2.1 L2 Motivation Self System and EMI

Students’ affective responses to EMI environments might differ due to the language of instruction, and are thus worth examining. Dörnyei’s L2MSS (2005, 2009) is nowadays one of the most popular frameworks (if not the most) for motivational research. It consists of three components, namely: ideal L2 self, ought-to L2 self and the learning experience. The first dimension is linked to a learner’s image of themselves as successful, competent users of a FL and comprises internal, intrinsic motives to become one. The second dimension is related to an image of what kind of FL speaker the learner “ought to” be in order to fulfil the expectations of the society or significant “others”, and is thus related to societal and/or parental pressure, and external, extrinsic motives to succeed in learning the FL. The last dimension comprises the factors constituting the learning environment and all previous experiences of learning the FL, such as the classroom, the teacher, the group, etc. According to L2MSS, learners’ motivation to succeed in acquiring a FL is prompted by the vivid vision of themselves as competent FL users. Empirical research provided evidence that learners’ ideal L2 self is the best predictor of their motivation towards learning the FL (Kojima & Yashima, 2017; Lamb, 2017; Taguchi et al., 2009). This was found also in EMI environment (Lasagabaster, 2016), although some studies in Asian contexts reported that students’ ought-to self might have a significant effect (Rose et al., 2020). The L2 learning experience, albeit conceptually at a different level from the other two components, is also considered a powerful predictor of conscious intended effort to succeed in learning a FL (Dörnyei, 2019, p. 22). Thus, positive L2 learning experience that EMI students have had during their courses might boost their motivation better than their self-images, generated either internally or externally. This is why it is important to analyse students’ perceptions and factors that contribute to or hinder their successful EMI experience.

In general, studies conducted in EMI settings within the L2MSS framework have shown that students consider this environment motivating. Lasagabaster (2016, p. 14) surveyed EMI students’ motivation towards EMI courses in Spain and found that it was driven by their pursuing of imagined ideal L2 future selves, as well as by the EMI learning experience itself. More recently, Serna-Bermejo, and Lasagabaster (2023, p. 248) found that Spanish EMI students’ image of a future competent speaker both in professional and academic settings was the most powerful motivator to take on courses delivered through English, while they did not feel urged by their tutors or by the university itself. In a different context, Iwaniec and Wang (2023) reported findings from a survey conducted with Chinese university students and found that such instrumental motives as enhanced future job opportunities, learning subject specific language and content simultaneously, and opportunities for contact with an international community motivated them the most to undertake their studies in English (p. 1576). Factors related to their ideal L2 professional self seemed to drive their motivation to a much greater extent than their ought-to L2 self and EMI teaching practices.

Furthermore, Kopinska and Fernández-Costales (2023) analysed the interplay between motivation, anxiety and classroom interaction in EMI among students at two different universities in Spain. The findings of their cross-sectional qualitative study corroborated the prevailing role of students’ ideal L2 self as the main motivator in EMI. They also found that general anxiety-free atmosphere in EMI classroom fostered motivation and student engagement in classroom interaction, supporting positive learning experience in EMI courses. These results add to the robust empirical research that confirms the soundness of the L2MSS self-based approach, even in a multilingual context where English represents the L3 of the vast majority of students and is used as a means of instruction (i.e. EMI) and not simply in the more traditional EFL context (Lasagabaster, 2016).

Nevertheless, these studies offer a snapshot of EMI students’ motivation and little is known on how their self-images might evolve over time, or whether and in what way participation in various EMI courses during the degree can sustain or increase this initial motivation (Macaro et al., 2018; Xie & Curle, 2022). In fact, as noted by Kojima (2021), language-related challenges posed by EMI might trigger the loss of motivation, in particular among those students with limited English proficiency, in contexts such as Japan where a minimum number of EMI credits is obligatory in order to graduate. To our knowledge, only one study analysed how EMI students’ motivation evolved over the course of one semester, although it did not rely on the L2MSS framework. Zhang and Pladevall-Ballester (2023) investigated the development of 170 Chinese students’ EMI motivation and anxiety from three disciplines over one semester. The results of the survey showed that students generally displayed high EMI motivation levels accompanied by high levels of anxiety, although a slight decrease in both was observed over time.

