DOI Number: 10.5958/0976-5506.2018.01058.

How to Use Artificial Intelligence in the English Language

Learning Classroom

Myeong-Hee Shin
Department of General English, Talmage Liberal Arts College, Hannam University, 70 Hannam-ro,
Daedeok-gu, Daejeon, 34430, Republic of Korea

Abstract
Nowadays, it’s expected that the Fourth Industrial Revolution will bring crisis to English education. This
means we don’t need any English teachers in the classroom, nor do we need any English education, if
we have a machine for English instruction. In Korea, English education has been trying to change from
education styles based on cramming and memorization to ones based on communication. However, those
people who blamed the cramming and memorizing learning methods for a limited ability to communicate in
English are going to embrace the machine itself with the fourth revolution. The purpose of communication
is to convey a message. However, it can be delivered very differently depending on speakers’ emotions
and feelings which machines do not possess. Therefore, the Fourth Revolution does not have to replace
Downloaded From IP - 83.122.38.10 on dated 17-Oct-2024

English education. Instead, it should work in cooperation with English education. The purpose of this study
Members Copy, Not for Commercial Sale

is to develop an English class model that uses Artificial Intelligence through Flipped Learning. To do this,
www.IndianJournals.com

English class modules were introduced in blended classes with Flipped Learning class procedures. The
model of Cubillos and Ilvento 12 was used to measure self-efficacy, along with a checklist asking about
academic self-efficacy, learning motivation, and class satisfaction. Each item consisted of five Likert scales.
Questions were answered using the Likert expression 5 point scale, 1 point for ‘not at all’, 2 points for
‘not’, 3 points for ‘normal’, 4 points for ‘yes’ and 5 points for ‘very’, respectively. To find out students’
language achievement, a paired t test was used. The reliability of the test used in the study was tested through
Cronbach’s α. The reliability of the items was analyzed over .70.
As a result of the study, it was found that an English class model using Artificial Intelligence through Flipped
Learning has a positive impact on students’ self-efficacy. There was a statistically significant difference
(p<.05) between the experimental group and the control group in their speaking ability according to teaching
methods. Also, in case of academic achievement, there was a significant difference between the two groups
in listening and speaking achievement according to teaching methods. In terms of academic self-efficacy,
task difficulty (M = 2.95) did not reach the average (M=3.18). For motivation, external motivation was
higher (M = 3.89) than intrinsic motivation (M = 3.20), and in case of class satisfaction, evaluation and
feedback (M = 3.26) stayed in the lowest range. Evaluation and feedback are somewhat new to the students.
This study may be worthwhile in that it gives educators an opportunity to think about the student-centered
class and merits in presenting an effective English learning model as a new learning strategy. However,
this research has its limitations in that there is some difficulty in generalizing the results because in the
present study, the academic self-efficacy and learning motivation on academic achievement was statistically
significant, but the influence was very different from the theoretical background of the previous studies. Also
the small size of the subjects may also have influenced the result of this study, so subsequent studies will
require more comprehensive research including these factors.

Keywords: The Fourth Industrial Revolution, Artificial Intelligence, Flipped Learning, Drawn from title,
Word representing the work.
 558 Indian Journal of Public Health Research & Development, September 2018, Vol.9, No. 9

Introduction in the classroom is still a relatively new concept, but

one that many researchers and educators are exploring6.
In the era of highly developing technology, the Flipped Learning has recently received a lot of attention
artificial intelligence revolution is predicted to be the in the various fields of school classroom environments.
fourth revolution of human history after the information
revolution. Thanks to high-speed information networks, The application of Flipped Learning in the classroom
the rapid development of new technologies has greatly has shown to have a positive educational effect on a
changed our systems of education, society, culture and variety of areas such as academic achievement, attitude,
industry1. Then what is artificial intelligence? It is to collaborative learning, and self-directed learning7. Also
allow computers (artifacts) to imitate human intelligence. an English class model through Flipped Learning can
Artificial Intelligence can be seen as something outside help students improve their metacognition, which is to
of reality and may seem like something of the future2. know how to think or to control the cognitive process11.
However, it is more and more widespread in our everyday Also students’ spontaneous participation in class through
lives. Whether we know it or not, most of us already use flipped learning is likely to promote positive self-efficacy.
AI every day from simple tasks, such as taking a picture, Self-efficacy is a belief in the ability to achieve a task8.
parking our car, asking directions, or using our personal
home care systems to turn on or off the heater, or the lights. However, a variety of limitations in applying Flipped
Learning in school environments, especially in English
Will the Fourth Industrial Revolution bring a crisis to classroom environments, has also been the topic of
English education, as people suspect? It may mean that we much discussion9. As a way of resolving the limitations
don’t need any English teachers in the classroom, or we of applying Flipped Learning to English teaching, it
Downloaded From IP - 83.122.38.10 on dated 17-Oct-2024

