Emotional and Social Engagement in the English Language Classroom for Higher Education Students in the COVID-19 Online Context

: The COVID-19 pandemic has led to perspective shifts in the approach of English language classrooms in the online higher education context. The current empirical study aims to understand the behaviour of 394 university students enrolled in agricultural, veterinary, and overall life science programmes in one of the main Romanian universities while learning English as a Foreign Language (EFL) and English for Speciﬁc Purposes (ESP) in terms of their classroom engagement during the 2020–2021 academic year. Following the application of an online questionnaire, a principal component analysis (PCA) was undertaken to determine the factors that determine students’ engagement (predominantly emotional and social), while learning EFL and ESP in the online system during the COVID-19 pandemic. Furthermore, the PCA results were used to identify clusters of students expressing similar behaviours towards the English learning process. Two clusters were identiﬁed: Cluster 1 consisted of moderately engaged students, and Cluster 2 consisted of highly engaged students. The clusters differed by the anxiety students express during class, the perception of the online system, the level of involvement and enjoyment, as well as class environment and dynamics, with the second cluster being the only one revealing a difference between the face-to-face and online experience. This study on students’ behaviours, attitudes, and preferences places students in the centre of the process of foreign language teaching and learning, as their enjoyment and engagement lead to a less stressful environment, especially during challenging times, and ultimately to long-term language proﬁciency and sustainable educational development.


Introduction
Starting in 2019, the Coronavirus pandemic has plagued humanity high and low, as our perceptions of community, society, group, and even family have been altered. Therefore, it is only natural that our view on education has conversely changed, towards what has been considered the revolutionizing of the entire higher education system [1]. As part of worldwide strategies to limit the spread of the Coronavirus, policy efforts have been made to resume teaching activities online or by applying hybrid models almost overnight all over the globe [2]. Therefore, with the advent of the COVID-19 pandemic came the advent of the "new normal" higher education paradigm, which has ceased to predominantly rely on century-old face-to-face interaction but rather on a non-linear mixture of face-toface, online, and hybrid/blended models, which are bound to become as mainstay as the pandemic context itself. Online and blended/hybrid education models carry their significant advantages, in their increased accessibility and adaptability and in their student autonomy and independence. However, global policy makers, universities, members of academia, and students have also been faced with a plethora of challenges, from limited infrastructure [3,4] to preparedness of academia [5], student motivation [6], collaboration, interaction [7] and engagement [8][9][10], as well as learning effectiveness [11,12] in these "invisible classrooms" online [13].
There is increasing debate on the advantages, as well as the drawbacks, of the online system over the conventional face-to-face system. On one hand, the conventional system tends to be more teacher-centred and gives way to passive learning, whereas the online system is bound to be more student-centred, where students need to be engaged and set the pace [8], but also where students are less-collaborative and less personal [10]. Blended systems of learning have also been favoured, depending on the context [14].
Thus, research on how students perceive the new learning context is paramount in achieving a learning-conducive environment where student engagement is achieved and leads to their better academic performance and to a decrease in their dropout rate, as well as in enhancing belonging and fighting isolation towards students' personal development and well-being. Moreover, student engagement, although not exclusively so, is key in reaching sustainable learning outcomes in the online English learning context. The meeting of students' both academic and emotional needs and expectations for social belonging lie at the basis of sustainable educational development in the university environment.
This "new normal" context drives changes in higher education approaches, both overall as well as in the foreign and English language classrooms. The dimensions of online and blended/hybrid models in English as a Foreign Language (EFL) and English for Specific Purposes (ESP) learning during the pandemic have only begun to be investigated in the higher-education context, and therefore, there is a knowledge gap in this respect.
Within this framework of reference, the aim of the current study is to understand students' behaviours while learning EFL and ESP during the COVID-19 pandemic period. To achieve this aim, an empirical study was carried out on university students enrolled in agricultural, veterinary, and overall life science programmes in one of the main Romanian universities. The following research questions were addressed: (1) What factors determine students' engagement while learning EFL and ESP in the online system? (2) What is the students' typology based on the factors that influence their engagement while learning in the online system?
The remainder of the paper is organized into six parts. Section 2 synthesizes the relevant literature on learning and engagement-related variables, such as classroom environment, anxiety, enjoyment in the language online setting. Section 3 reveals the methodology used, and Sections 4 and 5 focus on results and discussions. The paper ends with the conclusions.