2.2 Anxiety in EMI setting

Anxiety can have a complex and nuanced relationship with motivation. While high levels of anxiety can be debilitating and reduce motivation, moderate anxiety can sometimes act as a motivator, pushing individuals to take action and achieve goals (Welesilassie & Nikolov, 2022). Excessive FL learning-related anxiety, a common phenomenon among many FL students, might negatively affect their learning (Horwitz, 2001; MacIntyre & Gregersen, 2012) and willingness to interact in the FL (MacIntyre et al., 1998). EMI students exhibiting high levels of debilitating anxiety may experience difficulties to follow their lectures and understand the course contents, which may lead them to avoid taking part in classroom activities, especially if oral interaction is involved.

Horwitz et al. (1986, p. 128) defined foreign language classroom anxiety (FLCA) as “a distinct complex of self-perceptions, beliefs, feelings, and behaviours related to classroom language learning arising from the uniqueness of the language learning process”. They described three dimensions of FLCA, namely: communication apprehension while interacting with others, test anxiety comprising fear of failure, and fear of negative evaluation related to the concern about receiving negative comments from the others (classmates and/or teachers).

Research focused on FLCA have reported high levels of anxiety of EMI students in Asia. For instance, Chou (2018) surveyed Taiwanese students enrolled in partial EMI courses (in which Chinese was preferred for classroom interaction) and full EMI courses through FLCA scale. Students receiving partial EMI were found to exhibit higher anxiety and lower self-confidence while performing orally than their full EMI counterparts (p. 628), highlighting such factors as deficient FL mastery, issues with specific vocabulary and content knowledge. Kojima’s (2021) analysis of interviews with six Japanese students revealed that speaking anxiety had a significant effect in EMI settings; likewise, a strenuous EMI experience including work overload and language-related hurdles hampered students’ motivation. Heightened levels of anxiety in EMI were also found in Lei and Hu’s (2014) and Zhang and Pladevall-Ballester’s (2023) studies in China, as well as in Hengsadeekul et al.’s (2014) research in Thailand. As noted by the former (p. 11), these findings came as no surprise, since students in their setting faced great stress using English in the EMI class, usually due to insufficient proficiency and scarce experience in oral practice in their previous educational stages. As Lei and Hu (2014, p. 121) posited, learning content through English may generate extra study burden, which in turn might result in higher anxiety levels experienced in class, although the causal relationship could be the reverse.

In the European context, Kopinska and Fernández-Costales (2023) conducted a qualitative analysis of students’ anxiety in EMI setting, based on FLCA scale (Horwitz et al., 1986). Their findings revealed a general lack of anxiety, which may have been accounted for by the inexistence of focus on language-related issues by EMI teachers. The participants, however, did express the desire to have their language errors corrected, which seems hardly fulfilled (Doiz et al., 2019; Lo, 2015). This is especially worrying, since as noted by Yuan et al. (2023, p. 534), students’ envisioned ideal EMI course would specifically focus on the “integration between content, language, and learning strategies, and a process-oriented design permeated with a translanguaging approach”.

Having this in mind, taking on courses in English may benefit EMI students in that it bolsters their speaking skills in the FL (Chou, 2018). This is so because classroom interaction in EMI provides the students with enhanced opportunities to use English, which in turn might contribute to lowering their anxiety levels and encourage them to participate actively in class. However, students’ FLCA in a cognitively and linguistically challenging EMI environment should be given more attention, given its role in holding back their participation in class (Chou, 2018). Moreover, there is still a need for longitudinal analyses of this affective factor, and its interplay with students’ motivation (Kopinska & Fernández-Costales, 2023).

4.1 Participants and the university context

EMI courses offered at UPV/EHU are optional and pertain to different degree programmes. There are no specific criteria for the enrolment and students choose between courses offered in Basque or Spanish, or English, if available. In EMI, subject content is delivered through English, although the lecturer might make use of some translanguaging if necessary. Assessment is done in English as well; however, language is not specifically evaluated (see Doiz & Lasagabaster, 2021) and there are no general guidelines provided by the institution regarding language policy in the classroom, which is established by individual lecturers (for further detail on the language policy at UPV/EHU see Lasagabaster, 2024).

