other views and native speakerism, and
the implications of the latter for English language teaching. Advances in Language
and Literary Studies, 12(6), 79-86. http://dx.doi.org/10.7575/aiac.alls.v.12n.6.p.79
Chaka, C. (2022a). Digital marginalization, data marginalization, and algorithmic
exclusions: A critical southern decolonial approach to datafication, algorithms,
and digital citizenship from the Souths. Journal of e-Learning and Knowledge Society,
18(3), 83-95. https://doi.org/10.20368/1971-8829/1135678
Chaka, C. [Chaka Chaka]. (2022b, August 10). The geopolitics of knowledge production
in applied English language studies: Transknowledging and a two-eyed critical
southern decoloniality [Video]. YouTube. https://youtu.be/xb9seQEUR3M
Chaka, C. (2023a, May 30). Doing research in the age of AI-powered chatbots: English
Studies’ M&D student support at UNISA [PowerPoint slides].
Chaka, C. (2023b). Stylised-facts view of fourth industrial revolution technologies
impacting digital learning and workplace environments: ChatGPT and critical
reflections. (Unpublished manuscript.)
Conroy, S. (2023). ChatGPT vs YouChat. https://www.wepcgpt.com/tips/chat-gpt-vs-
youchat/
Coleman, T. (2023). Is ChatGPT a threat to English class? The Week.
https://theweek.com/feature/opinion/1020101/is-chatgpt-a-threat-to-english-
class
Cutcliffe, J. (2022). ChatGPT, chatbots and artificial intelligence in education.
https://ditchthattextbook.com/ai/
Deacon, T.W. (1997). What makes the human brain different? Annual Review of
Anthropology, 26, 337-357.
Dilmegani, C. (2023). ChatGPT education use cases, benefits & challenges in 2023.
https://research.aimultiple.com/chatgpt-education/
Derico, B. & Kleinman, Z. (2023). OpenAI announces ChatGPT successor GPT-4.
https://www.bbc.com/new/technology-64959346
Eliaçık, E. (2023a). Google unveils its experimental conversational AI service Bard.
https://dataconomy.com/2023/02/how-to-use-google-bard-ai-chatbot-
examples/
Eliaçık, E. (2023b). Best ChatGPT alternatives: What awaits us beyond ChatGPT?
https://dataconomy.com/2023/02/beyond-chatgpt-alternatives-best-jasper/
Eliaçık, E. (2023c). The role of large language models in the AI war.
https://dataconomy.com/2023/02/best-large-language-models-meta-llama-ai/
Eliaçık, E. (2023d). You.com’s AI-powered features are already in use, unlike Bing and
Google. https://dataconomy.com/2023/02/you-com-ai-chatbot-ai-image-
generator-how/#YouChat_A_rival_to_ChatGPT
Fraser, L. (2023). Banking and legal experts think ChatGPT could be a major disruptor.
https://businesstech.co.za/news/business/661863/banking-and-legal-experts-
think-chatgpt-could-be-a-major-disruptor/
Harris, M. (2022). ChatGPT: The game-changing app every teacher should know about.
https://www.learnersedge.com/blog/chatgpt-the-game-changing-app-every-
teacher-should-know-about
Hern, A. (2022). AI bot ChatGPT stuns academics with essay-writing skills and usability.
https://www.theguardian.com/technology/2022/dec/04/ai-bot-chatgpt-stuns-
academics-with-essay-writing-skills-and-usability.
Hetler, A. (2023). Bard vs. ChatGPT: What’s the difference?
https://www.techtarget.com/whatis/feature/Bard-vs-ChatGPT-Whats-the-
difference