Learning and Language Learning Theories
How people learn is multifaceted and, it involves interdisciplinary dimensions related to psychology and education sciences, sociology, neuroscience, biology, and computer science [15]. The past century has given rise to explorations in behaviourism, particularly in how students behave in the process of learning, as they are the passive recipients of a range of stimuli used to achieve a desired outcome, which is ultimately language knowledge [16,17], predominantly acquired from the environment [18]. Cognitivism [19] considers the brain's cognitive processes, such as motivation and imagination [20], which come into play as central and explanatory for the learning process, allowing the learner to actively think about, participate with, and make sense of the language learning process [21]. Notably, there is the social constructivist theory, regarding teaching and learning as complex interactive social phenomena [15], which are facilitated by the teacher in a social environment or by a community designed for fostering problem-solving skills [22,23], where social experiences, collaborations, and reflection lead to practical and (inter)active learning [24]. Learning environments exploit different learning purposes, but the students' views on the activity is of the utmost importance, in terms of its usefulness for the ultimate communicative purpose [25]. The idea of communities of practice and social learning systems is further explored by Wenger [26], who explains that the learners' social, cultural, and contextual conditions are key in learning.
Significant consideration has been given to the advent of Carl Jung's views on individual personality traits (extroversion and introversion, for instance) and the part they play in all aspects of human life, including learning styles [27,28] towards dynamic teaching models and environments where efficient learning is achieved for all personality types alike [29,30], including in language acquisition and performance and particularly in student speaking performance [31][32][33][34] and classroom involvement [35], as well as preferences for different types of strategies and methods to be used in language learning [36].
Additionally, it is worth mentioning that, in the plethora of existing learning theories, adult learning theory-andragogy [37,38]-involves dimensions which are ultimately distinct from pedagogy (child learning). It is self-directed learning, where control lies with the learner and their experience, integrated in their social and professional contexts to enhance their professional abilities and goals. This is especially relevant in higher education, where career goals have been set and where learners are academically motivated, and as adults, they are problem-and relevance-centred [39] and thus able to express their opinions and preferences related to the content, methods applied, and classroom environment.
Attempts have also been made to design theoretical frameworks for online education alone, even before the pandemic context, but it has proven to be a difficult endeavour. However, notable to the field are certain theoretical attempts that capitalize on the importance of simultaneous environment-, learner-, and community-centredness in online learning [40], connectivism and network learning [41], and notably online collaborative learning [42], in an environment that fosters joint learning communities where social collaboration is the norm facilitated by the teacher.

Student-Related Variables: An Overview
Over the past few decades, there have been momentous attempts at investigating student-related variables in Second Language Learning (SLL), Foreign Language Learning (FLL), and English as a Foreign Language (EFL) at the higher education level, but not exclusively. These investigations have covered such two-fold pillars as the interpersonal and the intrapersonal [43][44][45] or the learner-internal and learner-external [46] variables, only to be ultimately linked to the learning process, academic success [47], and emotional well-being [44,45,48].
In this corollary, affective variables [32,49,50] are considered key to academic success, integrativeness, and motivation [51]. These affective variables have been centred around personality traits [33,43,52,53] that help students be proficient language learners and users. Central to these investigative approaches is Foreign Language Anxiety (FLA), a term which was coined in the 1980s [54] centred on psychological and emotional variables in the language acquisition literature [44,[55][56][57]. They were long considered unreliable in terms of their connection to proficiency in language learning, particularly due to their individual variance and personal dimensions [43,53,58].
However, with the advent of Positive Psychology in education [45,48,59,60], there was a momentum shift for these variables towards a more predominant focus on Foreign Language Enjoyment (FLE) as a predictor for confidence, success, and well-being in the foreign language classroom alongside other favourable, interpersonal, and intragroup motivation variables. The classroom environment has been considered influential for enjoyment-anxiety, engagement, and ultimately academic achievement [43,61] centred on intra-individual situational variation and the contributions of society, the atmosphere [62], teachers [46], encouragement, as well as peers [63,64]. Moreover, other sociolinguistic variables [53,55] were considered to be similarly influential in the foreign language learning context, such as gender [43,63,65,66], age, and university major [66] as indicating variables of FLA and FLE.