Nineteen students volunteered to take part in the study. They were all first year students at the UPV/EHU, aged 18–19, and their self-reported English proficiency was B2-C1 (three reported a C2 level). Ten students were from the Degree in History and nine from the Degree in Engineering. The choice of soft vs. hard sciences was made in order to enable comparisons between disciplines. These students were contacted as they enrolled in (one to three) optional EMI courses as part of their corresponding degree offer.

4.2 Data collection and analysis

Eight focus groups (FGs) were carried out in total; four with History students and four with Engineering students. FGs were conducted in different time points, as a means to capture the evolution of students’ motivation and anxiety over time. Specifically, the first four FGs (i.e., two with History students and two with Engineering students) were carried out in the year 2021/2022 (time 1 – T1). The second two FGs (i.e., one with History students and another one with Engineering students) were conducted towards the end of their second academic year, 2022/2023 (time 2 – T2), and the last two at the end of their third year, 2023/2024 (time 3 – T3). Nevertheless, not all students who took part in the first FGs participated in the following ones. Specifically, there were ten History students at T1, six at T2 and five and T3, while those from Engineering were nine at T1, four at T2 and five at T3. We decided to include the data from those participants who took part in FGs at least in two time points.

The students signed the consent forms to participate in the longitudinal project prior to the onset of the study. To maintain confidentiality, each participant was assigned a unique code comprising the letter “F“ for “female” or “M” for “male”, followed by “H“ for History or “E” for Engineering, and the corresponding number.

The FG interview guide was taken from Kopinska and Fernández-Costales (2023) and comprised 35 questions organised into two main parts. The first one focussed on motivation with items adapted from Kojima (2021), Ryan (2009), Taguchi et al. (2009) and Yashima (2002). The second part dealt with anxiety and was based on the FLCA scale adapted from Horwitz et al. (1986). The same interview guide was maintained in all FGs throughout the three-year period.

All FGs were carried out in Spanish. Their average duration was of one hour, the longest lasting an hour and 20 minutes and the shortest 45 minutes. They were audio-recorded and then transcribed verbatim with the aid of the free version of Speechnotes software (https://speechnotes.co/files/). Finally, they were content analysed following Dörnyei (2007; see also Cohen et al., 2017) using NVivo 14 software. Codes or labels were attached to interesting pieces of information identified in the data. They were consequently grouped into broader concepts or categories and counted for frequency. Resulting categories were then compared to exclude overlapping. The tables derived from this process are presented in the Appendix (available as online supplementary material). Quotes reported in the findings were either direct answers to the researcher’s questions, or were brought up during spontaneous discussions among participants.

5.1 Motivation: Ideal L2 self, ought-to L2 self, and intended effort to learn English

When asked about their ideal L2 self, whether they could visualise themselves in the future as competent speakers of English in a variety of settings (in the classroom, in the workplace, with friends), both History and Engineering students initially responded affirmatively. The same was observed for both specialities in the second year, whereas in the third year, a slight change emerged. While Engineering students maintained a strong ideal L2 self, History students showed a modest decline. FH1 declared she did not see herself as a competent speaker of English in the future, due to her difficulties especially with pronunciation (notice that in T1 she said she was studying to get the B2 level). Moreover, other two History students said that they could visualise themselves as competent users of English but acknowledged that their improvement in the language was not thanks to EMI courses at the university, but to out-of-university activities [1]. FH7 also added that her ideal English self, especially in the workplace, was still quite far in time [2].

[1] I don’t see myself teaching in English, maybe I need some time to prepare for it or maybe the first year I’ll do it badly, but still, with practice you get better at it. And English, it’s not that it’s all thanks to the [university] that it has helped, but I would say my English has improved by other means [outside the EMI classroom] (MH3, T3).

[2] Right now I see it as something that is far off, that is, I don’t see it as impossible, but it is something that for now I see as far off (FH7, T3).

With regard to participants’ ought-to self, from T1 onwards both History and Engineering students showed a sort of “interiorised perceived importance of” or “interiorised pressure for” learning English mainly coming from parents, as it is was summarised in quotes [3] and [4] below. That is, at the beginning parents were those who pushed them to learn, typically by enrolling them in a language school, but then they understood the importance of and the benefits associated with mastering English and chose EMI by themselves.

