Learning adaptation (LA) denotes the ability of students to adapt to the learning environment during the learning process ( He & Zhong, 2006 ). Foreign language learning adaptation (FLLA) is the state or tendency of learners to coordinate with their own conditions and external conditions and develop through continuous adjustment in the process of foreign language learning ( Guo et al., 2021 ). The better the learning adaptability, the better the English learning engagement ( Wang et al., 2023 ). With the innovation of new teaching forms, Guo et al. (2021) updated the five-dimensional structure including foreign language teacher adaptation, teaching content adaptation, teaching environment adaptation, online learning adaptation and teaching model adaptation. This model covers a number of emerging educational factors.

Flipped classroom is an approach to teaching and learning activities where students watch a video lesson outside the class through distance learning and have hands-on activities in the class ( Zainuddin & Halili, 2016 ). Traditional and flipped classrooms show different properties. In a traditional classroom, students usually have little time to discuss questions with the teacher and peers. However, the flipped classroom empowers students to spend more time discussing in the classroom by allowing them to be well-prepared before class. In addition, the introductory video in the pre-class phase of the flipped classroom provides better assistance to the slower learners because they can stop, pause, or rewind the lecture video about the solution to the problem ( Osman et al., 2014 ). A prerequisite for the successful implementation of the flipped classroom is that students should be prepared and learn the material before class, as it relies on the motivation of students to engage in active learning and application ( Burke & Fedorek, 2017 ).

However, the flipped classroom, as a teaching model with different characteristics from the traditional classroom, may present challenges to students’ adaptation. A study by Chen et al. (2014) found that many students had difficulty adapting to the flipped classroom. Most part-time students suggested that the course was heavy as they did not have time to watch videos online. Johnson (2013) also found that students were not prepared for the transition from the traditional classroom to the flipped classroom. Although they verbalized their preference for the flipped classroom, lecture videos were not watched by the majority of students at home. This suggests that some students do not show initiative and motivation in the out-of-class phase and enter the classroom phase unprepared, which may lead to being unable to adapt to flipped classrooms.

The flipped classrooms generally improve learning efficiency by enabling more in-depth classroom discussions and providing students with more opportunities for interaction. However, it also acknowledges that the model’s impact depends on content adaptation—students perform better when they are cognitively and emotionally engaged with the learning material. Compared to other interactive models like problem-based learning (PBL) and task-based language teaching (TBLT), the flipped classroom provides a balance between structured guidance and interactive engagement but may not necessarily lead to better results in certain skill areas, such as oral fluency development.

Learning adaptation and achievement are theoretically linked. Adaptation level theory proposed by Helson (1964) implies that the basis of an individual’s judgement of a stimulus is based on their previous experiences and their recollection of how they perceived similar stimuli in the past. Therefore, students at the higher adaptation level will strive for higher grades because they are not personally satisfied with scoring at the same level as students at lower adaptation levels ( Pollio et al., 1988 ).

The potential relationship between learning adaptations and achievement has also been empirically revealed across different instructional models, although the findings are found to be relatively inconsistent, as shown in Table 1 . In the traditional teaching model scenario, Holliman et al. (2019) confirmed a significant positive correlation between academic adaptation and GPA in 90 UK university students. Collie et al. (2017) investigated 186 UK university freshmen and found that academic adaptation and GPA were not significantly correlated, suggesting that adaptation may be influencing achievement through other pathways. Collie & Martin (2017) reported significant correlations between student-reported adaptations, teacher-reported adaptations and maths achievement in 371 Australian secondary school students.

In the online education context, Martin et al. (2021) found a significant positive correlation between academic adaptation and maths achievement with 1548 Australian high school students. However, Feraco et al. (2023) investigated 435 Italian grades 6-12 students but did not find any significant correlation between learning adaptations and maths achievement. In summary, most studies in the field of education have only focused on offline scenarios (i.e., traditional teaching model) and online teaching model scenarios, whilst flipped classroom combining offline and online scenarios have not yet been addressed. In addition, a lot of research findings have been produced in the subject of Mathematics, but few related studies have been seen in the subject of English. Most importantly, the outcomes are rather controversial. There is no general agreement about the correlation between learning adaptations and achievement. Therefore, the purpose of this study is an attempt to elucidate the relationship between FLLA and English achievement in flipped classroom context. The study aimed to address the following research questions:

RQ1: What is the profile of students’ FLLA?

RQ2: Are there any significant differences in students’ FLLA levels across different gender and grade?

RQ3: How does the FLLA predict English achievement?

The Foreign Language Learning Adaptation Scale developed by Chinese scholars Guo et al. (2021) was applied. The 18-item scale measures students’ adaptation from five factors, including foreign language teacher adaptation, teaching content adaptation, teaching environment adaptation, online learning adaptation and teaching model adaptation. Compared with other scales, this scale is more domain-specific and more appropriate for the flipped classroom model with online and offline features. The scale was measured on a 6-point Likert scale, ranging from “1 (strongly disagree)” to “6 (strongly agree)”. The study by Guo et al. (2021) has confirmed the sound validity (χ²/df = 1.15, CFI = 0.99, TLI = 0.99, RMSEA = 0.03, N = 206) in the Chinese context. In this study, the scale also demonstrated good reliability (Cronbach’s alpha = 0.85).

