Conversational Agent Framework in Mathematics Education

Tharsiniy Ramasamy

, Choo Kim Tan

, Choo Peng Tan

and Ah Choo Koo

Faculty of Information Science and Technology, Multimedia University, Melaka, 75450, Malaysia

Facult

alaysia

tharsiniyramasamy@gmail.com, {cktan, cptan, ackoo}@mmu.edu.my

Keywords: Attitude Towards Mathematics, Conversational Agent, Traditional Teaching, Technology in Education,

Mathematics Anxiety.

Abstract: Students' negative attitude toward mathematics is always associated with a decline in math performance. This

often discourages students from pursuing STEM fields and careers. Efforts such as incorporating technology

in educational settings to alleviate math anxiety by boosting students’ motivation to explore and appreciate

the subject were made. Conversational agents present an opportunity to enhance educational support for

teaching and learning. However, despite almost six decades of development, the use of conversational agents

in education remains limited. The impact of integrating a conversational agent on students' attitudes toward

mathematics is essential to be investigated. In this study, an experimental research approach was adopted,

involving undergraduate students at a Malaysian university. The control group received only traditional

classroom instruction, while the experimental group received traditional instruction supplemented by

interaction with a conversational agent. Following the intervention, participants completed a questionnaire

assessing their attitudes toward mathematics. This paper aims to present the research design and framework

of a conversational agent in mathematics education.

1 INTRODUCTION

Mathematics education is a cornerstone for

developing essential analytical skills that benefit

students across academic disciplines and professional

paths (Boaler, 2016). Beyond fulfilling academic

requirements, mathematics promotes logical

reasoning and problem-solving skills that are critical

in fields ranging from science and engineering to

economics and technology. Despite its importance,

many students find mathematics challenging, leading

to issues such as high dropout rates, anxiety, and

negative attitudes toward the subject. These struggles

can discourage students from pursuing math-

intensive careers and contribute to the gender and

skills gap in STEM fields (Szczygiel & Perez, 2021;

Ashcraft & Moore, 2009). Math anxiety, which

affects students at various educational levels, can

severely impair performance by hindering cognitive

processing and reducing motivation (Dowker, Sarkar,

& Looi, 2016). Addressing these challenges requires

understanding and intervention at both the

instructional and emotional levels.

Interactive teaching strategies, such as open-

ended discussions and personalized feedback, have

been shown to support learning and engagement.

However, large class sizes and time constraints can

limit opportunities for individualized interactions

(Hattie & Timperley, 2007). Additionally, students

who are more reserved may hesitate to participate in

class discussions, preferring more private modes of

communication (Gleason, 2020). This dynamic can

lead to disengagement and lower performance among

students who need additional support, emphasizing

the importance of adaptable, student-centred

approaches in mathematics education.

Research highlights the potential of technology in

addressing these needs. Integrating digital tools such

as interactive software and conversational agents has

been shown to engage students and create a more

personalized learning experience. Studies indicate

that technology can increase student motivation,

facilitate instant feedback, and provide a safe

environment for practice, all of which help to reduce

math-related anxiety (Cheung & Slavin, 2013).

Conversational agents, in particular, present a

promising solution by enabling interactive learning

experiences where students can ask questions and

receive real-time support. These agents also promote

a more relaxed learning environment, allowing

Ramasamy, T., Tan, K. C., Tan, C. P. and Koo, A. C.

Conversational Agent Framework in Mathematics Education.

DOI: 10.5220/0013340600004557

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 4th International Conference on Creative Multimedia (ICCM 2024), pages 61-67

ISBN: 978-989-758-733-7; ISSN: 3051-6412

students to practice complex problem-solving

without fear of judgment, thereby enhancing

confidence and enjoyment in mathematics (D’Mello

& Graesser, 2013).

Students’ attitudes toward mathematics evolve

based on their experiences of success and failure.

Positive experiences build confidence and a

favourable outlook on learning, while negative

experiences can foster anxiety and disinterest. This

balance of emotions underscores the importance of

creating a nurturing and adaptive learning

environment that celebrates student progress and

addresses setbacks with constructive feedback

(Pekrun, 2020).

The current study explores the role of

conversational agents in mathematics education,

particularly their potential to positively impact

students' attitudes toward the subject. With the

advancement of technology, conversational agents

provide an opportunity for a more engaging and

supportive educational experience. By integrating

these tools, educational institutions can contribute to

an inclusive and effective learning environment that

supports students' academic journeys and prepares

them for success in math-related fields.