[3] I think that when you are younger you do... for example, they enrolled me in an academy and at that moment it was like you are ruining me... on top of that it was on Fridays for three hours... you are ruining me all afternoon on Fridays... but then in the end I even appreciated it because in the end you have a good level of English, you more or less control it, you can go from one place to another, no problem, but when you start, it’s a bit harder (FH6, T1).

[4] Pressure [from parents] when we were young, now we have internalised it, yes. [...] Then I grew up and I started to like it, and now I like it and I would like to speak it well. (FE2, T1).

The students did not seem to perceive any pressure from society; they were aware that English is highly valued, but they were also aware that its importance may vary in the work sphere (e.g. for construction workers, they said, it is not that important). Like in the case of parents, though, societal pressure as well has apparently been internalised, as clearly showed in FH4’s words [5]. This was not really seen as an external obligation but rather as an opportunity, as something that will “open doors” (FH5).

[5] No, I mean, I don’t feel pressured to say, Jeez, I have to do this. But I think that yes, well, it’s a matter of adapting because it’s the right thing to do and that’s it. In other words, I don’t have a negative attitude [towards English] that I don’t feel like it, that I have to do it. You can live perfectly well without knowing English, but well, it doesn’t hurt to have it (FH4, T3).

No changes have been identified regarding students’ ought-to self in the second year, nor in the third year. However, in the case of Engineering, participants appeared to feel higher societal pressure in T3. Specifically, FE7 expressed her concern explaining that the level of English required by society in general had increased considerably in the past few years, and such circumstance had put greater pressure on students to reach an ever-higher level of English proficiency. FE5 claimed that English is “essential” nowadays, especially as far as such fields as engineering and technology are concerned [6].

[6] Yes, obviously. It’s not me who thinks so, basically all the standards we have set to measure someone’s ability to be prepared for a job point in that direction. And evidently, nowadays, knowing English is an asset, and rather than an asset, a compulsory basic competence, especially in the field of engineering and technology (FE5, T3).

On the other hand, all students declared feeling no pressure whatsoever from their EMI teachers, and acknowledged that their teachers’ mastery of English was not high enough for them to be too demanding concerning students’ English output. As mentioned by one of the students, “as long as we participate and speak, I don’t think they care much [about our level of competence in English]” (FH10, T1). Another student noted that “what they [the teachers] are worried about is the content; they are not worried about your English proficiency” (FH2, T2), and that “there are teachers who don’t reach the level of fluency that I would expect from them” (FH6, T2). According to the participants, their EMI teachers were not the ones who put pressure on students but “I think it’s the other way around” (FE7, T3; this student declared a C2 level in T1). Students also affirmed that teachers tended to accept it when students asked them questions in Spanish or Basque.

This was clearly emphasised in T3 both by the History students on international exchange in Italy (FH4 and FH5) and those who remained at their home university (MH3 and FH7). FH5 declared that Italian EMI teachers’ English competence was pretty low too, so much so that more often than not they switched into Italian, and therefore they did not exert any pressure on students to speak in English. MH3, instead, said something slightly different; he reiterated that teachers had a low level of English but at least they tried to encourage/push students to speak it. Therefore, whereas no pressure was felt on the competence side, a gentle pressure was exerted on the usage side.

When dealing with participants’ intended effort to learn English, in the first year students from both specialities declared they did not think they invested more effort than their fellow classmates in EMI subjects. Some students explained that the need to put in a great deal of effort did not depend on the medium of instruction (English), but on the content of the subject. Additionally, FE5 recalled a physics professor telling them explicitly that he did not teach English, so they should not hope to learn English in his lessons.

Furthermore, students reported not attending any English course outside of the university, apart from two (FH1 and FH4) who were attending language schools. All of them mentioned they watched movies or series in English, or read books in their free time. They also explained they would do these activities even if EMI courses at university were not available. When asked whether they felt they strove to learn and practice English in EMI classes, all of them responded negatively. MH8 mentioned an interesting concept, “involuntary effort”, to describe his practicing of English [7], and his classmates agreed on that. All of them declared they would attend more EMI classes if they had the chance to do so.

