Comparing L2 learning outcomes in traditional and English as a lingua franca study abroad contexts

ABSTRACT The number of students embarking on a study abroad experience continues to grow steadily. A new setting for study abroad in English has developed recently where students may be studying in an environment where English is used as a lingua franca but is not an official language of the country. This setting is referred to as English as a lingua franca study abroad or ELFSA (Köylü 2016). The present study explores the outcomes that may arise from participating in an ELFSA programme and it compares the merits of this type of context to those arising from traditional, anglophone settings. Using a written task and a proficiency test, the L2 gains of 48 Catalan/Spanish undergraduates, learners of English as an L2, were investigated before and after their international stay in either an anglophone country (n = 29) or a non-Anglophone (n = 19) one. Results show that students improved on the four measures investigated and no differences were found between the groups in terms of their L2 gains.


Introduction
Many students embark on study abroad (SA) programmes with the objective of increasing their second language (L2) skills (Mitchell and Güvendir 2021).While the Covid-19 pandemic precluded students from living their adventure abroad during 2020-2021, participation in SA programmes is slowly recovering.Nowadays, international programmes abroad offer participants the opportunity of developing their language skills and knowledge, and they may also influence other areas such as learners' level of interculturality (Heinzmann et al. 2020), or employability (Mitchell and Güvendir 2021;Mizikaci and Uğur Arslan 2019;Sisavath 2021).
Overall, SA programmes have the potential to provide students with a wide range of benefits.While there are different types of SA experiences, most research within the SA literature examines students whose international experience takes place in a country where their L2 is the official or most widely spoken language.For example, students going to the UK to learn English (Zayteva, Miralpeix, and Pérez-Vidal 2019), to Spain or Mexico to learn Spanish, or to France to learn French (Mitchell, Tracy-Ventura, and McManus 2017).We refer to this setting as 'traditional' SA (Köylü 2021).
However, when English is the students' L2, it is no longer necessary to travel to anglophone countries in order to practise the language (Glaser 2017;Mitchell 2021).Particularly within Europe, thousands of students take part in SA experiences in countries where English is not the official language even though English is their target L2 (Llanes, Arnó, and Mancho-Barés 2016).

Literature review
Participating in an academic sojourn in the country where the students' L2 is the official language (i.e.traditional SA) is argued to be beneficial because it fosters 'L2 use in the authentic culture' (Tragant 2012: 16).Moreover, it can play a role in the students' personal development and their linguistic abilities (Mitchell, Tracy-Ventura, and McManus 2017;Mizikaci and Uğur Arslan 2019;Sisavath 2021).
In terms of language development, the SA literature suggests that staying in the foreign country will be beneficial for the learners' L2 overall proficiency (Coleman 1996;Hessel and Vanderplank 2018;Isabelli-Garcia et al. 2018).As an illustration of this, the European Language Proficiency Survey (Coleman 1996), one of the largest existing projects investigating L2 general proficiency, examined the language skills of 18,264 British and Irish students, learners of French, Spanish, Germany, and Russian as an L2, over their four years at university.Coleman (1996) stated that, while changes in the students' proficiency were not meaningful within the first two years at university, those students who spent their third year abroad showed significantly higher scores than those who stayed in their home country.He suggested that a period of residence abroad contributed significantly to the students' L2 general development.
More recently, Hessel and Vanderplank (2018) examined whether changes in the L2 English proficiency of a group of 136 German undergraduates were attributable to spending time abroad.They compared a group of students who participated in a SA experience at a British university for three (n = 44) or nine months (n = 52) to a group who remained at home (AH) (n = 40).Findings from their study indicated that, after the first three months, learners abroad already significantly outperformed the AH group, demonstrating the advantages of participating in a SA experience.However, during the subsequent months, the differences between the SA and AH groups decreased and at the end of the study, were no longer significant.The authors argued that simply being abroad might not always result in L2 gains and that what actually boosts development is actively using the L2.They recommended that higher education institutions and SA organisers promote opportunities for participants to practise and develop their L2 skills and knowledge so that SA experiences could be even more advantageous.Several authors have found a relationship between participants' learning outcomes and the activities in which they take part while abroad, indicating that what learners do during their sojourns can explain their linguistic improvement, or lack thereof, at least to some extent (Coleman 2015;Isabelli-Garcia et al. 2018;Mitchell, Tracy-Ventura, and McManus 2017).