18
Hughes, A. (2023). ChatGPT: Everything you need to know about OpenAI's GPT-3 tool.
https://www.sciencefocus.com/future-technology/gpt-3/
Kamran, S. (2023). Microsoft Bing vs Google Bard: ChatGPT cloning battle has begun.
https://techcommunity.microsoft.com/t5/itops-talk/microsoft-bing-vs-google-
bard-chatgpt-cloning-battle-has-begun/m-p/3737979
Knight, W. (2023). Meet Bard, Google’s answer to ChatGPT.
https://www.wired.com/story/meet-bard-googles-answer-to-chatgpt/
Kumar, H.S.A. (2023). ChatGPT: How artificial intelligence is changing the way we
communicate. https://skill-lync.com/blogs/chatgpt-how-artificial-intelligence-
is-changing-the-way-we-communicate
Leavy, P. (2017). Research design: Quantitative, qualitative, mixed methods, arts-based, and
community-based participatory research approaches. The Guilford Press.
https://doi.org/10.1111/fcsr.12276
Li, W. (2018a). Translanguaging as a practical theory of language. Applied Linguistics, 39(1),
9-30. https://doi.org/10.1093/applin/amx039
Li, W. (2018b). Translanguaging and code-switching: What’s the difference?
https://blog.oup.com/2018/05/translanguaging-code-switching-difference/
Makoni, S. & Severo, C. (2022). Southern perspectives of language and the construction
of the common. Language & Communication, 86, 80-86.
https://doi.org/10.1016/j.langcom.2022.06.003
Meghmala. (2023). GPTZero: An app to detect whether text is written by ChatGPT.
https://www.analyticsinsight.net/gptzero-an-app-to-detect-whether-text-is-
written-by-chatgpt/#:~:text=Student%2Dmade%20app%20GPTZero%20-
determines,for%20unethical%20usage%20in%20academia.
Metz, K. (2022). Experts warn about the danger of chatbots using ChatGPT and LaMDA
to create ‘hallucinations’ without considering what is right. The New York Times.
https://www.nytimes.com/2022/12/10/technology/ai-chat-bot-chatgpt.html
Mora, R. A., Tian, Z. & Harman, R. (2022). Translanguaging and multimodality as flow,
agency, and a new sense of advocacy in and from the Global South. Pedagogies:
An International Journal, 17(4), 271-281.
https://doi.org/10.1080/1554480X.2022.2143089
Motha, S. (2020). Is an antiracist and decolonizing applied linguistics possible? Annual
Review of Applied Linguistics, 40, 128-133.
https://doi.org/10.1017/S0267190520000100
Nawrocki, R. A. (2011). Simulation, application, and resilience of an organic neuromorphic
architecture, made with organic bistable devices and organic transistors.
(Publication No. 891) [Master’s thesis, University of Denver].
https://core.ac.uk/reader/217241530
Ndlangamandla, S. C. & Chaka, C. (2022). Relocating English Studies and SoTL in the
Global South: Towards decolonizing English and critiquing the coloniality of
language. Journal of Contemporary Issues in Education, 17(2), 39-56.
https://doi.org/10.20355/jcie29495
Nkhobo, T. & Chaka, C. (2021). Exploring instances of Deleuzian rhizomatic patterns in
students’ writing and in online student interactions. International Journal of
Learning, Teaching and Educational Research, 20(10), 1-22.
Nkhobo, T. & Chaka, C. (2023). Syntactic pattern density, connectives, text easability, and
text readability indices in students’ written essays: A Coh-Metrix analysis.
Research Papers in Language Teaching and Learning, 13(1), 121-136.

