AI Digital Textbooks (hereinafter AIDT) are educational software characterized by AI-based learning diagnosis and analysis, personalized learning reflecting individual learning levels and pace, and learning courseware designed from the learner's persp...
AI Digital Textbooks (hereinafter AIDT) are educational software characterized by AI-based learning diagnosis and analysis, personalized learning reflecting individual learning levels and pace, and learning courseware designed from the learner's perspective. AIDT was introduced into educational settings based on expectations of supporting learners' self-directed learning and enhancing class participation and satisfaction by providing various multimedia learning materials.
In particular, AIDT for English language instruction was introduced with the goal of improving learners' comprehension and supporting individualized learning through functions such as AI-recommended learning paths, vocabulary and text interpretation, explanations of grammatical concepts, native speaker voice provision, error analysis, and immediate feedback. However, empirical research on how AIDT for English language instruction actually affects learners in school educational settings remains insufficient. Accordingly, there is a need to verify the educational effectiveness of AIDT for English language instruction.
Meanwhile, high school students represent a period when motivation for English learning is relatively higher compared to other age groups, as they recognize the importance of English proficiency in the process of preparing for university admission and future careers, and they constitute a group familiar with digital environments and highly receptive to new technology-based learning tools. These characteristics suggest the necessity of verifying the educational effectiveness of AIDT for English language instruction targeting high school students.
Therefore, this study aims to empirically analyze the effects of English instruction utilizing AIDT on high school students' affective characteristics and academic achievement. To this end, a nonequivalent pretest-posttest design was applied to compare and analyze the AIDT group (English instruction utilizing AIDT) and the print textbook group (English instruction utilizing print textbooks) targeting first-year high school students. The specific research questions are as follows:
1. Are there differences in the effects on high school students' affective characteristics between the AIDT-utilizing English instruction group and the print English textbook instruction group?
1-1. Are there differences in the effects on self-directed learning ability between the two groups?
1-2. Are there differences in the effects on class participation between the two groups?
1-3. Are there differences in the effects on class satisfaction between the two groups?
1-4. Are there differences in the effects on post affective characteristics according to high, medium, and low levels of pre affective characteristics between the two groups?
2. Are there differences in the effects on high school students' academic achievement between the AIDT-utilizing English instruction group and the print English textbook instruction group?
2-1. Are there differences in the effects on academic achievement between the two groups?
2-2. Are there differences in the effects on post academic achievement according to high, medium, and low levels of pre academic achievement between the two groups?
3. How does AIDT-utilizing English instruction affect academic achievement mediated by affective characteristics?
3-1. Is there a mediating effect of self-directed learning ability?
3-2. Is there a mediating effect of class participation?
3-3. Is there a mediating effect of class satisfaction?
4. What are students' and teachers' perceptions and responses to AIDT-utilizing English instruction?
The main research methods and procedures are as follows. First, the research participants consisted of a total of 407 first-year students from general private high schools A and B, respectively constituting the AIDT group utilizing AIDT for English instruction and the print textbook group receiving print English textbook instruction. Both groups conducted English instruction based on the same curriculum and instructional content, with the AIDT group utilizing AIDT's learning support functions such as learning diagnosis, error analysis, and individualized learning path recommendations throughout instruction, while the print textbook group received English instruction utilizing print textbooks as usual during the same period. To confirm the effects of AIDT on affective characteristics, self-directed learning ability tests, class participation tests, and class satisfaction tests modified and supplemented by the researcher based on prior studies were utilized. To confirm the effects on academic achievement, scores from the March and September English mock examinations of the National Joint Academic Achievement Assessment were utilized. Private tutoring time and individual study time were used as control variables. Data analysis employed difference-in-differences, analysis of covariance, and structural equation modeling to verify mean differences between groups. The difference-in-differences analysis examined both a basic model without control variables and a controlled model with private tutoring time and individual study time as control variables to confirm the effect patterns of AIDT-utilizing English instruction considering learners' individual learning background factors. In addition to analyzing all students, to examine the possibility that the effects of AIDT-utilizing English instruction may differ according to learners' initial levels, pre affective characteristics and pre academic achievement levels were divided into high, medium, and low levels, and level-specific effects were analyzed together. Furthermore, to investigate perceptions and responses to AIDT-utilizing English instruction, focus group interviews were conducted with students and teachers in the AIDT group, as well as other first-year high school teachers with experience in AIDT-utilizing English instruction, to qualitatively analyze AIDT utilization experiences and changes in the learning process.