2 RELATED WORKS

This section presents literature review related to the

study such as the attitude towards mathematics and

conversational agents.

2.1 Attitude

Attitude refers to an individual's emotional, cognitive,

and behavioural responses to a specific situation or

object (Rosenberg & Hovland, 2022). Like other

academic subjects, mathematics can evoke various

emotions, such as enjoyment, frustration, anxiety, or

admiration. Research indicates that a positive attitude

towards mathematics is often linked to better

academic outcomes, while a negative attitude can

hinder performance. Consequently, fostering a

positive attitude toward mathematics is crucial, as it

can significantly influence a student's depth of

understanding and willingness to engage with the

subject (Ramirez et al., 2022).

Students encounter various challenges in

mathematics, such as difficulties with conceptual

understanding, problem-solving, and motivation.

These challenges often lead to feelings of anxiety or

inadequacy, collectively known as mathematics

anxiety. This anxiety is commonly rooted in negative

or discouraging past experiences, which can shape

students' beliefs about their mathematical abilities

and impact their willingness to persist in the subject

(Dowker, Sarkar, & Looi, 2016). A recent study by

Fletcher and Dowker (2023) found that students who

perceive mathematics as difficult and disconnected

from real-life applications often hold negative views

of the subject. Furthermore, student attitudes are

influenced by their interactions with teachers.

Educators who display empathy, encouragement, and

constructive feedback can foster a more positive

learning environment, whereas those who appear

overly critical may inadvertently contribute to student

anxiety and avoidance of the subject (Beilock &

Maloney, 2022).

Perceptions of mathematics are complex,

encompassing cognitive, emotional, and behavioral

components. Researchers have emphasized the need

to adopt a multidimensional approach to better

understand these perceptions (Kolar & Hodis, 2021).

For example, in a study of European middle school

students, three primary dimensions were identified in

students' perceptions of mathematics: personal beliefs

about mathematical competence, emotional reactions

to the subject, and attitudes towards the utility of

mathematics in everyday life (Lucangeli, Galli, &

Mammarella, 2022). Furthermore, the Trends in

International Mathematics and Science Study

(TIMSS) has shown that student attitudes toward

mathematics include enthusiasm, confidence, and

perceived relevance, all of which correlate with

performance and engagement (Mullis, Martin, Foy, &

Hooper, 2020).

Measuring attitudes towards mathematics is

critical in educational research, as it helps identify

factors influencing students' engagement and

performance in the subject. Various methods have

been employed for this purpose, including self-report

measures, behavioral observations, psychological

assessments, and implicit association tests (Eagly &

Chaiken, 2022). Self-report measures are among the

most used to gauge levels of agreement or

disagreement with statements related to mathematics.

These tools provide accessible ways to capture

students’ attitudes but may be influenced by social

desirability biases. Behavioral measures, which

focus on observable actions like the time spent on

mathematics tasks or the willingness to seek

additional practice opportunities, provide an

alternative view of student engagement. These

measures are beneficial in assessing attitudes through

action rather than self-report, though they may be

limited in capturing internalized attitudes toward

mathematics.

ICCM 2024 - The International Conference on Creative Multimedia

Psychological measures, such as monitoring heart

rate, cortisol levels, or skin conductance, offer

insights into the physiological responses associated

with mathematics-related anxiety or excitement. Such

physiological indicators are valuable in cases where

students might find it difficult to articulate their

feelings toward the subject, as they reveal underlying

emotions that may not be overtly expressed.

Additionally, implicit measures, such as the Implicit

Association Test (IAT), have been used to assess the

subconscious associations individuals hold toward

mathematics, particularly in distinguishing between

positive and negative attitudes (Greenwald et al.,

2021). This method helps address potential biases

found in self-reported data, providing a deeper

understanding of implicit attitudes.

Attitudes are multi-dimensional and could not be

observed directly and easily, thus the task of

measuring students' attitudes toward mathematics is

complex. The reliability and validity of the tools used

are often questioned, as attitudes are inherently

subjective and may be challenging to quantify

accurately (Peterson & Flanders, 2019). Traditional

methods, which often posit a link between positive

attitudes and mathematical success, rely on

measuring attitudes as a critical component in

understanding and improving student performance.