don’t need any English education, if we have a machine is suggested that a combination of Movie and Flipped
for English. In Korea, English education has been trying Learning be introduced into English lessons10.
Members Copy, Not for Commercial Sale

to change its education style, which is currently based

www.IndianJournals.com

on cramming and memorization, to communication- The purpose of this study is to develop an English
oriented education3. Meanwhile, it is expected that with class model that uses Flipped Learning. To do this, English
the advent of the Fourth Industrial Revolution, Koreans class modules are introduced by blended classes with
will be liberated from the difficulties of learning English. Flipped Learning class procedures. Three steps of English
However, those people who blamed the cramming class procedures are provided: the pre-class, the in-class,
and memorizing learning method for a limited ability and the post-class. Moreover, the in-class procedure is
to communicate in English are going to embrace the divided into three stages that include the opening, central,
machine itself with the fourth revolution4. For example, a and closing stages. In each stage, the contents of class
translator easily translates a sentence, but it does not grasp activities and interactions between teacher-student and
any specific context or the specific nuances, and provides student-student are presented as class modules.
a completely different expression.
The research hypotheses of this study are as follows;
Communication is to convey a message, however, it
can be delivered very differently depending on the speakers’ An English class model using Artificial Intelligence
emotion and feeling, neither of which machines possess. In through Flipped Learning has a positive impact on
other words, language conveys the meaning, expression, students’ self-efficacy.
value and understanding that one culture shares5. Therefore,
the Fourth Industrial Revolution does not replace English An English class model using Artificial Intelligence
education, but rather it should work in cooperation with through Flipped Learning positively influences students’
English education. It shouldn’t be the case that students use English language achievement.
the machine without thinking. Instead, the expressions that There is a correlation between academic self-efficacy
are unknown in the context of the conversation are referenced and academic achievement, and increasing self-efficacy
through the automatic interpreter. The role of English will have a positive effect on academic achievement.
language education, or English teachers, is more important
than ever in leading students to become true communicators
who can actively change and use appropriate expressions in Procedure
consideration of emotions.
Based on the relevant literature, the present study
Our technology has not reached consciousness yet, focused on Korean university students (with different
but it is not used in a creative and innovative way every majors) taking on-campus English language learning
day, including in the classroom. Artificial intelligence courses.
 Indian Journal of Public Health Research & Development, September 2018, Vol.9, No. 9 559

A. Subjects: The subjects of this study were 80 about academic self-efficacy, learning motivation, and
students who took a 3-hour elective English class satisfaction. Each item consisted of five Likert
language learning course at a university in Korea. scales. The format of the question is the Likert expression
Data collection in this study was conducted for 5 point scale, 1 point for ‘not at all’, 2 points for ‘not’,
approximately four months, from March to June 3 points for ‘normal’, 4 points for ‘yes’ and 5 points
2017. First, the researchers explained the purpose for ‘very’, respectively. To find out students’ language
of the study to each student, and meta-cognitive achievement, a paired t test was used. The reliability of
tests were conducted. The academic achievement the test used in the study was tested through Cronbach’s
was based on the final exam scores. α. The reliability of the items analyzed was over .70.
B. Procedure: A Specific research procedure In this study, we used the GPA of the subjects for the
consisted of preparation, research, and first semester of 2017. For the purpose of this study, all
implementation stages. For the experimental the data used in the analysis were analyzed using SPSS
group, a task-based instructional model was win 21.0 and independent t test.
used for class activities, such as brainstorming,
idea meeting, finding information, discussing, Result and discussion
solving problems, comparing, and presentation.
There are four main steps for an English class A. Difference in English Efficacy According to
model using Artificial Intelligence through Teaching Method: There was a statistically
flipped learning; pre-class, in-class, post-class, significant difference (p<.05) between the
and reflection. Artificial Intelligence was used at experimental group and the control group in the
Downloaded From IP - 83.122.38.10 on dated 17-Oct-2024

the stage of the pre-classes, such as various video speaking area compared with the difference of
contents with which students can practice English the English efficacy according to the teaching
Members Copy, Not for Commercial Sale

language. It is a step that allows learners to share method as table 1 shown. However, there was no
www.IndianJournals.com