However, the research findings concerning the impact of SA experiences have to be interpreted with caution given that they can vary depending on the measures and the group under investigation (Iwasaki 2007).As an example, changes in students' oral skills seem to appear faster than changes in the students' reading or written skills (Serrano, Tragant, and Llanes 2012), and one of the few consistent findings within SA research is that there is a lot of individual variation (Iwasaki 2007).This is particularly true in the case of writing development, with some researchers reporting benefits in this area (Barquin 2012;Sasaki 2007), while others conclude that SA will not necessarily help students to enhance their writing skills (Köylü and Tracy-Ventura 2022;Serrano, Tragant, and Llanes 2012).
An example of a study that showed gains in the students' writing is that of Sasaki (2007).The author compared the learning outcomesin terms of general proficiency and L2 writing abilityof a group of Japanese students learning English abroad to those of a group who remained AH.She concluded that, while both groups improved their general proficiency, only the group abroad showed significant gains in L2 writing.Sasaki explained this finding in light of the fact that participants in her study were required to write a lot during their sojourn, whereas the students AH had far fewer opportunities to write in English.Positive findings can also be found in Barquin (2012), who investigated the writing development of 30 Catalan/Spanish bilinguals, using an argumentative essay.Results indicated that a 3-month SA experience was significantly positive for the students' written fluency (they wrote longer essays after their stay), and lexical diversity (as measured through Guiraud's Index).
In contrast, other studies show that gains in the students' writing skills are slow to emerge and will not always be a by-product of studying abroad, especially when it comes to written syntactic complexity measures.Serrano, Tragant, and Llanes (2012) investigated a group of Spanish students in terms of oral and written fluency, complexity, and accuracy, and found that gains in the written measures only appeared after students had spent two terms abroad.They concluded that students may need to spend a certain amount of time in the foreign country if one of their objectives is to improve their written skills.More recently, Köylü and Tracy-Ventura (2022) found that intensive AH instruction was significantly more beneficial than a SA experience in terms of writing development because the international students in their study failed to practise their written skills while abroad.The authors suggested that students do not necessarily increase their L2 writing mastery by travelling abroad.
In sum, findings in relation to SA and writing development suggest that variables such as amount of L2 practice and duration of the programme will be key to determining the outcomes that emerge from an international stay (Mitchell 2021).However, the few existing studies in the area do not provide clear-cut conclusions in relation to L2 writing development.Moreover, all the previously cited studies focus on students who had a traditional SA experience.Although one might assume that staying in a country where the L2 is not the official language might not offer the same opportunities for practice, this ignores the reality of much study abroad (Glaser 2017), particularly where English is the students' L2, given that this language is regularly used as a lingua franca in non-English speaking countries (Mitchell 2021;Mitchell, Tracy-Ventura, and McManus 2017).
The role of English as a lingua franca has made it necessary for universities worldwide to increase their international profile and offer opportunities to practise and study in English (Jenkins 2014).In France, for example, a programme called 'Bienvenue en France' has been created, placing emphasis on the internationalisation of French institutions by creating a great number of content courses taught in English and offering exceptional services to incoming international students (Campus France 2020).Another aspect that adds to the popularity of ELFSA programmes is the fact that they may promote multilingualism and multiculturalism by promoting the practice of languages other than English while abroad (Köylü 2016).Moreover, it appears that students in ELFSA contexts can create communities of practice in which they feel less anxious when using L2 English, and they develop their level of interculturality further (Kalocsai 2014;Köylü 2021).The ELFSA setting has started to attract the attention of SA researchers, and has led to studies on its potential linguistic benefits.