19
Ofgang, E. (2023). What is GPTZero? The ChatGPT detection tool explained by its creator.
https://www.techlearning.com/news/what-is-gptzero-the-chatgpt-detection-
tool-explained
OpenAI. (2022). Introducing ChatGPT. https://openai.com/blog/chatgpt#OpenAI
OpenAI. (2015-2023). OpenAI. https://openai.com/about
Ortiz, S. (2023). The best AI chatbots: ChatGPT and other interesting alternatives to try.
https://www.zdnet.com/article/best-ai-chatbot/
Piller, I. (2021). Language social justice. In J. Stanlaw (Ed.). The International Encyclopedia of
Linguistic Anthropology (pp.1-7). John Wiley & Sons.
Pittalwala, I. (2023). Is ChatGPT a threat to education?
https://news.ucr.edu/articles/2023/01/24/chatgpt-threat-education
Randolph, L. J., Jr. & Johnson, S. M. (2017). Social justice in the language classroom: A call
to action. Dimension, 99, 99-121.
Rosa, J. & Flores, N. (2021). Decolonization, language, and race in applied linguistics and
social justice. Applied Linguistics, 42(6), 1162-1167.
https://doi.org.10.1093/applin/amab062
SGA Knowledge Team. (2023). The new buzz in town: What is ChatGPT and why has it
taken the world by storm? https://www.sganalytics.com/blog/what-is-chatgpt-
how-to-use-it-its-limitations/
Sharma, S. (2023). Generative AI and large language models: The AI gold rush.
https://www.idsa.in/system/files/issuebrief/ib-sanur-sharma-010323.pdf
Solé, E. (2023). Teachers sound off on ChatGPT, the new AI tool that can write students’
essays for them. https://www.today.com/parents/teens/chatgpt-ai-teachers-
students-cheat-essays-rcna63352
Stiennon, N., Ouyang, L., Wu, J., Ziegler, M.D., Lowe, R., Voss, C. & Christiano, P.
(2020). Learning to summarize from human feedback.
https://proceedings.neurips.cc/paper/2020/file/1f89885d556929e98d3ef9b8644
8f951-Paper.pdf
Tech Desk (2023). GPTZero to help teachers deal with ChatGPT-generated student essays.
https://indianexpress.com/article/technology/gptzero-app-helps-teachers-
catch-chatgpt-8378062/
Tyrrell, J. (2023). ChatGPT screening: OpenAI text classifier versus GPTZero app.
https://techhq.com/2023/02/chatgpt-screening-openai-text-classifier-versus-
gptzero-app/
ul Haq, F. (2023). Bing, Bard, and ChatGPT: Why AI advances are great news for devs.
https://www.educative.io/blog/bing-bard-chatgpt-ai-software-developers
Wiggers, K. (2023). Most sites claiming to catch AI-written text fail spectacularly.
https://techcrunch.com/2023/02/16/most-sites-claiming-to-catch-ai-written-
text-fail-spectacularly/
YouChat. (2023). Who are you.
https://you.com/search?q=who+are+you&tbm=youchat&cfr=chat

20
©Authors
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivatives 4.0
International License (CC BY-NC-ND 4.0).
Vol. 22, No. 6, pp. 20-44, June 2023
Received Jan 19, 2023; Revised Apr 19, 2023; Accepted Jun 10, 2023
Influence of Teacher-Targeted Bullying
Behaviour on Teachers in Selected Schools in
South Africa
Japsile Sambo and Sumeshni Govender
University of Zululand, KwaZulu-Natal, South Africa
Abstract. South African schools seem to be experiencing an increase in
incidents of teacher-targeted bullying (TTB). Previous studies conducted
in South Africa and abroad have proved that teacher-targeted bullying
is a serious challenge in schools and needs immediate intervention. This
study explored the influence of learners’ bullying behaviour on teachers
in selected schools in Ehlanzeni District. A semi-rural area in
Mpumalanga Province, South Africa. A qualitative approach was
applied in this investigation, and a sample of six teachers was
purposefully selected from three high schools. The data collected from a
questionnaire, was analysed using thematic analysis. Findings revealed
that teachers have experienced verbal bullying, physical bullying,
psychological bullying, cyberbullying, and sexual bullying, in incidents
which normally take place in the classroom or elsewhere on the school
premises. These experiences reportedly evoked disturbing emotions,
affected teachers’ lives at school, their personal lives, and undermined
their professional standing. Findings further indicated that there are no
strategic interventions put in place by schools or the Department of
Basic Education to assist teachers who are victims of teacher-targeted
bullying. It can be concluded that TTB poses a significant threat to
teachers’ mental health, physical health, performance and to the
teaching profession as a whole. If the Department of Education does not
introduce intervention strategies to try and resolve this issue, there may
be an increase in the number of teachers resigning from their jobs, which
will escalate the crisis.
Keywords: bullying; learner; teacher; teacher-targeted bullying
1. Introduction
Bullying affects teachers worldwide. This study explored the extent of teacher
bullying in selected schools in Ehlanzeni District, Mpumalanga Province, South
Africa. Previous international studies on teacher-targeted bullying (TTB)
conducted in Estonia, Finland, Ireland, Malaysia, New Zealand, South Africa
and Turkey (Kõiv & Aia-Utsal, 2021), revealed that there is a greater need for the