their knowledge and expand it into collective statistically significant difference (p>.05) between
intelligence through cooperative learning among the experimental group and the control group in the
colleagues at the stage of in-class. In this process, listening area. The standard deviations also showed
creative application: discussion, presentation, that the experimental group had a statistically
problem solving, etc. were used. Group significant difference to that of the control group.
discussion to make presentations was done for As the results show, an English class model using
the use of English language through the selection Artificial Intelligence through Flipped Learning
of topics according to the presented situation and has a positive impact on students’ self-efficacy
conditions. In this stage, this study was focused
on students’ metacognition which affects their Table I: Difference in english efficacy according to
self-efficacy. Also peer instruction was applied. teaching methods
Students helped their peers who had difficulty
Mean (SD)
understanding what they’ve learned. At the stage t p
of post-class, peer feedback and teacher’s feedback Experimental Control
were done. Students had a final discussion time Listening 3.48(.72) 3.40(.68) .45 .64
with their group members around 1) What did Speaking 3.56(.70) 3.21(.60) 2.21* .03
you learn? 2) What did you feel? 3) How will the
*p<.05
things you’ve learned be applied? For the control
group, the most general teaching model in English B. Difference in Academic Achievement According
class was used, such as using textbooks, and any to Teaching Method: In the case of academic
necessary ppts. Students focused on controlled achievement, there was a significant difference
practice activities for understanding and applying between the two groups in listening and speaking
new language materials. It means that students in
achievement according to teaching methods. As
the control group practiced listening and speaking
table 2 shown, the average of the experimental
through repetitive practice.
group was 20.04 in the listening area, and the
average of the control group was 18.00. Also, in
Data analysis the speaking area, the average of the experimental
The model of Cubillos and Ilventor 12 was used to group was higher (17.04) than the average of the
measure of self-efficacy, alongside a checklist asking control group (13.70).
 560 Indian Journal of Public Health Research & Development, September 2018, Vol.9, No. 9

The standard deviations also showed that in the C. Academic self-efficacy, motivation, and
listening area, the control group was higher than satisfaction level of: As the table 3 shown, the
that of the experimental group. However, in the results of the students’ academic self-efficacy,
area of speaking, the experimental group was learning motivation, and class satisfaction were as
higher than that of the control group. follows: motivation (M = 3.55), class satisfaction
(M = 3.36), and academic self-efficacy (M = 3.18).
Table Ⅱ: Difference in Academic achievement In terms of academic self-efficacy, task difficulty
according to teaching methods (M = 2.95) did not reach the average (M=3.18). ).
In terms of motivation, external motivation was
Mean (SD) higher (M = 3.89) than intrinsic motivation (M =
t p
Experimental Control 3.20). In the case of class satisfaction, evaluation
Listening 20.04 (2.32) 18.00 (3.21) 2.31* .02 and feedback (M = 3.26) were the lowest, and
Speaking 17.04 (4.70) 13.70 (4.00) 2.42* .02 language expression (M=3.27) was the second
lowest of other sub-factors, such as discussion,
*p<.05 presentation, and media.

Table Ⅲ: academic self-efficacy, motivation and satisfaction level of class

Academic self-efficacy Motivation Class satisfaction

M(3.18) SD(.46) M(3.55) SD ( .49) M(3.36) SD(.55)
Downloaded From IP - 83.122.38.10 on dated 17-Oct-2024

Presentation
imagination
Confidence
Members Copy, Not for Commercial Sale

regulation

Feedback
Difficulty

External
Intrinsic

Discuss

Media
www.IndianJournals.com

Using
Task

Self-

M 2.95 3.27 3.34 3.20 3.89 3.27 3.39 3.42 3.36 3.26
SD .73 .55 .73 .59 .66 .60 .77 .67 .76 .66
D. Correlation between Academic Self-efficacy and Academic Achievement: Table 4 shows that the
correlation between academic self-efficacy and learning motivation is statistically significant at .45, the
correlation between academic self-efficacy and satisfaction of class is .26 which is statistically significant, and
the correlation between academic self-efficacy and academic achievement is also statistically significant at .22.
The correlation between self-efficacy and academic achievement is .20, and the correlation between learning
motivation and academic achievement is .28, and these are also both statistically significant as well. However,
there was no statistical significance between class satisfaction and academic achievement as table 4 shows.

