Understanding college students’ acceptance of machine translation in foreign language learning: an integrated model of UTAUT and task-technology fit

UTAUT model

Existing research has developed several models to explain technology acceptance, such as the Theory of Planned Behavior (TPB), the Technology Acceptance Model (TAM), and UTAUT. Among these, the UTAUT model has been widely recognized as the most comprehensive one. As defined by Venkatesh et al. (2003), the UTAUT model integrates eight major technology acceptance models, including the TAM, Social Cognitive Theory (SCT), and TPB, which are commonly employed to determine users’ intentions and behaviors regarding technology adoption. The model posits that users’ intention to use information technologies is mainly influenced by core variables (performance expectancy, effort expectancy, social influence, and facilitating conditions), moderating variables (gender, age, experience, and voluntariness), and outcome variables (behavioral intention to use and actual use behavior).

Research has widely-used UTAUT to explain and predict users’ technology adoption intentions, overcoming the shortcomings of a single technology acceptance model (Dwivedi et al., 2019). Due to its wide applicability, the UTAUT model has been applied to diverse research contexts related to technology adoption. Scholars often combine the UTAUT model with other theories to enhance its explanatory power by incorporating additional external variables. For instance, Li and Zhao (2024) combined UTAUT with the social presence theory to investigate factors influencing students’ intention to use of MOOCs, demonstrating that UTAUT effectively predicts student satisfaction. Wut et al. (2024) explored students’ intention to join blended learning courses in higher education based on the Community of Inquiry framework and UTAUT model, showing that students’ behavior intentions to engage in blended learning significantly influenced their attitude towards blended learning.

Building on the original UTAUT model, this study incorporates user attitude as a core variable and adds experience and TTF as moderating variables to deepen our understanding of MT acceptance among college students. First, user attitude is often regarded as a critical factor that directly influences an individual’s willingness to adopt information technologies in technology acceptance models (e.g., TAM) (Andrews et al., 2021). Numerous studies have demonstrated that attitudes significantly impact the acceptance of specific technologies in foreign language learning (e.g., Li et al., 2019; Almusharraf and Bailey, 2023). In the context of MT, students’ attitudes toward the perceived reliability and the quality of translations may profoundly affect their willingness to use it (Jolley and Maimone, 2022). Consequently, this study incorporates attitude as a core variable to explore college students’ acceptance of MT in foreign language learning.

Second, experience with MT use has been shown to directly affect willingness and motivation to use MT, as well as have an impact on the relationship between other variables such as perceived usefulness and perceived ease of use (Rossi and Chevrot, 2019; Yang and Wang, 2019). Venkatesh et al. (2012) highlighted that most studies have thoroughly explored the effects of UTAUT’s core constructs, often overlooking moderators such as experience that could have impacts on the relationships between different determinants within the model. By including experience as a moderating variable, this study seeks to investigate its impact on shaping students’ perceptions and adoption of MT.

Third, although UTAUT has proved to be effective in explaining psychological and behavioral factors affecting technology use, it fails to account for the specific characteristics of the tasks that users perform or the fit between tasks and the technology itself. This theoretical gap can be effectively addressed through the TTF model, which has been empirically validated as a robust framework for evaluating how well a technology optimally supports users in performing specific tasks (Zhou et al., 2010; Wan et al., 2020). Therefore, the present study incorporates TTF to examine its role in influencing the adoption and behavioral intention to use MT. Finally, this study excludes facilitating conditions from the UTAUT model due to its lack of relevance to the study. In the context of MT, facilitating conditions (e.g., server capacity or technical support) might be assumed to be largely handled by MT service providers, such as mobile app developers or browser extension platforms. Also, the widespread availability and ease of access to MT tools further diminish the degree to which organizational and technical infrastructure support the use of MT (Wang & Wang, 2021). Therefore, this study excludes facilitating conditions in the research model. By integrating user attitude, experience, and TTF into the UTAUT model, the study could provide a more comprehensive understanding of college students’ acceptance of MT in foreign language learning and how well MT fits specific language learning tasks.