21
recognition of TTB by learners within the whole-school organisation (Hoffmann,
2013). Jacobs and de Wet (2018) stated that bullies are disobedient learners who
lack parental support or guidance. Management styles may motivate classroom
and school TTB, and an absence of constructive values in society, along with a
disrespect for authority figures, perpetuate bullying. According to Jacobs and de
Wet (2018), teachers are essential in any country’s Department of Education.
They therefore need protection from learners who are bullying them. The
education system should look at more productive ways that need to be
implemented to show that learners’ bullying of their teachers is taken as
seriously as any other bullying that occurs in the country. This would assist
those teachers who were bullied in regaining their self-confidence and dignity.
Teacher-targeted bullying seems to be a discourse that is prevalent in South
African schools. This discourse has not been given much attention since most
studies focus on bullying that is inflicted on learners by other learners or
perpetrated by their teachers. Teachers in South Africa experience verbal,
physical, indirect and cyber bullying (Jacobs & de Wet, 2018). Bullying leads to
serious challenges for the teachers who are experiencing it, so the government
and schools may want to consider ways of supporting teachers who face teacher-
targeted bullying (Santos & Tin, 2018). Some teachers manage to cope with and
survive the fact that they are bullied in their work environment, but others may
not, since human beings react differently to different situations. There are
limited strategies and policies which are currently being implemented by the
Department of Education that are aimed at eradicating teacher-targeted
bullying.
The term ‘teacher-targeted bullying’ has been used by researchers in reference to
bullying which is committed by learners against their teachers in school. This
term was adopted in this study. Specifically, the research study attempted to
answer the following questions:
1. What types of bullying behaviour do learners use against teachers in selected
schools?
2. What influence does learners’ bullying behaviour have on teachers in
selected schools?
3. Which strategies would assist teachers who experience learners’ bullying
behaviour to better manage the situation in selected schools?
2. Literature Review
2.1. Background on Teacher-Targeted Bullying
Teacher-targeted bullying has been defined as a variety of aggressive behaviour
involving the victimisation of teachers by learners (Moon & McCluskey, 2016).
Teachers are not only bullied by learners, but they also experience bullying from
their colleagues and school management team (such as the principal and
administrative staff) at the school (Pyhältö, Pietarinen, & Soini, 2015; Woudstra,
2015; Jacobs & de Wet, 2018). Bullying occurs where there are interactions, and
since teachers interact with all the learners, the School Management Team (SMT)
and other teachers, there is a possibility of bullying taking place. Teacher-
targeted bullying is a form of workplace bullying because it occurs in teachers'

22
work environments. However, this type of bullying is different from other
perceived workplace bullying, in that the individual who has a lower position
within the organisation inflicts the bullying behaviour on someone of a higher
position (Santos & Tin, 2018).
Work on TTB started at the end of the previous century, but research on this
type of workplace bullying has barely moved beyond the creation of awareness
that learners direct bullying at their teachers. This is an area that has not been
given much attention in the past in terms of research (Hoffmann, 2013;
Woudstra, 2015; Jacobs & de Wet, 2018; Qiao, 2018; Santos & Tin, 2018;
Woudstra, et al., 2018; Payne & Gottfredson, 2019; Yang, et al., 2019). Few
studies were found to acknowledge the presence of TTB (Hoffmann, 2013;
Woudstra, 2015; Qiao, 2018; Woudstra, et al., 2018; Billett et al., 2019). There is a
need for greater recognition of TTB by learners within the school context
(Hoffmann, 2013). A limited but increasing number of studies have revealed that
TTB is a serious matter (Santos & Tin, 2018).
2.2. Types of Learner Bullying
A systematic literature review has revealed that the most predominant types of
TTB of teachers are verbal bullying, emotional bullying, physical bullying,
cyberbullying, sexually orientated, insistent class disturbance, intimidating and
threatening behaviour, and personal property offenses (Garrett, 2014).
2.2.1 Verbal Bullying
Verbal bullying is an act of mistreatment by virtue of either spoken or written
words (de Wet & Jacobs, 2018). This is a form of direct bullying which includes
making threats, teasing, name-calling, improper sexual remarks, taunting,
making fun of a person, spreading rumours, insults, sarcastic comments,
discriminatory and refusing to talk to someone (Kõiv & Aia-Utsal, 2021). In most
cases, if no action is taken to resolve verbal bullying, it can grow into physical
bullying (de Wet & Jacobs, 2018). Studies in teacher-targeted bullying have
found verbal bullying to be the foremost type of bullying learners use on
teachers (Woudstra et al., 2018; Billett et al., 2019). This type of bullying is
experienced mostly by females (Bayer et al., 2018).
2.2.2 Psychological Bullying
Psychological bullying refers to harming an individual through emotional abuse,
therefore causing significant stress and interfering with an individual’s
capability to develop healthy and stable patterns of relating to other people
(Hlophe, Morojele & Motsa, 2017). It is often conducted by an individual or a
group of individuals, who repeatedly and intentionally use words or actions
which cause psychological harm to another individual (Antiri, 2016).
This is a form of indirect bullying which includes actions like intentionally
leaving somebody out, excluding somebody from a group or activities,
embarrassing somebody in public, and spreading rumours as well as gossiping
about someone (Smith & Thompson, 2017). Hlophe et al. (2017) stated that
stealing and damaging other individuals’ belongings amounts to indirect