To this end, a variety of instruments have been

developed, each focusing on different attitude

dimensions. For example, Wong and Wong (2022)

created an instrument that assessed students' attitudes

toward math based on enjoyment, motivation, and

confidence, and Stroet et al. (2021) designed a scale

emphasizing self-efficacy and interest, while Tapia

and Marsh (2002) developed a four-factor

questionnaire. Despite the challenges, many studies

continue to utilize Likert scale-based self-report

questionnaires to assess students’ attitudes toward

mathematics, as these remain the most practical for

capturing diverse attitude dimensions in educational

settings.

2.2 Conversational Agents

Conversational agents (CAs), also known as chatbots

or intelligent tutoring systems, are software programs

designed to simulate human conversation, allowing

users to interact using natural language. These

systems function as dialogue interfaces capable of

understanding and responding to user inputs in ways

that resemble human communication (Shawar &

Atwell, 2021). CAs are typically employed in

applications such as chatbots for customer service or

virtual assistants on mobile devices. They utilize

computational linguistics, allowing them to process

user queries and generate contextually appropriate,

human-like responses.

At the core of CAs are technologies like natural

language processing (NLP) and machine learning

(ML). NLP allows the agent to interpret the meaning

behind human language, enabling the bot to engage in

more sophisticated dialogue. Meanwhile, ML

empowers the CA to learn from each interaction,

improving its ability to tailor responses and

understand user preferences over time. While

chatbots have been around since the 1960s, recent

advancements in artificial intelligence, NLP, and ML

have expanded their potential, increasing their

integration into various sectors, including education

(Wu et al., 2022).

In the context of education, the role of

conversational agents is an important consideration.

Educators can deploy CAs in diverse roles, such as

tutors, mentors, or peer collaborators, based on the

instructional goals and learning environments. For

instance, some researchers have explored how CAs

function as personalized assistants in virtual

classrooms, adapting their support to the individual

needs of students (Jou & Huang, 2021). Additionally,

studies have highlighted that CAs can be designed to

facilitate interactive learning, enhancing the

educational experience by engaging students in

dynamic problem-solving activities (Yin et al., 2022).

One recent study by Abdullah et al. (2024)

examined the use of CAs as virtual tutors in higher

education. Their findings suggested that CAs, when

used as part of an interactive learning platform,

improved students' understanding of complex

concepts by providing immediate feedback and

personalized support. The study further emphasized

the advantages of CAs over traditional e-learning

tools, noting that their ability to respond in real-time

to students’ queries fosters a more engaging,

responsive, and individualized learning experience

(Zhou & Wang, 2023).

Moreover, CAs are found to be particularly

effective in non-technical fields where students may

struggle to grasp complex material without additional

guidance. For example, a recent investigation by

Chen et al. (2023) into CAs in humanities courses

highlighted their potential to assist non-technical

students by guiding them through difficult concepts

and ensuring timely interventions when students face

challenges. This adaptability positions CAs as a

powerful tool for expanding access to personalized

education and offering support that complements

traditional teaching methods.

Conversational Agent Framework in Mathematics Education

The application of conversational agents (CAs)

has been widespread in various industries, but their

integration within educational settings, particularly

mathematics education, remains relatively

underexplored. Despite growing interest in CAs, their

adoption in mathematics education is still nascent, as

highlighted by recent studies (Guszcza, Smetana, &

Waguespack, 2020; Zhang & Choi, 2023). This gap

presents a valuable opportunity for further research

and development to enhance the integration of

conversational agents as pedagogical tools in

mathematics learning.

CAs are proving to be beneficial in education by

providing immediate feedback, responding to student

queries, and offering personalized support throughout

the learning process (Wang, Yu, & Yang, 2023).

Moreover, they contribute to enhancing educational

efficiency by automating routine tasks, streamlining

access to learning materials, and saving valuable time

for both educators and students (Patel et al., 2023).

The use of CAs also optimizes learning outcomes by

ensuring that students receive individualized

attention, which is particularly important in large

classrooms or online learning environments.

Incorporating challenging questions into CAs has

been found to improve student confidence and foster

a sense of accomplishment (Xiao et al., 2023). As

students tackle more complex problems, they gain a

better understanding of their progress, which not only

enhances their problem-solving skills but also builds

their self-assurance in overcoming mathematical

obstacles. These incremental successes encourage

students to continue engaging with challenging

content, promoting a positive cycle of learning and

growth.

