Second language acquisition
According to this theory proposed by Krashen (1982, 1985, 1989), learners will naturally advance to higher levels of language competence when exposed to sufficient comprehensible input. Conversely, a dearth of such input can impede or halt this progression, potentially leading to fossilization, wherein the learner falls short of attaining native speaker proficiency (Krashen, 1985, p. 43). This stagnation can occur through encountering input that is either overly simplistic, failing to introduce new syntactic structures or a wide range of vocabulary, or excessively complex, rendering comprehension difficult (Krashen, 1985). In both scenarios, learning outcomes may be compromised. Integration of media and online learning improves English skills Krüger, M. (2023). Computer assistant language learning gives autonomous learning. With the advancement of the application of computers in learning a new learning theory known as Connectivism was conceptualized by George Siemens (Siemens, 2005). Computer-assisted language learning with the application of computers and software is more inclined to self-directed learning. Such learning is a concept of a popular theory of learning known as the constructivism learning theory that learners should take control of self-learning.
The application of using computers and mobile phones set a light on mixed findings among scholars. Siemens (2005) proposed the effectiveness of the integration of computers and the Internet in learning. However, numerous scholars challenge casting negative views of computers and the internet in learning a language mostly pointing to the distraction of using computers, mobile phones, and the internet degrading the achievement score of English language learning (Ugur & Koc, 2015; Goundar, 2014; Thomas, 2016; McCoy, 2016; Wexler, 2019; & Molnar et al., 2019; & Feng et al., 2019).
Artificial Intelligence (AI) in Education
The intersection of artificial intelligence (AI) and education has garnered significant attention in recent years. This review synthesizes insights from various scholarly works to provide a comprehensive understanding of this burgeoning field. Several authors have investigated the applications of AI in language learning and education. Thenmozhi et al. (2023) emphasized the positive impact of technology integration, showcasing studies on virtual reality (VR), augmented reality (AR), natural language processing (NLP), and AI-based automation systems in language education. Similarly, Mishra and Kumar (2020) focus on NLP in AI, exploring its theoretical foundations and practical applications in facilitating communication between humans and computers. Moreover, Sanyal and Chakrabarti (2020) discuss AI in the Indian context, examining its evolution, opportunities, and risks. They highlight India's significant stake in the AI revolution and the need for ethical considerations in AI development. Furthermore, Kaur and Gill (2019) provide insights into AI and deep learning for decision-makers, aiming to demystify AI concepts and their implementation in businesses and organizations.
It also explored AI's impact on mental health care and social media. Luxton, Fletcher, and Cai (2020) discuss AI applications in behavioral and mental health care, touching upon language processing and cognitive interventions. Similarly, Feldman, McDonald, and Webber (2019) focus on natural language processing for social media, examining NLP techniques for analyzing social media data. In the realm of education, AI offers transformative potential. Clark (2019) discussed the integration of AI technologies in the classroom, highlighting their role in personalizing instruction and streamlining administrative tasks. Boulay et al. (2022) emphasized the human-centered perspective in examining AI's impact on learners and educators, addressing ethical and pedagogical implications.
Furthermore, Mavrikis and Weatherby (2016) explored AI integration in educational settings, focusing on personalized learning and adaptive tutoring systems. Dickinson et al. (2012) covered various aspects of NLP and computational linguistics in education, providing theoretical frameworks and practical applications. Additionally, several works explored computational lexical semantics and natural language generation. Dizier and Viegas (2010) provided an overview of computational approaches to lexical semantics, while Reiter and Dale (2000) discussed natural language generation systems.
The widespread integration of AI in English education not only enhances students' overall learning levels but also enables teachers to dedicate more time to educational research, understanding students' needs, and refining teaching methodologies (Zhang & Cao, 2022). Artificial intelligence aims to develop machines with intelligence comparable to or surpassing that of humans, capable of tasks such as natural language processing, perception, reasoning, manipulation of objects, acquiring knowledge, and learning, ultimately streamlining activities to improve efficiency (Fitria, 2021). This advancement in technology, coupled with the accessibility of digital platforms through computers and cell phones, extends learning opportunities globally and promotes the utilization of AI (Fitria, 2021). Personalized content, facilitated by adaptive systems driven by big data and AI, revolutionizes digital learning technology, tailoring English learning experiences to individual users' needs and schedules (Fitria, 2021).