Table Ⅳ: Correlations of Academic Self-efficacy and Academic Achievement

Learning Satisfaction of Academic

Self-efficacy
Motivation class achievement
Self-efficacy .45* .26* .22*
Academic Achievement .20* .28* 1
*P<.01
Conclusion and self-directed learning. So in this study, we focused
on how to apply Artificial Intelligence in the classroom
This paper proposed how to use Artificial using Flipped Learning, partly applying it at the stage of
Intelligence in the English language learning classroom. the pre-class.
According to the previous literatures, the application of
Flipped Learning in the classroom has shown to have a As a result of the study, an English class model using
positive educational effect on a variety of areas such as Artificial Intelligence through Flipped Learning has a
academic achievement, attitude, collaborative learning, positive impact on students’ self-efficacy. There was a
 Indian Journal of Public Health Research & Development, September 2018, Vol.9, No. 9 561

statistically significant difference (p<.05) between the Ethical Clearance: Taken from Hannam University
experimental group and the control group in speaking Source of Funding: Hannam University(2018)
according to teaching methods. Also, in case of academic
achievement, there was a significant difference between Conflict of Interest: Nill
the two groups in listening and speaking achievement
according to teaching methods. In the listening area, References
the average of the experimental group was 20.04 and
1. Zhou, Y. Supporting multilingual Internet
the average of the control group was 18.00. Also in the
searching and browsing, The University of
speaking area, the average of the experimental group
was higher (17.04) than the average of the control group Arizona.2006. DOI:10.1007/s10798-008-9051-y
(13.70). The results showed that the students in the 2. James. Education for the future: Participatory
experimental group with an English class model using research as a link between non-formal adult
Artificial Intelligence through Flipped Learning improved education and community development in Brazil,
their English listening and speaking achievement. University of Minesota, Ph.D.dissertation. 2000.
In terms of academic self-efficacy, task difficulty 3. Max Tegmark. Benefits & Risks of Artificial
(M = 2.95) did not reach the average (M=3.18). ). Intelligence, Future of Life Institute. Future of
It means that students had difficulty doing their pre- life, Ph.D.dissertation. 2016.
class task, which is very different in an English model 4. Jerry Kaplan. Humans Need Not Apply, Yale
using Artificial Intelligence through Flipped Learning, University Press. 2016.
compared to traditional teaching methods. In the area of
Downloaded From IP - 83.122.38.10 on dated 17-Oct-2024

motivation, external motivation was higher (M = 3.89) 5. M.H. Shin. Preferred Ways of Study based on
than intrinsic motivation (M = 3.20), and in case of class Korean Cultural Background, IJST.2016; 9(37).
Members Copy, Not for Commercial Sale

satisfaction, evaluation and feedback (M = 3.26) stayed DOI:10.17485/ijst/2016/v9i37/102554

www.IndianJournals.com

in the lowest range. The correlation between academic 6. Choi, Rack-In, Cho,Jung-Gil. An Enhance
self-efficacy and satisfaction of class is .26 which is Scheme of Effectiveness on Lifelong Education
statistically significant, and the correlation between E-Learning through Application of Artificial
academic self-efficacy and academic achievement is also Intelligence (AI), JKIIT. 2012;10(12).
statistically significant at .22.
7. Borkowski, J., Carr, M., & Pressely, M.
It can be seen that students had difficulty doing their “Spontaneous” strategy use: Perspectives from
tasks, yet they are able to solve problems themselves by metacognitive theory. Intelligence.1987;,11; 61-75
investing more learning time using Artificial Intelligence
8. Bandura A., Self-efficacy: The exercise of control.
and combining it with both teacher-student and student-
New York: W.H. Freeman. 1997. DOI: 10.4236/
student interaction than using traditional classroom
teaching methods. According to student interviews, using jss.2014.211021 4,488
smart devices with Artificial Intelligence helped increase 9. Myxter, J. The effects of flipping the classroom
practice time outside of class. With less work for teachers, environment and student achievement. Master
artificial intelligence devices can give them more time thesis, University of Minot State.2014.
to create engaging and creative classroom activities, and 10. Strayer, J. How learning in an inverted classroom
work on their own professional development.
influences cooperation, innovation and task
This study may give educators a chance to think orientation. Learning Environments. 2012;15(2);
about student-centered class and the merits of presenting 171-193. DOI 10.1007/s10984-012-9108-4
an effective English learning model as a new learning 11. Brown, A.L. Metacognition, executive control,
strategy. However, this research has its limitations in self-regulation, and other more mysterious
that there is some difficulty in generalizing the results,
mechanisms. In F. E. Weinert & R. H. Kluwe, (Eds.)
because in the present study, the academic self-efficacy
Metacognition, motivation, and understanding.
and learning motivation on academic achievement was
Hillsdale, NJ: Lawrence Erlbaum. 1987;67-116.
statistically significant, but the influence was quite
different from the theoretical background of the previous DOI: 10.12691/education-4-2-5.
studies. Also, the small number of the subjects may also 12. Cubillos, J.H. & Ilventor, T.The Impact of Study
have influenced the result of this study, so subsequent abroad on Students’ Self-efficacy Perceptions.
studies will require more comprehensive research to FLA.2012; 45(4); 494-511. DOI: 10.1111/
address these factors. flan.12002. 494.