Schneider and Radu (2022) defined AR as a technology that is composed of digitally generated 3D content with real-world objects that users are looking at. Such an computer-generated 3D content is overlaid on a view of the real world via a specialized display. Thus, in this sense, AR could enable the invisible visible (Akçayır & Akçayır, 2017) . This is especially conducive to support learners to understand complex concepts, which are not directly observable. Moreover, differentiating from virtual reality, AR contains real-world aspects, which make the learners’ experience always situated in the physical environment.

AR has been witnessed an accelerated growth since 2010 and it has been proved its the potential for enhancing students’ learning performance, given its features in providing contextual visualization and interactive learning experiences (Cai et al., 2022) . In the realm of language education, AR has brought benefits for both first and second language learners (Qiu et al., 2021) . For instance, in the study of (Cai et al., 2022) , they presented a meta-analysis of AR in language education for both L1 and L2 settings for the past eighteen years, with the main focus on the influences of AR on learning gains as well as motivation of learners. While current AR applications in language learning predominantly address alphabetic languages like English (Fan et al., 2020) , the exploration in non-alphabetic languages, especially Chinese, is relatively underdeveloped. Research has predominantly demonstrated AR’s effectiveness in English language learning, particularly in reading and vocabulary (Cai et al., 2022; Parmaxi & Demetriou, 2020; Xie & Huang, 2023) . However, its influence on Chinese reading comprehension has not been thoroughly investigated. Echoing Garzón and Acevedo (2019) , the effectiveness of AR varies across subjects, but in language learning, it shows promises in aiding younger learners in alphabetic languages (e.g., Cai et al., 2022; Chen & Chan, 2019; Lantavou & Fesakis, 2018 ). Nonetheless, the specific impact of AR on Chinese as a Foreign Language (CFL), especially in reading comprehension, remains to be discovered.

Reading engagement involves the reader’s ability to interact with the text in a strategic and motivated way (Guthrie & Wigfield, 2000) . This concept has been defined by Fredricks et al. (2004) as encompassing three dimensions, i.e., behavioral, emotional, and cognitive. The impact of reading engagement is significant, affecting knowledge acquisition (Guthrie & Wigfield, 2000) , learner performance (Guthrie et al., 2013) , and various other cognitive, affective, and behavioral outcomes. Consequently, reading engagement plays a crucial role in the learning process related to reading. Similarly, reading emotion is another vital element influencing reading comprehension (Hyland & Jiang, 2016) . Research shows that emotions like boredom, enjoyment, and anxiety significantly affect language learners’ capacity to comprehend and process text (Hyland & Jiang, 2016) . Positive emotions, such as enjoyment, are seen as beneficial for reading and enhancing academic performance (Goetz et al., 2007) , while negative emotions, like anxiety, are associated with poorer academic outcomes. Additionally, boredom is found to have a negative relationship with motivation, cognitive engagement, self-regulation, and academic success (Rowe & Fitness, 2018) .

Collectively, these studies suggest that reading comprehension is shaped by the dual factors of reading emotions and engagement. Addressing these aspects effectively could lead to improved teaching methods and better learning outcomes in reading comprehension. For instance, a study by Kamarainen et al. (2013) revealed that the use of AR could enhance learners’ positive perceptions towards science and boost content comprehension. This was supported by Schneider and Radu (2022) , who noted that participants using AR tools developed more favorable attitudes towards physics learning.

Smagorinsky and Smith (1992) defined that literacy scholars have centered on three kinds of literacy in a broader sense, namely general, task-specific, and community-specific. In general literacy, from the cognitive perspective, reading comprehension has been long recognized as a creative act of structure building (Bartlett, 1932) where learners selects related information, mentally organizes it into a coherent structure, and itegrates it with relevant prior knowledge activated from long-term memory (Mayer, 2011) . Skilled readers effectively leverage their pre-existing knowledge to optimize how they select, organize, and assimilate new information (Smagorinsky & Mayer, 2022) . Research indicates that learners with relevant prior knowledge typically outperform those without in reading comprehension tests (Bransford & Johnson, 1972; Pearson et al., 1979) . Furthermore, learners are more likely to retain information that aligns with what they already know (Lipson, 1983; Pichert & Anderson, 1977) . Furthermore, task-specific literacy refers to the differences of declarative and procedural knowledge employed between learning to read different genres, such as argumentation and narration (Smagorinsky & Mayer, 2022) . Akin to Smagorinsky and Mayer (2022) , we acknowledge that general literacy is necessary, yet there remains insufficiency as people grow mature and begin to read an increasing variety of genre texts. Moreover, the cognitive variation existed in reading diverse types of genre articles (Smagorinsky & Mayer, 2022) . Thus, task-specific literacy is needed to acquire as they grow academically.