Research employing qualitative methods, particularly content analysis, identifies four key themes regarding the use of AI in language teaching and learning, demonstrating its efficacy in pedagogy (Ali, 2020). The integration of information technology with English curricula further deepens the application of AI in teaching, presenting new avenues for optimizing the English teaching process and creating intelligent, personalized learning environments (Bin & Mandal, 2019). In the context of middle school English teaching, leveraging relevant curriculum theories, literature analysis, and field investigations, an implementation plan for an AI-based College English instructional system is proposed, aiming to enhance system functions and humanize the English teaching experience (Bin & Mandal, 2019).
Moreover, the role of ChatGPT in text assessment is explored, emphasizing its ability to offer real-time, personalized feedback to learners, thereby enhancing overall learning experiences. Notably, these applications are based on the latest iteration, ChatGPT 4, distinguishing them from earlier versions (Koraishi, 2023).
Finally, Fitzpatrick (2023) underscores the responsibility of educators in preparing students for the AI-driven future, providing practical strategies for incorporating AI tools into teaching practices.
In conclusion, the literature reviewed highlights the diverse applications and implications of AI in education, language learning, mental health care, and social media. Moving forward, further research and collaboration are essential to harnessing the full potential of AI in enhancing teaching and learning experiences.
The integration of AI in English vocabulary learning particularly in rural Indian schools might not have been investigated. Indian schools are more inclined towards traditional learning methods. Not only that educators were not in favour of using AI or ChatGPT. Therefore, the current study was trying the learning outcomes of the integration of AI into English vocabulary learning.
Research Design
The current study was an experimental teaching conducted to collect the data using AI and for the control group teaching the traditional method. It was an empirical and quantitative research. It investigated the effectiveness of using AI in English vocabulary learning using a t-test.
Research set up
For a comprehensive investigation, two intact classes comprising a total of 115 class 12 students participated in the study at Government Higher Secondary School, Boleng, Siang District- Boleng Block," adopts a field study method. These students were divided into an experimental group and a control group. The experimental group received instruction mediated by AI, whereas the control group received traditional language instruction. Pre-tests and post-tests were administered to assess English learning achievement in vocabulary development. The experimental group of students used ChatGPT, Online Dictionary, and Google Translate. The control group of students was taught the traditional method of lecture and traditional dictionary. There was pre-test and post-test t-test data analysis. The experimental teachings were conducted from January 2024 to the first week of March 2024 in one and a half-month period. The research design and steps are shown in a flowchart in Figure 1.
Sample size
There were N=112 students of participated in the study. The experimental group had 56 students and the control group had 56 students (Table 1).
Table 1 Sample Size and Gender of the Participants for Experimental and Control Groups
|
Gender
|
Experimental group
|
Control Group
|
|
Number of Girls
|
48
|
11
|
|
Number of Boys
|
8
|
45
|
|
Total
|
56
|
56
|
Note. Total number of participants was 112 (experimental group=56, control group=56)
Data collection
At the onset of the beginning of the experiment, after population and sample sizes were determined, a pre-test was conducted to ascertain the level of students’ vocabulary level. After teaching for six weeks, post-test scores were collected for the experimental and the control groups. The scores for the vocabulary test are listed in Table 2 below.
Table 2 Pre-test and Post-test Scores of the Experimental and the Control Group of Students
|
Pre-test scores
|
|
Post-test scores
|
|
Sl.No.
|
Experimental Group
|
Control Group
|
Sl.No.