Two of the earliest examples of studies investigating the ELFSA context are those by Köylü (2016) and Llanes, Arnó, and Mancho-Barés (2016).In her dissertation, Köylü (2016) conducted an exhaustive mixed-methods study in which she compared a group of Turkish L1 undergraduates (N = 46) learning English in three different contexts: traditional SA (n = 7), ELFSA (n = 24), and AH (n = 15).Several interesting findings emerged.First, the SA and the ELFSA contexts were found to be equally advantageous in terms of linguistic gains, but only those students in the ELFSA group created an 'English as a lingua franca' identity while abroad which made them feel more multilingual after the stay.Finally, studying English AH was found to be more beneficial than doing so abroad in terms of written fluency, which highlights the fact that positive changes in some written measures will not automatically appear after being abroad, and will vary depending on the extent to which students practise their skills (Köylü and Tracy-Ventura 2022).
Llanes, Arnó, and Mancho-Barés (2016) investigated the written skills and general proficiency of a group of Catalan/Spanish bilinguals (N = 39) who travelled to different non-anglophone European countries and found that students improved significantly on two of the four measures under investigation (lexical complexity and general proficiency).They concluded that sojourns in ELFSA countries can be as advantageous as those in English-speaking countries, considering that many previous studies investigating traditional SA experiences and changes in the students' written skills had not been too positive either.However, Llanes, Arnó, and Mancho-Barés (2016) did not include a group sojourning in traditional SA contexts, so direct comparisons between the traditional and the ELFSA contexts could not be made.
Two other examples of studies reporting changes after an ELFSA experience are those by Martin-Rubió andCots (2018), andLlanes (2019).Martin-Rubió and Cots (2018) investigated a group of six Catalan students sojourning in Denmark in terms of oral fluency and accuracy, and they found that after their stay, students were able to produce speech of longer duration with fewer pauses and fewer mistakes..Moreover, the participants in their study reported feeling more self-confident concerning their English skills.However, the findings in Martin-Rubió and Cots ( 2018) are based on descriptive statistics and a very small number of participants so they need to be interpreted with caution.Llanes (2019) investigated a group of 18 participants sojourning in Italy, Denmark, Germany, Belgium, Finland, and the Netherlands in terms of oral development (fluency, accuracy, lexical richness, and grammatical complexity) and general proficiency.Participants in Llanes (2019) showed some improvement in the five measures included in their study.Nonetheless, this improvement was only significant in terms of L2 general proficiency, oral fluency, and lexical richness, but not in grammatical complexity or accuracy.
Altogether, the few previous studies examining the ELFSA context seem to offer a positive picture for the non-anglophone setting.In other words, although some may argue that the ELFSA setting will not offer students the same opportunities for L2 development as traditional SA, the research literature does suggest ELFSA experiences too can be beneficial for L2 learning.However, thus far, research in this area is particularly scant, with only a few studies investigating its linguistic impact quantitatively.Moreover, the number of studies directly comparing the ELFSA context to a traditional SA one is even lower, which makes it difficult to draw any conclusions on whether the settings are comparable.In an attempt to increase the literature regarding ELFSA programmes and to offer a more comprehensive study of this context, the following research questions guided this study: RQ1: Will semester-long traditional SA and ELFSA experiences have a positive impact on a group of English L2 learners as measured by a general proficiency test and a written task?RQ2: Will a semester-long ELFSA experience lead to outcomes comparable to a traditional SA experience in terms of L2 general proficiency and writing?

Participants and learning contexts
A total of 48 Catalan/Spanish bilinguals participated in the present study.All of them were learners of English as an L2 and they were studying at two Catalan universities to become either language teachers (n = 13) or translators (n = 35).None of them had any prior experience abroad.During their undergraduate studies, they were required to participate in a SA experience, which consisted in living in a foreign country and studying at a foreign university for one academic term.Students could choose their preferred destination and were then allocated according to their average mark.Most participants decided to travel to different European countries through the ERASMUS programme (n = 43), while five took part in a university-specific programme which enabled them to study in the USA.