Task-technology fit theory

Task-Technology Fit (TTF) is defined as the degree to which technology supports users in performing their tasks effectively (Goodhue and Thompson, 1995). This theoretical framework posits task and technology characteristics as two antecedents that function together to promote better task performance (Lin and Wang, 2012). It argues that users are more likely to adopt a technology if its features are in accordance with task requirements and improve their performance. As such, TTF provides a solid foundation for analyzing factors related to technology use and adoption, particularly when combined with models such as TAM and UTAUT (Rahi et al., 2021).

In recent years, TTF has been applied to evaluate the effectiveness of AI-driven tools and their impact on students’ learning performance. For instance, Wang et al. (2023) applied TTF to assess user satisfaction with new online learning spaces, indicating that TTF had significantly impacted user satisfaction and the continuance intentions toward using these spaces. This highlights the importance of task-technology fit in ensuring the efficiency of educational technologies. Recent studies have integrated UTAUT with TTF to provide a more comprehensive model for technology adoption in educational and domain-specific learning contexts (e.g., Tian and Yang, 2023; Du and Lv, 2024). These studies suggest that TTF could significantly influence college students’ perceptions of educational technologies, highlighting the importance of integrating TTF with other theoretical frameworks to provide a more comprehensive explanation for the relationship between TTF and the acceptance of technologies.

TTF has also been applied to translation studies, particularly in evaluating how effectively MT support professional translation tasks. It demonstrated that when MT systems align well with current domain-specific tasks, they can significantly enhance translation satisfaction and performance (Cadwell et al., 2018; Cui et al., 2023). Yang (2024) examined the role of MT fit and machine translation literacy in educational settings. It showed that both technology and task characteristics positively impacted MT fit, highlighting the significance of integrating TTF into MT research for examining the impact of MT in both foreign language learning and professional translation settings. As a widely-used learning tool, MT can support learners with diverse learning requirements, such as reading comprehension, vocabulary acquisition, and text construction (Lee, 2020; Yang et al., 2023). However, the effectiveness of MT depends on its ability to support domain-specific learning tasks. A high task-technology fit implies that MT not only meets the needs of students while performing specific learning tasks, but also results in better learning performance and satisfaction (Yang, 2024). In light of this, it is critical to take into account not just learners’ intentions to use MT, but also how well they align with the specific learning tasks. Thus, this study integrates UTAUT with the TTF to comprehensively explore students’ MT adoption in foreign language learning.

Research hypothesis

Performance expectancy (PE)

Performance expectancy (PE) refers to the extent to which users believe that a technology will improve their task performance, similar to the concept of perceived usefulness (Venkatesh et al., 2016). Previous research has shown that PE has a positive role in promoting learners’ behavioral intention to adopt various innovative technologies (Abbad, 2021). In this study, PE reflects the degree to which students believe that MT can improve the effectiveness of their foreign language learning. Yang and Wang (2019) posted that the quality of MT played a significant and positive role in affecting users’ willingness to adopt it for language learning tasks. Similarly, Lee (2023) highlighted that learners’ perceived efficacy of MT in enhancing writing performance emerged as a critical predictor of their adoption intentions. When students perceive that MT can enhance their academic performance or learning efficiency, their behavioral intention to use it increases. Thus, the following hypothesis is proposed:

H1: Performance expectancy has a significant positive impact on the behavioral intention to use machine translation.

Effort expectancy (EE)

Effort expectancy (EE) is a key component of the UTAUT model that describes the perceived ease of using a technology, closely related to perceived ease of use (Venkatesh et al., 2003; Tian and Yang, 2023). Specifically, the less effort users expend in using an information technology system, the higher the intention to adopt the technology. In translation studies, the adoption of MT among students is largely influenced by how effectively they generate precise and accurate translations that require minimal post-editing (Lee and Briggs, 2021; Tian and Yang, 2023). Research indicates that factors such as ease of use, speed, and operational convenience influence users’ willingness to use MT (Yang et al., 2021). Accordingly, this study proposes the following hypothesis:

H2: Effort expectancy has a significant positive impact on the behavioral intention to use machine translation.