[7] It’s almost an involuntary effort. I mean, you do it because it’s in English, but it’s some sort of involuntarily practice (MH8, T1)

But it is not a conscious effort, no, at least not for me (MH9, T1)

They also acknowledged it turns out to be difficult to loose contact with English outside the class because English is “so globally assimilated” (FH6) that it can be found everywhere, in series, videos, music, etc.

Despite their willingness to attend more EMI classes if they had the chance to do so, this choice would be to obtain an official certification (i.e., international BA) and not necessarily because EMI courses were better than those delivered in Spanish or Basque. Since they can freely choose to take on EMI courses, and “it doesn’t require any extra effort, that’s fine, if you get the recognition” (FE2, T1). Some of them planned to participate in an international exchange programme the following year, or do an internship or an MA abroad.

Finally, in T3 an improvement was detected in the effort invested in EMI classes among History students. Apart from FH1, who was on an exchange programme in another Spanish city and confessed that since her pronunciation was so bad she just avoided speaking English, the other participants’ comments were rather positive. FH4, who was on Erasmus in Italy with FH5, acknowledged that they spoke English a lot with their flatmates, and that they had to prepare long presentations in English as part of their coursework. They recommended a study abroad experience especially because they saw it as an occasion to speak more English and improve fluency in the language. No major changes were detected over the three years in the amount of effort invested in EMI classes by Engineering students. Actually, ME1 and FE5 declared that instead of taking advantage of attending EMI classes to practice their English, they nonetheless often chose to speak Basque among themselves, because they found it easier and more habitual. This circumstance is a rather clear sign that their intended effort to learn English was low. Table 1 in the Appendix offers a visual summary of the development of the ideal L2 self, the ought-to L2 self and the intended effort to learn English in History and Engineering students over the three years.

5.2 Motivation: EMI learning experience

With reference to the EMI learning experience, the following main themes emerged from the content analysis: motivating and demotivating factors in EMI, and advantages and disadvantages derived from EMI.

As shown in Table 2 (see Appendix), the main motivating factors identified within the EMI learning experience were as follows: 1) possibility of communicating with Erasmus students or with students from abroad, 2) feeling luckier than students enrolled in Basque Medium Instruction (BMI) courses (whose course load was much heavier), 3) higher degree of interaction in class, 4) absence of a final exam with a fixed format for everyone (as is usually the case in BMI courses) and have the opportunity instead to give a presentation on a freely chosen topic, 5) more homogeneous groups (age-wise), 6) smaller group sizes, and 7) teacher (his/her attitude, teaching style and age).

History students in T1 especially emphasised the importance of the teacher (11 instances), followed by interaction (4) and group size (3). In T2, the most frequently mentioned motivating factor was the absence of a final standard exam, replaced by an oral presentation (6), followed by the teacher (5), feeling luckier than BMI students (3), and group size (2). Finally, in T3, the teacher was the most mentioned factor (6), followed by smaller group size (5), the possibility of communicating with Erasmus students or students from abroad (4), absence of a final exam (3), and a higher degree of interaction in class (2). The most recurring motivating factors throughout the three years are the teacher (as exemplified in excerpt [8] below), smaller group size, and the absence of a final exam.

[8] They promote interaction a lot and people kind of follow them, it becomes much more pleasant. (FH5, T3).

Likewise, Engineering students in T1 primarily stressed the importance of the teacher (9 instances), followed by group size (6), interaction (3) and more homogeneous groups (1). In T2, the most recurring motivating factor was the teacher (5), followed by group size (4), possibility of communicating with foreign or Erasmus students (2), and class interaction (1). Finally, in T3, smaller group size was the most mentioned factor (3), followed by the possibility of communicating with international students and class interaction (2 instances each), and the teacher (1). The most mentioned factors throughout the three years are the teacher -as shown in excerpt [9] below– and smaller group size.

[9] A good teacher. It’s the same as always, if you like the teacher and you may not like the subject but s/he can make it passable, but on the other hand if you don’t like the teacher, even if you like the subject, you’ll be... (FE7, T1).

Regarding the demotivating factors displayed in Table 3 (see Appendix), the following were mentioned by the participants: 1) being made fun of by classmates when intervening in class, 2) lack of interaction, 3) not understanding what was being said in the lesson, 4) teacher’s attitude and teaching style, and 5) excessive effort required.