Students’ English achievement is measured by their scores in the College English Test Band 4 test (CET-4 test) administered by the Chinese Ministry of Education. The test consists of writing, listening comprehension, reading comprehension, vocabulary and structure, with a time limit of 125 minutes and a total score of 710 (106.5, 248.5, 248.5 and 106.5 for each part). The reasons for using CET-4 scores as the measurement tool are as follows. Primarily, compared with international tests such as IELTS and TOEFL, the CET-4 test has a wider popularity among Chinese university students. The test is taken by almost every university student before their graduation. Secondly, as a test administered by the Ministry of Education, the CET-4 exam is authoritative and reliable. The proposition, invigilation and assessment are all strictly executed by experts. Therefore, CET-4 score is a favourable tool for measuring students’ English proficiency.

A total of 296 students were recruited voluntarily to partake in this study through convenience sampling after signing formal consent. The subjects were from a university in China adopting the flipped classroom model. The teaching approach was based on the model of Yuan & Moran (2018) , which included a pre-class, in-class, and post-class stages. In the pre-class stage, students undertake autonomous pre-study online with the assistance of their teachers. In the mid-class phase, students and teachers engage in in-depth discussions and presentations of outcomes offline. In the post-class stage, students review online with the assistance of the teacher and complete the assessment.

All students had 8 - 11 years of English learning experience with no overseas study experience. All students’ first language was Mandarin Chinese and the second language was English. The average CET-4 score of the students reached 452 (SD = 47.99). According to the criteria proposed by Jin et al. (2022) , the cut-offs for CET-4 scores corresponding to the B1 and B2 levels defined in the Common European Framework of Reference for Languages (CEFR) have been set at 388 and 549, respectively. Therefore, the students have achieved an intermediate level of English proficiency. Among the participants, 141 (about 47.6 %) were male and 155 (about 52.4 %) were female. They were from different grades, including 63 (about 21.3 %) freshmen, 71 (about 24 %) sophomores, 98 (about 33.1 %) juniors, and 64 (about 21.6 %) seniors.

The questionnaire was distributed to the students with the consent of the school leadership team. Participants were informed of the nature and purpose of the questionnaire, their unconditional right of non-participation and the confidentiality of the data before filling out the questionnaire. Data collection was completed in November 2023. SPSS 27 and AMOS were utilized during the data processing stage. Reliability analysis was employed to test the research instrument. Descriptive analysis was performed to answer RQ1. Independent sample t-test and ANOVA test were applied to answer RQ2. Pearson correlation analysis and structural equation modelling were used to answer RQ3.

Descriptive statistics were performed to analyze the overall picture of FLLA, as shown in Table 2 . According to Field (2009) , data with skewness values between 0 and ±3.0 and kurtosis below ±8.0 were assumed to be normally distributed. Therefore, the data was judged to be approximately normally distributed. It could be inferred from the table that students’ FLLA was in the upper-middle level as the mean value was higher than the median (M = 3.95 > 3.5), which suggested that students’ adaptation to the flipped classroom was favourable. In detail, students’ language teacher adaptation, teaching content adaptation, teaching environment adaptation, and online learning adaptation were all at the upper-middle level (M > 3.5). However, students’ teaching model adaptation was at the lower-middle level (M < 3.5). Language teacher adaptation had the highest mean (M = 4.79) and teaching model adaptation had the lowest mean (M = 3.33).

Note: FLLA = Foreign language learning adaptation; LTA = Language teacher adaptation; TCA = Teaching content adaptation; TEA = Teaching environment adaptation; OLA = Online learning adaptation; TMA = Teaching model adaptation.

To investigate whether there were gender differences in FLLA levels, an independent sample t-test was performed, as shown in Table 3 . It can be seen that the average FLLA level of males was 3.93 (SD = 0.59) and females was 3.98 (SD = 0.59). There was no significant difference in FLLA levels between males and females (p > 0.05).

The ANOVA test was performed with the aim of exploring whether there is a significant difference in FLLA levels between grades. The results are presented in Table 4 . It is clarified that there is no significant difference in FLLA levels between grades (p > 0.05). It can be concluded from the means plots ( Figure 1 ), which represents the mean value of FLLA levels in different grades, that the FLLA levels of sophomores peaked (M = 4.05), followed by juniors (M = 3.96), seniors (M = 3.91), and finally freshmen (M = 3.86).

In order to investigate the effect of FLLA on English achievement in flipped classroom, the study started with the Pearson correlation analysis between the factors of FLLA and English achievement, as shown in Table 5 . It could be inferred that FLLA and English achievement were significantly and positively correlated with a small to medium effect size (r = 0.22, 0.1 < r < 0.3, p < 0.01) ( Cohen, 1988 ). More specifically, both teaching content adaptation and teaching model adaptation were significantly and positively correlated with English achievement (p < 0.01), with medium to large effect sizes for teaching content adaptation (r = 0.35, 0.3 < r < 0.5) and small to medium effect sizes for teaching model adaptation (r = 0.23, 0.1 < r < 0.3). However, the correlations between language teacher adaptation, teaching environment adaptation, online learning adaptation and English achievement were not significant (p > 0.05).

Note: EA = English Achievement.

In order to present a clearer picture of the relationship between FLLA and English achievement, structural equation modelling was utilized. The standardized results applying maximum likelihood estimation are shown in Figure 2 . It can be inferred that only teaching content adaptation is a significant predictor of English achievement (β = 0.32, p < 0.001). The other dimensions of foreign language learning adaptation, i.e., language teacher adaptation, teaching environment adaptation, online learning adaptation, and teaching model adaptation could not significantly predict English achievement (p > 0.05).