23
bullying. Intimidation, manipulation and stalking of an individual are also part
of psychological bullying (Antiri, 2016). This type of bullying often happens
amongst close individuals, and impacts one’s self-esteem (Hlophe et al., 2017).
The perpetrator of this bullying stokes an individual up emotionally to try to
make him or her uncomfortable, disturbed and mentally destabilised (Antiri,
2016).
2.2.3 Physical Bullying
Physical bullying is any activity conducted by a bully to inflict physical harm on
a victim (Thompson, 2019). The bully must physically make contact with the
victim so that this act can happen (Potocki, 2015). This is a direct form of
bullying which includes hitting, kicking, pinching, spitting, tripping, pushing,
taking somebody's belongings, breaking or damaging somebody’s belongings,
and making somebody do things they do not want to do (Uz & Bayraktar, 2019).
A perpetrator of physical bullying normally performs the bullying act when the
victim is not aware that bullying is about to happen (Potocki, 2015). Homes,
schools and workplaces are always affected by numerous occurrences of
physical bullying (Antiri, 2016). Learners who commit physical bullying are
more likely to worsen in their fierce behaviour and criminal offences
(Thompson, 2019). Physical bullying is easy for people to detect in school
because of the visibility of physical actions (Thompson, 2019). This type of
bullying is experienced mostly by males (Bayer et al., 2018). A challenge for
schools to maintain a safe and orderly learning environment is presented by
physical bullying, hence it tends to be given more attention from school
employees than other types of bullying (Antiri, 2016).
2.4.4 Cyberbullying
Cyberbullying is any technology-mediated bullying behaviour recognised in
social media, websites and instant messaging which comprises repetitive
behaviour like mailing, posting, messaging, sending images, videos with
abusive content; the deliberate prohibiting of a person in the online space;
spreading false information, and hacking of private accounts such as email
(Palaghia, 2019). It is a type of indirect bullying which favours the perpetrator's
invisible and anonymous state (Navarro et al., 2015). Cyberbullying behaviour is
perpetrated by individuals who experienced or who are experiencing real-world
abuse or strained relationships with others that trigger tension such as hostile
interactions between parents and children (Yamin, Shalahudin, Rosidin, &
Somantri, 2019). Victims of real-world abuse may conduct cyberbullying to
express anger towards the individuals who bullied them in the real world,
hoping they are aware of their mistakes, wanting to overthrow and humiliate
them, feeling hurt and wanting to retaliate, seeking attention and pleasure
(Yamin et al., 2019). Lack of inhibition and social disengagement are significant
factors which may be responsible for online violence (Navarro et al., 2015). The
main reason for teenagers to cyberbully people is because to them, it is a way of
joking. They may also want revenge, because they are angry, or hate the person
they direct the bullying towards (Yamin et al., 2019). Teachers are not often
cyberbullied, but the few reported incidents have damaging effects on those