Additionally, CAs also enhance student

motivation and engagement by offering greater

control over their learning experience. Studies have

shown that interactive tools such as CAs can sustain

students' interest and participation, creating a more

enjoyable and supportive learning environment (Dale

& Choi, 2021; Lee et al., 2022). This personalized

approach helps students to absorb knowledge more

effectively and reduces feelings of frustration or

boredom. Furthermore, CAs’ ability to provide

instant feedback is crucial for improving both

academic performance and motivation. By allowing

students to receive immediate corrective feedback,

CAs help them rectify mistakes and solidify their

understanding of complex mathematical concepts

(Wang & Zhang, 2021).

3 METHODOLOGY

This section will discuss the design, sample, materials

and instrument used in the study as well as the

research procedure to provide a lucid picture.

3.1 Design

This study employed a quasi-experimental design

with convenience sampling. Due to limitations in

scheduling and venue arrangements, as well as the

desire to maintain classroom norms, pre-existing

groups were used instead of randomly assigning

students to experimental and control groups. Within

these groups, students who were easily accessible and

willing to participate were selected. A consent form

was distributed to all students, and only those who

consented to participate were included in the study.

3.2 Sample

A total of 200 undergraduates who were voluntarily

participated in the study. The sample consisted of

both male and female students from various ethnic

backgrounds such as Malay, Chinese, Indian, and

others, as well as a mix of local and international

students. The experimental group comprised 115

students, while the control group consisted of 85

students. Students were asked to voluntarily select

their group—either experimental or control—

resulting in unequal sample sizes. A quantitative

research methodology was employed to collect and

analyze data on students’ attitudes toward

mathematics.

3.3 Conversational Agent and Material

A conversational agent, integrated into Facebook

Messenger, was developed for educational purposes,

functioning as a tutor in this study. This agent was

learning, experiential learning, social dialogue, and

scaffolding. It broke down complex learning material

into manageable, bite-sized segments, making it

easier for students to grasp concepts quickly without

feeling overwhelmed. Motivation in the form of

inspirational messages was given to the students by

the agent. Constructive feedback and motivational

quotes were integrated to bolster students'

confidence. The agent assisted students stay focused

during their problem-solving process. Alongside

clear explanations, examples, and exercises, the agent

provided step-by-step solutions, reinforcing learning.

Furthermore, it allowed students to attempt questions

ICCM 2024 - The International Conference on Creative Multimedia

multiple times, regardless of previous mistakes,

fostering a sense of growth and resilience in their

learning journey.

3.4 Mathematics Attitude Dimension

The objective of this project is to assess the influence

of integrating a conversational agent on students’

attitudes toward mathematics, thus the instrument,

Mathematics Attitude Dimension, was developed. It

was a questionnaire which comprised a combination

of adopted, adapted, and self-designed items. This

instrument included 11 questions and served as pre-

test and post-test, focused on students’ perspectives

regarding their attitude towards learning the

mathematics topic ‘Integration,’ using a 5-point

Likert scale. “The importance of mathematics” and

“anxiety towards mathematics” were two dimensions

focused on this instrument, and were adapted from the

four-factor questionnaire developed by Tapia and

Marsh (2002).

Before the study, participants were asked to sign

a consent form, confirming their voluntary

participation in the study. They were then asked to

provide confidential general information such as age,

gender, race, and nationality. The questionnaire was

thoroughly reviewed and validated by experienced

mathematics lecturers from the private university to

ensure the quality and relevance of the questions for

the study. To assess the reliability of the instrument,

Cronbach’s Alpha was calculated using SPSS,

yielding a value of 0.921, which indicates excellent

reliability according to Amirrudin et al. (2021). This

high Cronbach’s Alpha not only confirms the

instrument’s reliability but also demonstrates its

internal consistency, ensuring its effectiveness in

measuring each variable. Table 1 presents an example

of the questions used in the mathematics attitude

dimension questionnaire.

Table 1: Example of Questions in Mathematics Attitude

Dimension.

Dimensions

Item

Importance of

mathematics

Mathematics helps to develop my

thinking skills.

Anxiety towards

mathematics

To even consider having to

complete a math problem makes

me anxious.

3.5 Procedure

The field study was conducted in a period of 5-weeks

with three distinct phases. Before the first phase, the

students were briefed by the researcher on the study's

objectives, and they were asked to sign a consent

form indicating their voluntary participation.

Participation was entirely optional. Those students

who agreed to participate in the study were then asked

to participate in either the control or experimental

groups voluntarily. The first phase involved

administering a pre-test to both control and

experimental groups’ students before they were

introduced to the topic of 'Integration' by the

university lecturer.