Taken together, in this paper, we take from both perspectives of general and task-specific literacy to ground our research design. Reading in Chinese, a logographic system, poses unique cognitive processing challenges compared to phonetic systems, especially for foreign language learners aiming for reading proficiency. In learning sciences field, scholars have explored the role of AR on learning mainly using three perspectives, namely conceptual, socio-contextual, and affective (Schneider & Radu, 2022) . From the perspective of cognition, learning is defined as the procedure of individual information processing, and knowledge was regarded as the representation of information within an individual mind. In viewing this lens, AR could be designed to contain different traits that may benefit learning. Interactive 3D visualizations offered by AR technologies enable representations of educational content that are unattainable with conventional 2D screens or textbooks. The use of AR in teaching reading comprehension is drawing increasing interest from researchers, such as Parmaxi and Demetriou (2020) . This is because AR creates an engaging and interactive environment beneficial for both general reading comprehension and specific literacy tasks. Drawing on the schema theory, AR can help learners establish and access prior knowledge, enhancing their comprehension abilities (Dai, 2003) . Additionally, from an affective perspective, studies have examined AR’s impact on learner engagement, emotions, and motivation. For example, Schneider and Radu (2022) observed that AR could present timely information within authentic settings and present content in an engaging, playful manner, thereby boosting engagement, personal agency, and self-efficacy.

In summary, previous research has highlighted the potential of AR in language education, yet its effectiveness in aiding CSL learners in reading comprehension is still an unknown area. Moreover, given the importance of reading engagement and emotions in reading comprehension, this study aims to address the following research questions to achieve the objectives of this study:

1) What is the effect of AR-enhanced technology on Chinese as a Foreign language learners’ reading emotions?

2) To what extent does AR-enhanced technology influence Chinese as a Foreign language learners’ reading engagement?

This research was carried out in part of the research project named ARCL (see Xie et al., 2024 ) during an online Chinese Reading course in the latter part of the Fall 2023 semester. The course was led by a proficient Chinese language instructor with over 10 years’ teaching experience in instructing Chinese as a Foreign Language learners. The study involved 54 undergraduate students, with a gender distribution of 26 males and 28 females, and the mean age of participants is 23.6. All participants were non-native Chinese speakers and they were selected based on convenience sampling (Johnson & Christensen, 2019) . All participants were learning CFL at a university in Northeast China. Each Chinese reading class lasted for 120 minutes and the intervention period was one week. None of them had exposed to AR in their previous classrooms. Moreover, all participants passed their HSK Level 4 examination. Each student provided consent to participate, and they received 30 RMB and a certificate of participation upon completion of the study.

For answering the research questions above mentioned, this study employed a pretest-posttest control-group design (see Figure 1 ) (Johnson & Christensen, 2019) . In this paper, our main focuses were only on learners’ reading emotions and engagement. Thus, for centering the objectives of the study (see Xie et al., 2024 ), we illustrated the process as follows. First, the instructional materials in this investigation were delivered via a thematic approach, with “vacation” designated as the core theme. Prior to the classes, all participants underwent a pretest using the HSK examination (i.e., a standardized test) with and completed a questionnaire containing demographics, reading emotions, and reading engagement. The pretest results revealed no significant disparities in reading comprehension (p = 0.24), overall Chinese proficiency (p = 0.48), reading engagement (p = 0.38), and reading emotions (p = 0.44) between the two groups. The experimental group then engaged in reading tasks employing an educational AR tool known as CoSpaces, which facilitates the construction, programming, and exploration of user-generated content in AR. Throughout the study, students accessed CoSpaces on their mobile devices or tablets.

During the initial lesson, both student groups analyzed an expository article titled “My Winter Holiday,” guided by the instructor (see Xie et al., 2024 ). This phase focused on understanding the principal learning objectives of the class, mastering relevant vocabulary and phrases pertaining to the vacation theme, and learning the future tense in Chinese. In the subsequent lesson, participants

in the experimental group were tasked with developing content linked to the reading materials using the AR tool under the instructor’s supervision. The students were organized into groups of four to five to design scenarios depicting their anticipated winter vacations and elaborate on their preparatory activities. Subsequently, they were required to orally present their winter vacation projects in class with their mobile devices or tablets. Conversely, the control group adopted a traditional reading approach, where they conceptualized their ideal winter vacations and discussed them orally with their peers. Students from both groups were instructed to apply the insights gained from the first lesson to effectively present their envisioned winter vacations. After the intervention, we administered a post-test HSK reading exam and an online survey for both groups of participants. Please note that the Figure 2 shown below did not display the pre-and-post test of HSK exam. This is due to the fact that it is not our main data focus of this study (see Xie et al., 2024 ).

In order to address research questions 1 and 2, the study deployed a questionnaire for both groups before and after the AR intervention. This survey collected data on the demographic information, emotional responses to reading, and engagement levels in reading among participants. We utilized two reliable and validated measures, which were adapted from Hamedi et al. (2020) . This instrument included the English Language Learners’ Reading Emotions Scale (ELL-RES) and the English Language Learners’ Reading Engagement Inventory (ELL-REI). The ELL-RES scale is divided into three subscales, namely “Reading Boredom”, “Reading Anxiety”, and “Reading Enjoyment”. Please note that in the subscales of Reading boredom and Reading anxiety, the higher the score, the lower the boredom and anxiety. The ELL-REI scale is composed of three subscales, which are “Behavioral Engagement”, “Cognitive Engagement”, and “Emotional Engagement”. Both scales utilized a five-point Likert scale for participants’ responses, ranging from 1 (strongly disagree) to 5 (strongly agree), to measure the participants’ attitudes and experiences regarding their reading emotions and engagement levels.