24
who have experienced them Cyberbullying is a tremendously harmful
psychosocial phenomenon in constructive school life (Navarro et al., 2015).
2.5. Influences of Learner Bullying on Teachers
Bullying has a destructive influence on its victims. Sometimes it results in harm
for the bullying perpetrators as well (Siregar et al., 2019). Regardless of the type
of bullying that an individual has experienced, they all tend to have a similar
influence on people. Bullying is hurtful towards its victims and may cause
health problems for some victims; some might end up avoiding the places where
the bullying took place in order to minimise the chances of being bullied again.
This in turn disrupts their lives as it limits some of the basic human rights
enshrined in the Bill of Rights of the Constitution of the Republic of South
Africa. These include freedom of movement, freedom of association and the
right to a healthy environment (RSA, 1996).
2.6 Strategies to Assist Teachers
South Africa is currently facing the challenge of protecting teachers and creating
sufficient resources to abolish learner-to-teacher bullying so that teachers'
mental health will be improved (Woudstra, 2015). In South Africa, there are
currently limited laws in place that strictly protect teachers from learners'
bullying them at work. However, the Constitution of the Republic of South
Africa (RSA, 1996), legislation and common law, provide a legal framework in
relation to TTB (Jacobs & de Wet, 2018). Learners should obey the laws of the
country, which state that teachers have rights too, and need to be treated with
respect and dignity within their workplace (Jacobs & de Wet, 2018).
3. Research Methodology
The study used a qualitative research design to explore the influence of learners’
bullying on teachers. An interpretive paradigm was used to gain insights into
TTB.
3.1. Participants
The researchers used a total of six participants from three selected high schools
(two participants per school) in Sikhulile Circuit in the Ehlanzeni District,
Mpumalanga Province. Purposive sampling, which falls under the non-
probability sampling method, was applied in this study.
Table 1: Participants’ biographical information
Participants Age Female Male South
African and
SiSwati-
speaking
Teaching
experience
(in years)
Educational
qualifications
Subjects
Participant 1 45 ✓ ✓ 19 Standard
Teacher
Diploma (S.T.D)
Life
Orientation
Participant 2 31 ✓ ✓ 8 Bachelor of
Education (FET)
& BA Honours
English (First
Additional
Language)
Education (FET)
Mathematical
Literacy

25
Education (FET)
Economic and
Management
Science
Participant 5 46 ✓ ✓ 24 Honours in
Education
Geography
Education (FET)
Geography
The researchers used the following criteria for participants’ inclusion in the
study:
1. Participants had to express a willingness to be interviewed.
2. Participants had to be employed as teachers at the time of the study and be
teachers by profession.
3. Participants had to be teaching in one or other of the three selected schools.
4. Participants had to perceive themselves as teachers who had experienced
learners’ bullying.
3.2. Data collection instrument
The research questionnaire included two sections, A and B. Section A required
biographical details from the participants, and they had to fill in this section by
hand. The biographical details included information on age, gender,
qualification, number of years as a teacher and subjects they teach. Section B was
the interview schedule, which had eight sets of questions intended to yield
qualitative data on learners’ bullying of teachers. Semi-structured interviews
were used to obtain data in this study and participants’ responses were recorded
using an audio recorder. The research instrument’s reliability and validity were
ensured through the use of an audio recorder to record the interviews; notes
were taken to supplement the data recorded. The interview schedule (see
Appendices 1 & 2) was provided in two versions (English and SiSwati) for
participants to choose the language with which they were comfortable. It
contained a set of open-ended and closed-ended questions. All the questions
were based on the research objectives. The open-ended questions required in-
depth responses from participants. This semi-structured interview guide is a
diagrammatic demonstration of questions or topics, which allowed the
interviewers to explore further if necessary (Jamshed, 2014). The researchers
thus gathered qualitative data which answered the research objectives and
questions. All the participants were given the same interview schedule, and they
answered the same set of questions.
3.3. Data collection procedures
The data was collected over a period of three days, one day for each school; two
participants were interviewed per day. Before each interview, a briefing session
was held to discuss the purpose of the study, allowing participants to make an
informed decision regarding their participation. After the briefing session,
participants signed the informed consent for participation in the study and for
the audio recorded notes. Each interview was conducted in a private space so
that the participants would feel comfortable and free to share. A 45-minute
session was allocated for each participant and an audio recorder was used to
capture the information. The participants had a copy of the interview schedule