In the second phase, a conversational agent was

incorporated into the teaching and learning process

for the mathematics syllabus for the experimental

group. Students in the experimental group were

briefed on the purpose of the conversational agent

before using it for learning. While both groups

received traditional instruction on the 'Integration'

topic from the same lecturer, only the experimental

group had access to the conversational agent. This

intervention lasted for 3 weeks, during which the

experimental group used the conversational agent

independently, at their own pace, and at times and

locations that were most convenient for them outside

of their regular class schedule.

The administration of the post-test was conducted

on both groups in the last phase after the completion

of the 'Integration' topic. The post-test assessed the

mathematics attitude dimension to determine the

effectiveness of the conversational agent in

improving students’ attitudes toward mathematics.

4 CONCLUSION

This paper has outlined a conversational agent

framework, alongside a research design employing a

quasi-experimental with convenience sampling

method approach. The design utilizes a mathematics

attitude dimension instrument. Reliability and

validity tests have shown that the instrument is

reliable and valid, ensuring that the subsequent data

collection process will yield meaningful and valuable

interpretations. As the project is still ongoing and data

analysis is in progress, no finding is available at this

stage to conclude the study. It is hoped that

favourable findings will be obtained and will be

significant to many parties such as educators,

students, curriculum designers, etc.

Conversational Agent Framework in Mathematics Education

ACKNOWLEDGEMENTS

This research was funded by a grant from Ministry of

Higher Education of Malaysia

(FRGS/1/2020/SSI0/MMU/02/6). We would like to

acknowledge and thank Ministry of Higher Education

of Malaysia for the fund, the university and the

lecturer for permitting the data collection and all

respondents who participated in this research.

REFERENCES

Abdullah, M., Ahmed, H., & Saleh, M. (2024).

Conversational agents as virtual tutors in higher

education: Improving student engagement and

understanding. Journal of Educational Technology,

45(3), 112-125.

Amirrudin, M., Nasution, K., & Supahar, S. (2021). Effect

of variability on Cronbach alpha reliability in research

practice. Jurnal Matematika, Statistika dan Komputasi,

17(2), 223-230.

Ashcraft, M. H., & Moore, A. M. (2009). Mathematics

anxiety and the affective drop in performance. Journal

of Psychoeducational Assessment, 27(3), 197–205.

Beilock, S. L., & Maloney, E. A. (2022). Understanding

and addressing the math anxiety of students and

teachers. Annual Review of Psychology, 73, 583–611.

Boaler, J. (2016). Mathematical mindsets: Unleashing

students' potential through creative math, inspiring

messages, and innovative teaching. Jossey-Bass.

Chen, L., Zhang, L., & Yu, H. (2023). The role of

conversational agents in facilitating humanities

education: Supporting non-technical students. Journal

of Educational Research and Practice, 10(2), 89-103.

Cheung, A. C. K., & Slavin, R. E. (2013). The effectiveness

of educational technology applications for enhancing

mathematics achievement in K-12 classrooms: A meta-

analysis. Educational Research Review, 9, 88–113.

D’Mello, S., & Graesser, A. C. (2013). AutoTutor and

affective AutoTutor: Learning by talking with

cognitively and emotionally intelligent computers that

talk back. ACM Transactions on Interactive Intelligent

Systems (TiiS), 2(4), 1–39.

Dale, E., & Choi, H. (2021). The role of interactive tools in

sustaining student engagement in education.

Educational Technology & Society, 24(2), 45-56.

Dowker, A., Sarkar, A., & Looi, C. Y. (2016). Mathematics

anxiety: What have we learned in 60 years? Frontiers

in Psychology, 7, 508.

Dowker, A., Sarkar, A., & Looi, C. Y. (2016). Mathematics

anxiety: What have we learned in 60 years? Frontiers

in Psychology, 7, 508.

Eagly, A. H., & Chaiken, S. (2022). The psychology of

attitudes. Harcourt Brace Jovanovich College

Publishers.

Fletcher, C., & Dowker, A. (2023). Exploring the links

between mathematics anxiety, attitudes, and

achievement in primary school students. Journal of

Educational Psychology, 115(1), 144–155.

Gleason, N. W. (Ed.). (2020). Higher education in the era

of the fourth industrial revolution. Springer.

Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K.

(2021). Measuring individual differences in implicit

cognition: The implicit association test. Journal of

Personality and Social Psychology, 74(6), 1464-1480.

Guszcza, J., Smetana, L., & Waguespack, M. (2020).