|
Experimental Group
|
Control Group
|
|
1
|
9
|
9
|
1
|
16
|
12
|
|
2
|
12
|
9
|
2
|
19
|
11
|
|
3
|
13
|
10
|
3
|
15
|
12
|
|
4
|
11
|
7
|
4
|
18
|
12
|
|
5
|
13
|
12
|
5
|
18
|
11
|
|
6
|
14
|
11
|
6
|
19
|
12
|
|
7
|
10
|
10
|
7
|
15
|
13
|
|
8
|
8
|
9
|
8
|
17
|
11
|
|
9
|
13
|
12
|
9
|
19
|
12
|
|
10
|
7
|
5
|
10
|
16
|
11
|
|
11
|
10
|
6
|
11
|
16
|
12
|
|
12
|
11
|
9
|
12
|
16
|
11
|
|
13
|
12
|
8
|
13
|
17
|
11
|
|
14
|
6
|
2
|
14
|
18
|
12
|
|
15
|
13
|
4
|
15
|
16
|
11
|
|
16
|
12
|
12
|
16
|
15
|
11
|
|
17
|
12
|
14
|
17
|
17
|
12
|
|
18
|
11
|
12
|
18
|
18
|
13
|
|
19
|
14
|
5
|
19
|
16
|
12
|
|
20
|
12
|
12
|
20
|
17
|
14
|
|
21
|
15
|
11
|
21
|
18
|
13
|
|
22
|
13
|
9
|
22
|
18
|
12
|
|
23
|
11
|
13
|
23
|
17
|
11
|
|
24
|
12
|
13
|
24
|
18
|
14
|
|
25
|
12
|
10
|
25
|
16
|
13
|
|
26
|
12
|
11
|
26
|
16
|
14
|
|
27
|
9
|
8
|
27
|
17
|
15
|
|
28
|
12
|
12
|
28
|
18
|
13
|
|
29
|
9
|
6
|
29
|
16
|
18
|
|
30
|
4
|
3
|
30
|
12
|
8
|
|
31
|
8
|
12
|
31
|
15
|
12
|
|
32
|
10
|
14
|
32
|
16
|
11
|
|
33
|
8
|
15
|
33
|
18
|
14
|
|
34
|
13
|
13
|
34
|
16
|
11
|
|
35
|
12
|
15
|
35
|
16
|
15
|
|
36
|
10
|
14
|
36
|
16
|
12
|
|
37
|
11
|
12
|
37
|
15
|
11
|
|
38
|
10
|
10
|
38
|
15
|
13
|
|
39
|
9
|
12
|
39
|
18
|
17
|
|
40
|
11
|
10
|
40
|
15
|
11
|
|
41
|
12
|
14
|
41
|
13
|
16
|
|
42
|
5
|
9
|
42
|
15
|
14
|
|
43
|
7
|
12
|
43
|
14
|
15
|
|
44
|
9
|
17
|
44
|
14
|
12
|
|
45
|
9
|
13
|
45
|
16
|
14
|
|
46
|
10
|
10
|
46
|
17
|
13
|
|
47
|
12
|
12
|
47
|
18
|
9
|
|
48
|
16
|
7
|
48
|
18
|
5
|
|
49
|
10
|
12
|
49
|
17
|
4
|
|
50
|
11
|
13
|
50
|
18
|
8
|
|
51
|
11
|
10
|
51
|
15
|
17
|
|
52
|
7
|
9
|
52
|
16
|
7
|
|
53
|
11
|
11
|
53
|
18
|
6
|
|
54
|
7
|
14
|
54
|
17
|
9
|
|
55
|
14
|
13
|
55
|
18
|
6
|
|
|
|
56
|
18
|
8
|
Note. During the pre-test, the experimental and the control group had 55 students in each group.
Data analysis
The data analysis was conducted in two sets. The first set was for pre-test scores. A t-test was conducted using SPSS-26. A second set of t-tests was conducted using SPSS-26.
Pre-test
A pre-test was conducted for both the experimental group and the control group before teaching English vocabulary learning using AI. There was no significant difference in the scores for the experimental group (M=10.6364, SD=2.46729) and control group (M=10.4909, SD=3.14980) conditions t(108)=.270, p=0.788, at 95% Confidence Interval of the Difference. As the p-value is greater than 0.05, which was interpreted as p>0.05, the experimental and control groups were not different statistically before teaching using AI tools for vocabulary learning (Figure 2). This means the experimental and control groups’ vocabulary proficiency was more or less the same before beginning the experimental teaching.
Post-test
To ascertain if AI use in vocabulary is effective or not, a post-test was conducted for both the experimental group and the control group after teaching six English vocabulary learning using AI. There was no significant difference in the scores for the experimental group (M=16.5091, SD=1.51380) and control group (M=11.8421, SD=2.93859) conditions t(110)=10.508, p=.000, at 95% Confidence Interval of the Difference. As the p-value is less than 0.05, which was interpreted as p<0.05, the experimental and control groups were statistically significant after teaching using AI tools for vocabulary learning (Figure 3). This means the experimental group was more effective in scoring than the control group’s vocabulary proficiency level.
Findings
The findings show that using AI is more effective than using traditional methods of English vocabulary learning. Before starting the experimental teaching both groups (Experimental and Control groups) had a similar level of vocabulary proficiency shown by the t-test score, with a p-value greater than 0.05, which was interpreted as p>0.05 (Figure 2). After teaching for six weeks, the t-test for the post-test showed the p-value is less than 0.005, which was interpreted as p<0.05, which means the experimental and control groups were statistically significant which means that the experimental group was more effective in scoring than the control group’s vocabulary proficiency level after teaching using AI tools for vocabulary learning (Figure 3).