Of the 48 participants, 29 completed a traditional SA experience in the UK or the USA.The other 19 decided to go to different non-anglophone European countries (see Table 1 for more information).Both traditional SA and ELFSA experiences consisted of living in the foreign country and attending classes at a foreign university for one term.In most cases, the classes that students took while abroad (irrespective of their destination country) were content classes related to their degrees, and they ranged from teaching a second language to understanding other cultures through films.Overall, no major differences were initially observed between the settings.The initial English proficiency level of the students was examined through a general proficiency test (see Section 3.2), and an independent samples t-test showed that the two groups were statistically comparable (t(49) = 1.367, p = .178).Some differences were found in the participants' use of English while abroad, as reported by the students in an online questionnaire they completed after their stay.First, students in the traditional SA group received all their content classes in English, whereas some students in the ELFSA context decided to also practise a third language (L3) and, therefore, also received classes in that language (this was the case of six students in France and Germany).Consequently, it seems that participants in the ELFSA group may have received less formal instruction than students in the traditional SA setting.However, when asked about whether they took any English language classes, six students in the ELFSA context declared that they had enrolled in a course where all skills were practised to a similar extent whereas this was only the case for four students in the traditional SA group.
All students who participated in this experiment did so voluntarily and they all signed a consent form in which they agreed to be part of the study.

Language tests
During both data collection times (before and after the stay), students in the present study were asked to complete four language tests, two of which are relevant to this study: a general language proficiency test and a written task. 1  The Oxford Quick Placement Test (OPT) was used to measure participants' general L2 proficiency.This test consisted of a battery of 60 grammar and vocabulary multiple-choice questions which measured the students' general L2 knowledge.To investigate the students' writing skills, learners were asked to write a composition entitled 'My life: past, present and future expectations'.Although this prompt may seem too simple in nature, it was decided to use it because students should be able to write about it regardless of their age or proficiency level (Köylü 2016;Llanes and Muñoz 2013).Participants were given 15 minutes to complete this task, and they were asked to write as much as possible, although no specific word limit was required because one of the goals of the project was to examine the students' fluency, that is, the length of texts they wrote (See Author, 2022).Table 2 shows that students wrote texts of different lengths both at pre-and at post-test, which led to much individual variation.
In order to determine the extent to which participants used their L2 while abroad, students completed an online questionnaire once they came back from their stay.This questionnaire was used to comprehend the linguistic changes in the students' L2 and to provide a more comprehensive picture of their SA experiences.It was based on the Language Contact Profile (LCP) (Freed et al. 2004) and the Language Engagement Questionnaire (LEQ) (Mitchell, Tracy-Ventura, and McManus 2017).However, a few questions were added and it was distributed in Catalan (students' L1) because it was believed that they would write fuller answers if they were able to use their L1.In total, there were 37 open-and closeended questions which enquired about participants' engagement with English, their use of English and other languages while abroad, their perceived improvement, and their insights and reflections on the experience.A few differences were detected in relation to whether one context triggered the use of the English (and other languages) more than the other did.Some students in non-anglophone settings took advantage of their location to exercise an additional language, which led to receiving fewer content classes in English than those students placed in the UK and the USA.However, this was counterbalanced by two facts: the number of students taking language classes was bigger in the ELFSA group, and students in anglophone countries reported greater use of their L1 while abroad.Even though these results are not discussed in any detail in the present paper (but see Author, 2022), they have been used to better understand the results of this study.

Measures and procedure
This study has a pre-post-test design.Participants were tested before they departed for their destination countries and again once they came back.It is important to note that, in order to prevent any task-repetition effects, they were not informed about there being a post-test after their stays.As previously mentioned, the two language tests that were used in the present study were part of a larger battery of tests.Hence, although students took around 40 minutes to complete the two tasks used here, the whole procedure lasted for around 90 minutes.Once they finished the language tests, students were told that they could leave the room and they were sent an online questionnaire which they were asked to return as soon as they could.