The main research findings are as follows.
First, the AIDT-utilizing English instruction group showed more positive effects on affective characteristics (self-directed learning ability, class participation, class satisfaction) than the print English textbook instruction group. As a result of applying difference-in-differences to all students, the changes in self-directed learning ability, class participation, and class satisfaction from pretest to posttest in the AIDT group were significantly greater than in the print textbook group. For self-directed learning ability, the difference-in-differences estimate was positively significant in both the basic model and the controlled model, and class participation and class satisfaction were also positively significant in both models. Additionally, according to the analysis by high, medium, and low levels of pre affective characteristics, self-directed learning ability was statistically significantly higher in the AIDT-utilizing English instruction group than in the print textbook instruction group only at the 'low level', while there were no significant differences between the two groups at the 'medium level' and 'high level'. In contrast, class participation and class satisfaction were statistically significantly higher in the AIDT-utilizing English instruction group at all 'high, medium, and low levels'. These results suggest that AIDT for English instruction has a positive influence on overall affective characteristics such as self-directed learning ability, class participation, and class satisfaction, and particularly that class participation and class satisfaction showed improvement effects regardless of learners' initial levels, while self-directed learning ability showed relatively prominent effects among students with low initial levels.
Second, the AIDT-utilizing English instruction group showed more positive effects on academic achievement than the print English textbook instruction group. However, the group effect on academic achievement showed differential effects that varied according to pre academic achievement levels. For example, as a result of difference-in-differences analysis of all students in both groups, the difference in average changes in post academic achievement relative to pre academic achievement between the AIDT-utilizing English instruction group and the print English textbook instruction group was not statistically significant in both the basic model without control variables and the controlled model controlling for private tutoring time and individual study time. However, as a result of conducting difference-in-differences analysis by dividing pre academic achievement into high, medium, and low levels, the effect of the AIDT-utilizing English instruction group was positively significant at the 'low level' and 'medium level', while the group effect was negatively significant at the 'high level'.
Third, as a result of analyzing the effect of pre academic achievement level on post academic achievement for each of the AIDT group and print textbook group, the difference in post academic achievement according to pre academic achievement level was not statistically significant in the AIDT-utilizing English instruction group, whereas pre academic achievement level had a significant influence on post academic achievement in the print English textbook instruction group. These results suggest that in the case of print English textbook instruction, initial achievement level functions as a major factor determining subsequent academic achievement, showing a tendency for achievement gaps to become fixed, whereas AIDT-utilizing English instruction may have functioned in the direction of mitigating achievement gaps by weakening the influence of initial achievement level.
Fourth, the mediating effects of all affective characteristics were found in the influence of AIDT-utilizing English instruction on academic achievement. As a result of utilizing the bootstrapping technique to confirm whether the mediating effects of self-directed learning ability, class participation, and class satisfaction were statistically significant, when control variables were not set, the standardized coefficients of the direct effect of AIDT-utilizing English instruction on academic achievement were 0.153∼0.201, and the indirect effects were 0.074∼0.123, all of which were statistically significant. Meanwhile, when private tutoring time, individual study time, pre affective characteristics, and pre academic achievement were set as control variables, the direct effects were not statistically significant for all three variables, but the standardized coefficients ranged from 0.041∼0.086, and the indirect effects were 0.024∼0.056, all of which were statistically significant.
Fifth, as a result of conducting focus group interviews on students' and teachers' perceptions and responses to AIDT-utilizing English instruction, AIDT-utilizing English instruction was found to be helpful in improving self-directed learning ability, class participation, and class satisfaction for 'medium and low' level students or students not receiving private education rather than 'high level' students. Additionally, English instruction utilizing AIDT was found to be helpful in developing students' basic learning abilities.
In summary, this study is significant in that it empirically verified that AIDT-utilizing English instruction has positive effects on high school students' affective characteristics and academic achievement. However, since these research findings analyzed only the effects on English instruction targeting students from two high schools, there are limitations in generalizing them. Accordingly, there is a need to research the effects of AIDT targeting more high school students, including subjects such as mathematics and informatics.