Social influence (SI)

Social influence (SI) refers to the extent to which individuals’ judgments of technologies can be influenced by the actions and opinions of others around them, which is related to social norms in other technology acceptance models (Venkatesh et al., 2003). In this study, SI represents how much college students are influenced by their peers, teachers, parents, or social media in adopting MT for foreign language learning. Relevant research shows that the attitudes of teachers and peers towards the use of MT have an important impact on their intention to engage with MT (Ducar and Schocket, 2018; Stapleton and Kin, 2019; Wang and Wang, 2021). Additionally, institutional support for using MT in academic or professional settings can further encourage its adoption among translators (Paterson, 2022). Based on this, this study proposes the following hypothesis:

H3: Social influence has a significant positive impact on the behavioral intention to use machine translation.

Attitude (ATT)

Attitude (ATT) refers to an individual’s overall evaluation of a technology, including both positive and negative perceptions (Venkatesh et al., 2003). In technology acceptance models, attitude has been identified as an important intrinsic variable directly influencing users’ willingness to adopt technologies (Dwivedi et al., 2019; Andrews et al., 2021). Jolley and Maimone (2022) emphasized that students’ attitudes toward MT are critical for determining their intention to use it. In other words, when students have positive attitudes toward the use of MT, they are more likely to perceive it as useful, which in turn strengthens their intention to use it. Thus, this study proposes the following hypothesis:

H4: User attitude has a significant positive impact on the behavioral intention to use machine translation.

Behavioral intention (BI)

Behavioral intention (BI) reflects users’ tendency towards information technologies and the possibility of future use (Venkatesh et al., 2003). UTAUT theory suggests that BI will positively affect their actual usage behavior (Agyei and Razi, 2022). In the context of MT, relevant research has confirmed that students’ behavioral intention to use MT has a significant positive impact on their actual use of the technology (Rossi and Chevrot, 2019; Lee, 2020). The stronger the behavioral intention of students to use MT, the greater the likelihood of students using MT in the future. Therefore, this study proposes the following hypothesis:

H5: Behavioral intention has a direct and positive effect on the use behavior of college students in terms of machine translation.

Moderating variables: experience and TTF

This study incorporates experience and TTF as moderating variables in the model. First, experience is a core moderating variable in the original UTAUT model (Venkatesh et al., 2003). Previous studies highlight its role in moderating the relationships between UTAUT constructs and behavioral intention in the educational sector (Celik, 2016; Lakhal et al., 2021). Also, experience has been proven to be a key factor in understanding students’ intention to learn and use MT. Greater experience with MT helps students become familiar with the types of translations MT excels at and the common errors it produces, making experience a critical moderator in technology adoption models (Yang and Wang, 2019). Therefore, experience is regarded as a key moderator in technology adoption models. Second, TTF reflects the extent to which technology supports users when performing specific tasks (Goodhue and Thompson, 1995). Previous research has demonstrated the importance of task-technology fit in shaping individuals’ perceptions of the actual utility of the technology and has positive effects on behavioral intention (Zhou et al., 2010; Wan et al., 2020; Wang et al., 2024). Recent studies also confirm that TTF has been proven to act as a significant moderator that effectively improves the explanatory power of the UTAUT model (Shirolkar and Kadam, 2023; Du and Lv, 2024). In the context of MT, TTF evaluates how well MT aligns with translation tasks, assessing whether the MT systems are capable of enhancing translation quality by effectively integrating into the learning environment and satisfying the students’ requirements. The following hypotheses are proposed:

H6a: Experience moderates the relationship between performance expectancy and the behavioral intention to use machine translation.

H6b: Experience moderates the relationship between effort expectancy and the behavioral intention to use machine translation.

H6c: Experience moderates the relationship between social influence and the behavioral intention to use machine translation.

H6d: Experience moderates the relationship between attitude and the behavioral intention to use machine translation.

H7a: TTF moderates the relationship between performance expectancy and the behavioral intention to use machine translation.

H7b: TTF moderates the relationship between effort expectancy and the behavioral intention to use machine translation.

H7c: TTF moderates the relationship between social influence and the behavioral intention to use machine translation.

H7d: TTF moderates the relationship between attitude and the behavioral intention to use machine translation.