Following Llanes, Arnó, and Mancho-Barés (2016) four measures were extracted from the two tasks.First, the participants' proficiency level was examined through the OPT by using the students' raw scores on this test.Then, participants' writing was examined in terms of syntactic complexity, lexical complexity, and subordination.Syntactic complexity (SC) consisted of identifying the number of words per T-unit, and subordination (SUB) was measured as number of clauses per T-unit.A T-unit is defined as 'one main clause with all subordinate clauses attached to it' (Hunt 1965, 20).In terms of lexical complexity (other times referred to as lexical diversity), Llanes, Arnó, and Mancho-Barés (2016) and other studies within the SA field (Serrano, Tragant, and Llanes 2012;Zayteva, Miralpeix, and Pérez-Vidal 2019) have used Guiraud's Index of lexical richness.However, vocabulary researchers have argued against this measurement because it provides poor models of type/token ratios.Hence, they suggest using D-formulas instead (Jarvis 2002;Meara andMiralpeix 2016, 2018).Therefore, for the present analysis, it was decided that to calculate lexical complexity (LC), the online tool D_tools (Meara andMiralpeix, 2016, 2018) would be used in an attempt to avoid innate flaws in raw type/token ratios.The written data were transcribed and coded into CLAN (MacWhinney 2000), which was used to compute the SC and SUB measures.In order to ensure the reliability of the data, a second researcher was asked to code 10% of the data and the agreement was 100% considering that only clauses and T-units were coded for the present study.The Statistical Package for the Social Sciences (SPSS) version 24 was used for the statistical analyses of the four measures.

Results
The descriptive statistics in Table 3 show that there was a lot of individual variation (as shown in the SD column), particularly in the measure of syntactic complexity.The ELFSA group had lower means before the stay in all the measures under analysis, but these were statistically compared and results determined that the groups were comparable.The post-test scores indicate that participants in both groups made gains in the four measures under investigation.Of the four measures, two were normally distributed (lexical complexity and general proficiency) while two violated the assumption of normality (syntactic complexity and subordination).
To answer the first research question, a paired sample t-test was used for those measures that were normally distributed and a Wilcoxon signed-ranked test was run for those that were not.
In order to answer the second research question, which asked whether one context was superior to the other in terms of language learning, a one-way ANOVA was run with the normally distributed variables.No significant differences were found between the groups in terms of gains in general proficiency (F(1, 46) = .100,p = .753)and lexical complexity (F(1, 46) = .591,p = .446).Finally, the Kruskall-Wallis test was run for gains in syntactic complexity and subordination and, again, no differences were found in either of the variables (χ 2 (1) = .081,p = .776;χ 2 (1) = .423,p = .515respectively).These findings suggest that students' outcomes were not related to their type of stay.

Discussion
The research questions in the present study focused on L2 linguistic outcomes emerging from two different SA experiences: traditional SA and ELFSA.Results indicate that both contexts were similarly beneficial for the development of students' English proficiency and written skills, given that participants in both groups improved on all the measures under study and no significant differences were found between the groups.At first glance, these findings might be rather surprising given that, as Llanes, Arnó, and Mancho-Barés ( 2016) have argued, one could speculate that the ELFSA context would not provide participants with the same opportunities to receive English input as a traditional SA experience.However, in line with the previous literature on ELFSA experiences, the present study suggests that learning English in non-anglophone contexts can be equally beneficial in terms of English L2 development (Köylü 2016(Köylü , 2021;;Llanes 2019;Martin-Rubió and Cots 2018).
The finding that students made gains on all measures researched is in line with Llanes (2019) and Martin-Rubió and Cots (2018), who found that their students improved on all oral measures under their examination, at least to some extent.However, findings in the present study are particularly intriguing given that the objective was to measure students' L2 writing development, and previous research on L2 writing development has not been as uniform.Llanes, Arnó, and Mancho-Barés (2016), whose study investigated the same measures as the present one, found that participants only made gains in one of the three written measures (namely lexical complexity).The authors explained these gains in light of the students' participation in social and academic activities during the SA.However, in line with other previous studies, they suggested that the mere fact of being abroad did not require students to engage in many writing activities, which may be why they did not show an improvement in the other two measures.
SA researchers seem to agree on the fact that credit students (i.e.those who study abroad for one or two terms to earn academic credits) will not necessarily receive enough opportunities to practise and improve all their L2 skills (Mitchell 2021).While it is true that international students practise their oral skills considerably while abroad, in many cases they barely practise their writing due to the nature of the SA setting.Overall, lack of L2 written practice and exposure to L2 formal instruction could explain the results in studies such as Llanes, Arnó, and Mancho-Barés (2016) in which improvement was not found on most of the written measures.This was not the case for students in the present study, who seemed to have received enough opportunities to write for them to start writing longer and more complex sentences after their SA experience.