Conversational agents in education: Exploring their

role and impact in mathematics classrooms. Journal of

Educational Technology, 35(1), 12-25.

Hattie, J., & Timperley, H. (2007). The power of feedback.

Review of Educational Research, 77(1), 81–112.

Jou, M., & Huang, H. (2021). The use of conversational

agents as personalized tutors in virtual classrooms.

International Journal of Educational Technology, 41(4),

220-233.

Kolar, M., & Hodis, F. (2021). Developing a

multidimensional perspective on mathematics anxiety:

Implications for research and teaching. Mathematics

Education Research Journal, 33, 553–572.

Lee, Y., Kim, S., & Chang, J. (2022). Enhancing student

engagement and motivation with intelligent tutoring

systems. International Journal of Artificial Intelligence

in Education, 31(3), 111-124.

Lucangeli, D., Galli, F., & Mammarella, I. C. (2022).

Emotional and cognitive factors associated with

mathematics achievement: Implications for assessment

and intervention. Educational Studies in Mathematics,

110(2), 315–331.

Mullis, I. V. S., Martin, M. O., Foy, P., & Hooper, M.

(2020). TIMSS 2019 international results in

mathematics and science. International Association for

the Evaluation of Educational Achievement (IEA).

Patel, R., Kumar, A., & Reddy, S. (2023). The benefits of

automation in education: A study on the integration of

conversational agents. Journal of Educational

Computing Research, 57(6), 755-772.

Pekrun, R. (2020). Emotions and motivation in

mathematics education. In S. Lerman (Ed.),

Encyclopedia of Mathematics Education (pp. 341-348).

Springer.

Peterson, R. A., & Flanders, B. (2019). Quantifying student

attitudes toward mathematics: A meta-analysis. Journal

of Educational Psychology, 111(7), 1323-1334.

Ramirez, G., Shaw, S. T., & Maloney, E. A. (2022).

Improving math attitudes and reducing anxiety through

interventions: A comprehensive review. Learning and

Instruction, 76, 101530.

Rosenberg, M. J., & Hovland, C. I. (2022). Attitude

organization and change: An analysis of consistency

among attitude components. Yale University Press.

Shawar, B. A., & Atwell, E. (2021). A survey of

conversational agents: Chatbots and intelligent

tutoring systems in the age of artificial intelligence.

Computer Applications in Engineering Education,

29(2), 387-402.

Stroet, K., Opdenakker, M. C., & Minnaert, A. (2021).

Effects of need supportive teaching on early

ICCM 2024 - The International Conference on Creative Multimedia

adolescents' motivation and engagement in math.

Journal of Educational Psychology, 113(3), 573-587.

Szczygiel, M., & Perez, S. M. (2021). Understanding the

gender gap in mathematics performance: The role of

math anxiety, self-efficacy, and test anxiety.

Contemporary Educational Psychology, 65, 101959.

Tapia, M., & Marsh II, G. E. (2002). Confirmatory factor

analysis of the attitudes towards mathematics inventory.

Wang, L., Yu, Z., & Yang, X. (2023). The effectiveness of

conversational agents in providing personalized

educational support. Education and Information

Technologies, 28(4), 1579-1592.

Wang, Y., & Zhang, Y. (2021). The role of feedback in

improving academic performance with conversational

agents. Journal of Educational Psychology, 113(5),

823-834.

Wong, P. S., & Wong, C. W. (2022). Assessing attitudes

toward mathematics: Development of a

multidimensional scale. International Journal of

Educational Psychology, 11(4), 312-327.

Wu, X., Li, Y., & Sun, P. (2022). Artificial intelligence in

education: The role of conversational agents in

learning environments. Journal of Educational

Technology Development and Exchange, 15(1), 75-91.

Xiao, Z., Chen, H., & Li, F. (2023). Enhancing student

confidence in mathematics through conversational

agents. Journal of Mathematics Education, 46(2), 231-

245.

Yin, X., Zhang, F., & Liu, M. (2022). Interactive learning

with conversational agents: Enhancing problem-

solving skills in education. Journal of Educational

Computing Research, 59(7), 1123-1140.

Zhang, J., & Choi, B. (2023). The adoption of

conversational agents in mathematics education: A

review of recent trends. Educational Research Review,

14(1), 1-14.

Zhou, L., & Wang, Z. (2023). The advantages of

conversational agents over traditional e-learning tools

in fostering student engagement. Journal of Interactive

Learning Research, 34(2), 163-178.

Conversational Agent Framework in Mathematics Education