Given the large volume of data and the extensive predefined categories used in this scoping review, a quantitative analysis was deemed more appropriate to systematically map the research landscape on the pedagogical use of AVT in LE. This approach aligns with the research objective of providing a comprehensive overview and identifying gaps, as it enables efficient identification of patterns and trends across a broad set of studies. Furthermore, the complexity of the dataset made qualitative analysis impractical without oversimplification, ensuring that key findings are accurately captured.
Working from our research questions and data charting, this section first reports the results concerning the publication characteristics of the 162 studies, followed by contextual factors, and, finally, the research design.
The dataset comprises a total of 207 unique authors and 162 publications, of which only 12 are book chapters. Twenty-four publications are written in Spanish and the rest in English.
Between 1981 and 2013, publication rates were either low or absent for a number of years (see Fig. 2), indicating a nascent field. Subtitling for pedagogical purposes started earliest, in 1981 (Lambert et al., 1981). The first study on using didactic dubbing in LE appeared in 2006 (Burston, 2006), on AD used passively in 2015 (Sadowska, 2015), and on didactic subtitling for the deaf and hard of hearing (SDH) in 2019 (Talaván, 2019). From 2014 to 2020, publications increased gradually; however, since 2021, there has been a sudden and consistent rise, with no fewer than 20 publications annually.
While Table 1 lists the top five journals with more than three publications, Fig. 3 provides a broader perspective by illustrating the distribution of publications across research areas, offering complementary insights rather than overlapping content. Translation (30.25%) dominates, followed by LE (25.93%) and linguistics (22.22%).
In analysing yearly trends in the top five journals, all 11 articles in Parallèles were from a 2024 special issue on AVT and media accessibility in LE, excluding one on translator training. Similarly, five articles from a 2023 issue of Translation and Translanguaging in Multilingual Contexts addressed the same topic.
As for participating countries, whether conducting research or being under study, a study may be funded by one country, with authors developing their research activities in different locations, who also aim to extend the scope of their research to other territories. For example, Sokoli, Zabalbeascoa, and Fountana (2011) conducted a study where two authors were affiliated with a Spanish university and one with a Greek university. As part of the EU-funded LeVis project, this study involved six countries and eight languages under study. This example highlights the advantage of international collaborations in enhancing the scale of research.
In total, 29 countries conducted research on the pedagogical use of AVT in LE as the affiliated countries, while 30 countries were studied as the target countries, with 27 countries overlapping in both roles. For the subsequent analysis, the focus will be on the affiliated countries, as the minor numerical differences among the overlapping countries are negligible, and the consideration that a country's interest in a research area is crucial for its further development. It should be noted that the data regarding the number of publications is for reference only, as a publication may involve authors from different countries, but each has been counted separately to obtain the final number for each continent or country.
Spain leads by a significant margin with 79 publications (see Fig. 4), followed by England with 23 and Italy with 14. While these numbers are considerably lower than Spain's, they still represent a significant contribution compared to other countries. Ireland, China, Poland, the USA, and Belgium have comparatively similar and lower publication numbers, ranging between 7 to 10. Of the eight countries with the most publications, only China and the USA are non-European. All other countries have fewer than three publications.
The data was then examined, categorising the countries by continent. Europe had the most affiliated countries (12), followed by Asia (7), North America (3), South America (2), Europe/Asia (2), Africa (2), and Australia (1). The percentage of publications by continent was calculated against the total number of publications (see Fig. 5).
In total, 155 studies were conducted solely within their corresponding continents, while seven studies were the result of intercontinental collaborations between 2021 and 2024, four of which were between an Asian and a European country. These collaborations, which began during the surge of research on the pedagogical use of AVT in LE since 2021, were relatively isolated.
Europe accounts for most publications, with 129 out of a total of 162, representing 79.63% of the total. The distribution of studies across Europe is one of the largest, involving 12 affiliated countries, as shown in Fig. 4.
Asia has 19 publications, making up 11.73% of the total, representing a notable but significantly smaller contribution compared to Europe. In Asia, the highest concentration of research is in China, with 10 publications.
The moderate number of publications from North America suggests a slow pace of research, with major contributions mainly from the USA. The even lower numbers from Africa, South America, Australia, and Europe/Asia suggest that these regions either have fewer resources allocated for research or face other challenges limiting their research output.
Data was also computed on the number of funded publications and the number of unique affiliated institutions on an annual basis, as illustrated in Fig. 6. There is a clear overlap between the number of publications and the number of affiliated institutions until 2011, after which the blue line begins to diverge and becomes clearly distinguishable.
In total, 115 institutions have contributed to research on the application of AVT in LE. Out of the 162 publications, 66 received funding, while no data is available for the funding status of eight publications. Out of the 66 funded publications, 58 were supported by research projects, involving 26 distinct research projects. Notably, the TRADILEX project, led by UNED, produced 26 publications between 2021 and 2024, in collaboration with 15 universities from four countries.
The first publication with funding appeared in 2005. No further funding was provided until 2011, with interruptions in 2012, 2013, and 2015. However, the impact of the 8 entries with missing data, sporadically distributed across different years, is deemed minimal for the overall data analysis.
Figure 6 also reveals a strong positive correlation between the number of publications, affiliated institutions and funding, underscoring the continuous growth and development in academic research activities related to the pedagogical use of AVT in LE.