In line with participants in Zayteva, Miralpeix, and Pérez-Vidal's ( 2019) study, all students in the present study took content classes in English while abroad, which probably required them to frequently read and write in the L2.Moreover, some students in each group (n = 4 traditional SA; n = 6 ELFSA) also enrolled in various English language courses to strengthen their L2 abilities.In their study, Zayteva, Miralpeix, and Pérez-Vidal (2019) reported significantly positive outcomes in their participants' written productive vocabulary after a semester-long SA experience, and they explained these gains in light of the fact that, apart from simply attending lectures at a host university, students were also asked to submit different written assignments while abroad.The authors stated that it was possible that these opportunities for writing in English, together with the fact that students created new English-speaking communities, could have enhanced their L2 written abilities.As previously established, students in the present study were regularly exposed to formal instruction while abroad.Moreover, they also created their own communities in which English was the vehicular language.Consequently, it is possible that the combination of participating in L2 content classes, L2 language classes, and living in a foreign country with constant L2 contact boosted the students' writing development.
It is important to note that some students in the ELFSA group tried to develop their L3 skills and knowledge, taking some content classes in that language while abroad, which may have decreased their participation in English content classes.This was the case of six students, who were either in France or in Germany, and wanted to take advantage of their destination to practise their French or German skills.Although one might expect these students to show less improvement in English skills than their peers, findings in the present study suggest that practising an L3 while abroad does not impede English L2 development.Considering that we live in a multilingual and multicultural world where many languages and cultures coexist and students do not exclusively travel to anglophone countries or wish to develop only one L2, this finding favours the ELFSA setting (Borràs and Llanes 2022;Köylü 2021).Altogether, in line with previous research, findings in answer to the first research question suggest that, more than type of stay, what may play a role in the students' development are their efforts to use the L2 while abroad (Coleman 2015;Isabelli-Garcia et al. 2018).
Many SA researchers have indicated that what seems to shape the linguistic outcomes of the students after an international stay are their efforts to use the L2, their social networks and the activities they enrol in (Coleman 2015;Isabelli-Garcia et al. 2018).In other words, the amount of interactions and opportunities for practice that the students take advantage of while abroad will be a strong indicator of positive changes in the students' L2.Although some differences were found between the groups in terms of their English use (e.g. at university) participants in both groups claimed to have used English regularly in both types of stay.This could also have affected the positive results in the participants' L2 general proficiency given that students in both groups improved their overall English knowledge significantly.This is in line with the previous literature on SA, which has repeatedly indicated that students will improve their general proficiency greatly after an international academic sojourn (Coleman 1996;Hessel and Vanderplank 2018).This finding points to the benefits of SA not only when learners travel to a country where English is the official language but also when sojourns take place in non-anglophone countries where participants use English as means of communication (as in Llanes, Arnó, and Mancho-Barés 2016).Nonetheless, findings in the present study have to be interpreted with caution because a great deal of individual variation was also found in the results, especially in the measures of syntactic and lexical complexity (See Table 3).This variation reveals the individual differences that are often found in SA research, no doubt due to the fact that SA participants have a variety of agendas while abroad, and will undergo different, unique experiences even when placed within the same setting (Tullock and Ortega 2017).These differences will depend on their individual objectives, which will automatically determine the type and amount of formal and informal interactions they engage in, and the social networks that they create while abroad (Dewey 2004).Therefore, it seems important to explore the objectives of the students in relation to their sojourns, as that could be an indicator of their future success.More attention should be given to the students' out-of-class activities because these tend to be more valued by SA participants than formal classroom instruction and could possibly explain a great deal of the individual variation that is often found in SA research (Briggs Baffoe-Djan and Zhou 2020).