26
to peruse while questions were asked in real-time, in either English or SiSwati
(the participant’s choice). During the interviews, the researchers asked the
participants to clarify vague responses (if the participant’s response was not
clear) and asked probing questions to obtain more data. The researchers were
cautious because bullying is a sensitive issue and may create emotional distress.
Open-ended questions allowed the teachers to elaborate freely on the questions
that were posed.
3.4. Data analysis procedure
Thematic data analysis was used to summarise information gathered from
participants regarding learners’ bullying. According to Javadi and Zarea (2016),
thematic data analysis is used to extract meaning and concepts from gathered
data, which can be in the form of interview transcripts, notes, documents,
pictures or videos, and it comprises pinpointing, examining and recording
certain patterns or themes.
Stage 1: Getting Familiar with the Data. The researchers listened to the
participants’ recorded audios to be familiar with the data they had provided. All
the participants were SiSwati speakers, so they answered some questions in their
home language, and others in English. Translation of the information provided
in SiSwati was conducted by the SiSwati speaking researcher.
Stage 2: Transcription. The researchers transcribed the verbal data gathered from
the participants’ interviews in a sequential form, starting from the participant
that was interviewed first and ending with the one that was interviewed last. All
the transcribed data were compiled to form one transcript.
Stage 3: Finding Meaningful Data. The researchers read the transcript which
contained the transcribed data from all six participants, and highlighted
sentences, phrases, or paragraphs that seemed to be meaningful and relevant to
the research questions.
Stage 4: Reviewing the Highlighted Information. The researchers went back to
review the highlighted data on the transcript to ascertain if it really linked to the
research questions. Irrelevant data were eliminated and stored in a separate
document.
Stage 5: Developing Patterns. The researchers named each set of data, and the
information that was connected was grouped together to form patterns, which
were summed up using a phrase (Percy et al., 2015). The patterns were then
gathered and identified as sub-emerging themes which were related to the
research questions, and were coded (Hlophe, Morojele, & Motsa, 2017).
Stage 6: Naming Themes and Producing the Report. From the participants’
interviews, open-ended questions were arranged into themes. For each theme, a
comprehensive abstract analysis clarifying the scope and component of the
theme was written. This procedure was conducted for each participant’s data
(Percy et al., 2015).

27
3.5. Ethical considerations
The prescribed ethical principles were followed and all policy and guideline
documents that were provided by the University of Zululand, HPCSA and other
regulatory bodies, which stipulated ethical practices, were taken into
consideration when conducting research. These prohibit the researchers from
plagiarism, harm or violation of the rights of others who are directly and
indirectly involved in the research study.
a. Permission to conduct research
Permission to conduct the study was obtained from the Research Committee of
the University of Zululand, Faculty of Education and Department of Education,
Mpumalanga. The researchers wrote a letter which clearly outlined the study
details and objectives to the Head of the Mpumalanga Department of Education,
Ehlanzeni District, as well as the selected school principals, requesting
permission to conduct the study. The Head of the Mpumalanga Department of
Education, Ehlanzeni District, and the school principals provided the researchers
with written permission to conduct research at the schools. Informed consent
forms were signed by the participants as proof of willingness to participate in
the study and to record the interviews. They were assured that confidentiality
and anonymity of records would be maintained.
b. Informed consent
Information concerning participation, as well as information that might have
reasonably been expected to affect their willingness to participate in the study,
was distributed to all participants. The researchers provided informed consent
forms to participants in a language they clearly understood, and made them
aware that participation was completely voluntary, and that they were free to
withdraw from the study at any time should they so wish. The researchers made
sure that participants who were involved in the study all had the capacity to
consent.
c. Privacy, confidentiality and anonymity
The researchers did not record participants’ names or any private, identifying
details in the semi-structured interviews. Pseudonyms were used to protect the
information obtained from the participants. For confidentiality, the researchers
ensured that information collected from participants was not disclosed to any
individual who was not involved in the study, and stored information obtained
from the participants in a safe location. Questions that were sensitive or
embarrassing to the participants were avoided by the researchers (a senior
psychologist and student psychologist), who made participants aware that
should they feel emotional discomfort at any time, they could discuss this with
the researchers.
4. Results
4.1. Types of bullying behaviour learners use against teachers
a. Teachers’ experiences of teacher-targeted bullying
This study revealed that teachers experienced learners’ bullying behaviour in
different ways and some encountered repeated bullying experiences. Teachers

ILJTER.ORG Volume 22 Number 06 June 2023

ILJTER.ORG Volume 22 Number 06 June 2023

Recommended

Recommended

More Related Content

Similar to ILJTER.ORG Volume 22 Number 06 June 2023

Similar to ILJTER.ORG Volume 22 Number 06 June 2023 (20)

More from ijlterorg

More from ijlterorg (8)

Recently uploaded

Recently uploaded (20)

ILJTER.ORG Volume 22 Number 06 June 2023