In T1, History students mainly emphasised the role of the teacher (18 instances), followed by difficulties in understanding parts of the content/language due to mispronunciation (9), lack of interaction in class (5), and being made fun of by fellow students when intervening in class (1). In T2, the only demotivating factor mentioned was the teacher (2). Finally, in T3, the teacher was the most mentioned factor (6), followed by troubles with understanding (3). The teacher is therefore the most cited demotivating factor throughout the three years. This is clearly reflected in the excerpt [10] below:

[10] For example, at the beginning of the course I was really looking forward to the [name of subject] course, and now I can’t stand it. It was the subject I expected to like the most, and now it’s the one I like the least. So I think that the teacher makes a big difference... (FH10, T1).

As for Engineering students, in T1 they especially stressed the demotivating power of not understanding (10 instances), followed by the teacher (7), lack of interaction (2), and excessive effort required (1). In T2, the most demotivating factor was again not understanding (4), the teacher (2) and excessive effort (1). Finally, in T3, not understanding was once more the most mentioned factor (7), followed by lack of interaction (4), the excessive effort required (2) and the teacher (1). The most recurring demotivating factors throughout the three years were problems with understanding a key concept or a word (as shown in excerpt [11] below) and the teacher.

[11] I do understand the vocabulary but some words are mispronounced. So I’m still thinking the whole class, what could that be? And I say, ah! You know? So I’m still thinking about it the whole class and I don’t listen to the other stuff. Because I didn’t understand that part (FE3, T1).

Regarding the advantages and disadvantages derived from EMI, the students mentioned the following (see Table 4 in the Appendix): 1) greater attention by the teacher to individual students, 2) learning disciplinary vocabulary in a third language, 3) official recognition of having studied through English, 4) possibility of getting rid of the heavy course load and exam in Basque, 5) opportunity to practice English to avoid losing proficiency, and 6) use of a wider range of resources beyond the usual ones in Basque and Spanish. Four students mentioned “no advantage” of EMI, and one mentioned EMI having detrimental effects on students’ English competence on the disadvantages side, as reflected in the following passage [12]:

[12] And do you think that the classes you have in English help you to improve? (Interviewer)

Well, I think the opposite, actually. Because I’m starting to question things (FE7)

Really? But why is that? (Interviewer)

Pronunciation and so on. It’s just that when they repeat the same word five times, you end up internalising it. And it’s happened to me a couple of times this year. And well, not grammar, because my ideas are quite fixed, but pronunciation. The verb ‘determine’. ‘Determine’ all the time. ‘Determine’, ‘determine’ [/deter’main/]. I’ve reached a point where I think it’s like that. And it’s not. I’ve already got it kind of stuck in my head (FE7, T1).

While the most recurring advantage mentioned by History students throughout the three years is the possibility of getting rid of the heavy course load and exam in Basque “because those who do [the subject] in Basque are exploited, they do tasks all the time” (FH10, T1), the possibility of learning specific disciplinary vocabulary in a third language was the one mostly stressed by Engineering students [13].

[13] For me like the improvement is mostly in the technical vocabulary that you learn little by little. In terms of communicating and so on, well, after all, practice helps. But it hasn’t been a drastic improvement either (FE5, T3).

5.3 Anxiety: Communication apprehension, Test anxiety, Fear of negative evaluation

When analysing the Communication apprehension questions of the FGs, five main themes emerged from the content analysis (see Table 5 in the Appendix). Participants reported 1) feeling uneasy or uncomfortable due to pronunciation issues, 2) fear of not expressing content properly, 3) the presence of interlocutors or listeners (e.g., their fellow classmates), or 4) a personality trait that made them uncomfortable speaking in general, regardless of the language used. No feelings of uneasiness specific to EMI classes were reported, and this was in fact the most frequently mentioned issue.

As can be observed in Table 5, both History and Engineering students mostly declared not experiencing anxiety in EMI classes. When they did perceive some uneasiness, it was linked to the general fact of having to speak in front of others, regardless of the language [14].

[14] Yes, but not because it is in English, but in any language. In any subject, [it is a matter of] personality, it has nothing to do with the language (FH7, T1).