As shown in Fig. 7, experimental studies dominate the field, representing a significant 77.16% of all publications. This suggests that experimental methodologies are the most commonly employed design in this area of research. Applied studies make up the second-largest category, 16.67%, emphasising the practical application of research findings. In contrast, review studies (4.32%) and descriptive studies (1.85%) constitute a much smaller proportion of the research, indicating potential areas for further investigation. After cross-checking the data with information on AVT mode and its active or passive use, it was found that among the seven review studies, five focus exclusively on subtitling, and three examine only the passive use of AVT. Additionally, only three systematic reviews examined in this paper provide a meta-analysis, all of which focus solely on the passive use of subtitling (Montero Perez et al., 2013; Reynolds et al., 2022; Alotaibi et al., 2023), with Reynolds et al. (2022) also incorporating a scoping review. The remaining studies follow a narrative literature review approach. This underscores the lack of a comprehensive review of research on the application of AVT in LE. The absence of theoretical studies suggests that this approach is not currently a focus within the research landscape.
In terms of educational settings (see Fig. 8), higher education dominates with 62.6% of all publications. Those without a specified educational setting account for 17.28%, consisting of 17 applied studies, 5 experimental, 5 reviews and 1 descriptive. This is understandable because for non-experimental studies, specifying the educational setting, target agent, or target language is often unnecessary due to the nature of these research types.
Moreover, secondary education shows significant interest, making up 9.26% of publications, while primary education (5.56%) and combined settings (3.09%) are less common.
In terms of learning environments (see Fig. 9), formal learning dominates, accounting for 77.16% of all publications, highlighting a strong focus on structured educational settings. Formal settings are primarily associated with the application of DAT, with only 17 studies focusing on the passive use of AVT. Incidental learning comprises 6.17%, reflecting some interest in informal or unplanned learning contexts. Of the 10 studies explicitly addressing incidental learning, all focus on passive use, with 5 specifically examining subtitling.
As for the learning dimension (see Fig. 10), most publications focus on learning outcomes (72.22%), underscoring their importance in research about the application of AVT in LE. A notable proportion of studies examine the methods and strategies used in learning (21.60%), highlighting the interest in how learning is facilitated. There is some research on the learning process (6.17%), but it is much less prevalent compared to the other dimensions. The 35 studies on learning approaches include all three descriptive studies, four of the seven review studies, 27 applied studies, and one experimental study from the perspective of the participating teachers. All 10 studies focusing on the learning process use an experimental design.
Regarding the focused agent -- the subject under study (see Fig. 11), students are the subject of 75.93% of all publications, reflecting a significant emphasis on student-related research in this field.
To explore the relationship between the nature of the study, the learning dimension, and the agent, the relative frequencies between them were calculated to identify the most common combinations. Since these values cannot be ranked, a potential correlation could not be calculated.
Experimental designs are primarily linked to learning outcomes and students, accounting for 105 out of 162 studies (64.81%). Notably, the learning process has only been studied through experimental designs with students as the agents (6.17%). Applied studies are mostly connected to learning approaches and unspecified agents (10.49%), with no focus on the learning process or outcomes. Descriptive studies relate exclusively to learning approaches, with one case each for unspecified agents, students and teachers, making up 1.85% of all studies. Review studies are associated with learning approaches and unspecified agents (2.47%) or learning outcomes and students (1.85%).
Regarding the AVT mode (see Fig. 12), subtitling dominates, comprising almost half of all publications (50.62%), indicating it is the most researched mode in the context of AVT in LE. The second-largest category, combined modes, accounts for 19.14%, reflecting notable interest in integrating multiple AVT techniques. It is important to note that which modes were included under "combined" was not specified in this analysis, as their varied combinations would considerably complicate a general overview. In some cases, only two modes, typically subtitling and dubbing (Fernández-Costales, 2021), were involved, while in others, more modes were used (Plaza-Lara & Bobadilla-Pérez, 2024).
AD represents 14.20% and dubbing, 12.35% of all the publications. While these are less common than subtitling and combined modes, they still hold significant interest. This is notable given the relatively recent rise of using dubbing and AD for didactic purposes (see "Publication characteristics"). The lower number of publications on SDH (3.09%), which first appeared in 2019 (Talaván, 2019), and the single publication on free commentary (Lertola, 2021) -- highlight niche areas that may warrant further research.
A total of 19 target languages have been studied by far, with English leading at 99 publications, indicating a strong preference or necessity for English (see Fig. 13). There is a sharp drop after English, with Spanish at 21 and much lower numbers for other languages. Most languages have only one or two publications, reflecting notable linguistic diversity, but with relatively few studies for each language. Additionally, 17 publications do not specify their language.
After cross-checking the "not specified" metadata, it was found that 10 of the 17 publications are applied studies proposing a model for applying AVT in LE. All 7 review studies are labelled as "not specified," despite a few mentioning some specific target languages. Of the 19 target languages, 11 are Asian, with four focusing on Chinese and one on Russian. The rest focus on Western languages, excluding the 17 unspecified studies. A more precise geographical comparison is difficult since languages like English and Spanish are spoken in multiple regions across continents.
Regarding the use of AVT (see Fig. 14), active use of AVT -- DAT -- dominates, accounting for 64.20% of publications, highlighting a focus on methodologies where AVT is actively applied. Passive use represents 28.40%, indicating notable interest in AVT as a secondary or supportive component. The smaller percentage of combined use suggests that integrating active and passive methods is less common but still present.
The features examined in these 162 studies are quite diverse (see Fig. 15), with the total adding up to 169 as seven studies covered two features. Integrated skills dominate -- however, the specific combinations of skills are not detailed in this study due to its complexity and broad scope. They account for 30.86% of publications, indicating a focus on holistic approaches. Vocabulary follows at 14.20%, with speaking at 10.49%. Writing and listening both account for under 10% but over 5%. The "other" category (9.26%) includes skills outside our classification. Publications without a specified feature make up 7.41%. The remaining skills have very few publications, and data is lacking for two studies due to limited access.