One final finding in the present study is that no significant differences were found between those learners studying in traditional contexts and those studying in ELFSA countries in relation to their language learning overall.This is in line with Köylü (2021) who compared the general proficiency development of two groups after both types of stay, and found that participants significantly developed their proficiency regardless of context.Again, the results in relation to the second research question only highlight the fact that students are likely to have the same opportunities to increase their L2 skills regardless of their type of stay.As previously stated, more than type of stay, other variables may influence the students' learning development (e.g. the students' efforts to use the L2 or the social networks they create while abroad).This offers a positive picture for SA participants, at least for those learning English as an L2, given that it confirms the value of a wider range of settings for students' SA experience.
To conclude, the fact that there is empirical research evidence that the ELFSA context can be beneficial, or at least as beneficial as the traditional SA, is positive for the SA population for different reasons.First, the UK has recently left the European Union and, consequently, is no longer part of the ERASMUS programme.Therefore, it is likely to be more difficult and costly for students to participate in SA programmes in the UK, which reduces the options for SA in an anglophone European country.Second, the new immigration policies in the USA have resulted in a drop in the number of students deciding to carry out their SA experience in the country (Campus France 2020), thus reducing the attractiveness of that anglophone setting.Moreover, travelling to anglophone countries may well be a less affordable option for some L2 learners.Finally, SA in a non-anglophone country may well offer a more propitious environment for boosting L2 communication among international students as it appears that they feel less anxious when using their L2 among other L2 speakers (Köylü 2021).

Conclusion
The results in the present study point to the positive nature of SA regardless of the type of stay that students decide to embark on.Students in the present study either participated in a traditional SA experience in an anglophone country or in an ELFSA experience in a non-anglophone European country, and all of them improved their English skills as measured by a written task and a general proficiency test.Moreover, no significant differences were found in the outcomes of the students in both groups, which suggests that both types of stay were equally beneficial.The present study also suggests that participating in an ELFSA experience offers students the opportunity to learn an L3 and this does not interfere with their English L2 development.
Nonetheless, findings in the present study are only indicative and should be read with caution due to the small number of participants in each of the settings (n = 29 traditional SA, n = 19 ELFSA).Larger scale studies would provide more determining conclusions.Additionally, a limitation of the project is that there were not enough students in each of the separate ELFSA countries to effectively examine whether the specific destination country may have an impact on students' English learning.In other words, this study compared traditional SA to ELFSA, but it did not take into account the different countries where students had their experiences.It is possible that, while type of stay was not related to the students' outcomes, destination country may have been a variable impacting on L2 development.It would be interesting for future studies to take 'country/destination' 'destination' into consideration as a variable when conducting studies on the ELFSA setting.Further, although some students tried developing L3 knowledge and skills (French or German), progress in this third language was not examined.This study calls for future investigations to also take into account any development in languages other than English given that this might highlight further positive aspects to ELFSA experiences.
Another limitation of the present study is that it did not include any control group studying AH.Even though previous research within the SA field considers that comparing SA and AH contexts is like 'comparing apples and oranges' (Sanz 2014), SA research also suggests that, in the case of writing, students AH may outperform those abroad because practice is essential to writing development but where that practice takes place may be immaterial (Köylü 2016;Köylü and Tracy-Ventura 2022).Therefore, future studies would do well to include a control AH group, at least when examining L2 writing development.In addition, the present study only used two instruments to examine the students' L2 development.It is possible that a wider range of testing measures could have captured more nuanced results.Future investigations should consider all these limitations when exploring the possible outcomes emerging from participating in any type of SA experience.Moreover, given that previous studies show that the benefits of academic mobility go beyond academic achievements (Mizikaci and Uğur Arslan 2019), future research should investigate personal growth in relation to linguistic changes.
In spite of the aforementioned limitations, the present study offers several important implications.First of all, SA participants need to be made aware that studying in a non-anglophone country can still help them develop their English skills as long as they use their English while abroad.Many students may expect to increase their L2 skills by the mere fact of engaging in a SA experience and the reality is that they will not do so unless they actually make an effort to use the L2 while abroad.For this reason, offering pre-SA preparation courses is important as these may help students avoid unrealistic expectations and understand the kinds of activities and strategies that will help them make the most of the experience.

Table 2 .
Mean length of texts (no. of words).