Over the three years, no significant variations were detected in this dimension. The only notable changes in both groups were a slight decrease in the absence of anxiety feelings in T3 compared to T1, and the lack of mention of uneasiness derived from speaking in general.

When examining the data involving comments on Test anxiety, several themes emerged, as shown in Table 6 (see Appendix). Among the reasons students mentioned for their uneasiness in test situations were 1) general insecurity about being able to express content properly, mainly due to not being able to improvise in English, 2) the higher expectations they felt were placed on EMI students, and 3) discomfort with oral presentations due to having to speak in front of the teacher and their classmates. Consistent with their comments on Communication apprehension, the most frequently mentioned issue was not related to using English to communicate, but to other factors (e.g., disciplinary content, people listening). Students reported having no concerns about exams in English or making language mistakes, as they explained that, since teachers also made language mistakes, they did not feel judged. They also expressed no concerns about oral evaluations, with a few welcoming them as opportunities to learn from mistakes and improve.

Table 6 reveals that the majority of coding instances were found in T1, especially among History students, whose most frequent reason for perceiving uneasiness in test situations was not English per se but other factors, as reflected in the excerpt [15] below:

[15] I think the presentation is the worst (... because in an exam I do my own thing, you know, I’m only going to see it myself, I only know if I’m doing it right or wrong, but in a presentation I’m talking to everyone, it gives me more...I don’t know... (FH1, T1)

Also in T2, History students expressed the greatest concerns, particularly emphasising anxiety due to not perceiving themselves as able to express content properly and not being able to improvise when using English [16].

[16] The presentations in English. I mean, I get super tense... But it’s not like I’m going to prepare for a week or anything like that. I mean, I don’t like presentations in English [...] Because in Basque I’m completely free of... I know the theory and I can improvise or whatever. In English it’s like... well, I have to think [much more]. (FH4, T2).

Over the three years, the most noticeable variations were detected among History students, who, in T3, seemed to perceive a lower degree of test anxiety than in T1 and T2. Although Engineering students registered low test anxiety in all three time periods, a slight decrease of such feelings was detected from T1 to T3.

Regarding the Fear of negative evaluation dimension, two categories of themes emerged from the content analysis (see Table 7 in the Appendix). On the one hand, students mentioned the occasions when they felt uneasy due to fear of negative criticism; on the other hand, they explained the circumstances in which they felt no concern. Among the former, participants mentioned feeling uncomfortable speaking in front of classmates because they thought their peers had a higher command of English, due to interlocutors’ excessively high expectations, when having to ask a clarification question to the teacher, or when failing to understand what the teacher said. Consistent with their other comments to questions dealing with anxiety, some students also mentioned that their uneasiness did not depend on using English per se but on other factors (e.g., teacher’s attitude, content, having to speak in public regardless of the language used). As for the latter, among the occasions when some participants did not feel concern were speaking in front of potentially more English-proficient peers, not understanding what the teacher said (because they could ask questions), the speed of delivery of EMI lessons, and the teacher’s corrections of their English (especially because teachers tended not to correct them).

As shown in Table 7, there is some variation between the two disciplines, since at T1 History students do not seem to feel anxious while interacting in EMI class influenced by possible criticism on the part of fellow EMI students, while Engineering students do. In fact, the most frequently mentioned cause of discomfort was detected in T1 among Engineering students, who reported feeling uneasy speaking in front of peers since they believed their classmates had a better command of English. However, by T3, the situation had evolved, and these same students claimed they had no concerns when speaking in class, regardless of the mistakes they made, as reflected in excerpt [17] below:

[17] I consider that I don’t usually make them [mistakes]. If someone does, I notice them. I don’t judge them for it. If I make mistakes, well, I don’t care, especially if the teacher usually has a lower level than the students and 8 out of 10 sentences have some grammatical error. Well, I don’t care, but personally it doesn’t get me out of the subject (FE5, T3).

In contrast, History students reported having no concerns when speaking in front of potentially more English-proficient classmates since T1.

Over the three years, no important changes were detected in this dimension. The only notable development is the one mentioned above, i.e. Engineering students’ change of attitude towards peers’ evaluation.