Learning Engagement
Learning environment and learner engagement are intrinsically connected. It is each teacher's desideratum to achieve student engagement, particularly because it is considered that motivation and engagement are strong predictors for academic performance and learning efficacy [67]. Certainly, the online environment is more flexible in terms of participation and removing financial, time, or location constraints, but concerns need to be raised in terms of motivation [6], interactivity, communication [9], involvement, engagement, and even student performance considerations [68].
Engagement is defined as a state of heightened attention and involvement in which participation is reflected [69]. Certainly, the term itself implies that the learner, apart from the enjoyment and satisfaction derived from the learning process, finds the learning process motivating, assigning meaning, purpose, and usefulness to it, as well as connection in social relationships within groups of individuals [70]. Studies in foreign language classroom engagement [71,72] have explored classroom engagement as behavioural (how students pay attention towards completing tasks), cognitive (how they actively think, make connections, solve problems, and answer questions), and emotional (which is external and internal, and it amounts to students enjoying the classroom environment and learning materials). These three types of learning engagement have been linked to EFL learning in online higher education [8]. Social engagement is added to the behavioural, cognitive, and emotional trifecta [73,74] in the language classroom [75], because it refers to group and classroom dynamics and regards engagement as an expression of subjective interpersonal relationships existing during classroom interaction. The behavioural dimension is at times capitalized upon more significantly [69], as the more behavioural that engagement is, the more positive of an impact it has on the learning effect, whereas emotional engagement, as positive as it may be, is not a predictor of the ultimate learning effect [76].
However, a body of literature shows that psychological and emotional factors affect learning engagement, revolving around the key variables of enjoyment and anxiety. Affective dimensions, particularly enjoyment, are of relevance to predict determination as well as positive and enthusiastic engagement [77]. It is particularly Foreign Language Enjoyment (FLE) that fosters motivation and engagement in the learning environment as well as ultimately well-being, hope, empathy, and mindfulness [48,78]. Foreign Language Anxiety (FLA) includes worrisome-, negative-, and fear-related emotions associated with learning or using a language [54,79], and it significantly impacts language class involvement, achievement and efficiency [35]. Additionally, Foreign Language Classroom Anxiety (FLCA) includes self-perceptions, beliefs, feelings, and behaviours in the unique classroom learning environment [54], such as communication apprehension, fear of negative evaluation, and test anxiety.
Convergence towards the fundamentals of Positive Psychology presents the classroom environment as paramount [80] in foreign language learning, as encouragement from teachers and peers contributes to determining levels of anxiety and enjoyment and to securing engagement mediated by the mode of interaction in this environment [81]. This is where the dimensions of the online environment are challenged and challenging on multiple levels.
Overall, perceptions on student engagement in the online environment have been positive, as students are able to resume behavioural engagement and complete tasks, but they do not manage to achieve cognitive, emotional, or social engagement [7] as a result of the environment not allowing for the eliciting of a response to learners' psychological needs [82]. Behavioural learner engagement has been approached at the task level, with studies finding that cognitive social and emotional engagement can be achieved in EFL classrooms, particularly in those with goal-oriented tasks that are key for achieving learner engagement [83]. Behavioural learner engagement was shown to be task-dependent (nature, familiarity, purpose, or lack of difficulty of the task), but it is lacking when there is no social cohesion in the classroom or emotional factors involved [84]. These findings reveal the importance of the affective and emotional dimension, as psychological needs are highly dependent on teacher-learner dynamics, although to a lesser extent than personal fears and anxieties. In higher education, alongside behavioural and cognitive engagement, it was found that emotional and social engagement residing in learners' responses to teachers, peers, course content, and a sense of belonging in the campus community and in life is a prerequisite for engagement in activities and tasks and is ultimately key for academic achievement [56]. Additional studies suggest that the role of the teacher and teacher-student dynamics are crucial in analysing and adapting teaching methods and mediating the classroom environment in a non-stressful manner [85]. In this respect, efficient learner engagement and teacher-learner congruence and dynamics can also be achieved by providing (corrective) feedback towards language improvement and proficiency [86].
Additionally, when peer cooperative learning techniques are used, such as peerassisted learning [64] in higher education and familiarity in the ecology of the classroom with peers, anxiety is limited [61]. These are the favourable social conditions and social climate that influence confidence in speaking, alongside attitudes, common experiences, interpersonal and intergroup motivations that add to the above-mentioned individual personality, and emotional traits in learning engagement [87]. These extrinsic factors are called intra-individual variations, including cultural and cross-cultural differences, sociobiographical factors, and even the broader social context [44,63].
Moreover, if there is control over what tasks to perform and what topics to discuss, students are more engaged [88][89][90]. Further studies on higher education learning engagement showed that, when the content is familiar, personally relevant, and not too difficult, engagement is heightened [90] as opposed to anxiety upon solving difficult or unfamiliar tasks.
The measures used in this research were based on the Foreign Language Enjoyment scale [46,62,91,92], which currently includes 10 items from the primary FLE scale [44,63] out of the initial 21 items. It also includes 6 items pertaining to the Foreign Language Classroom Anxiety Scale [54] to assess personality and to assess emotional and sociolinguistic variables in foreign language anxiety and enjoyment, as well as ultimately in motivation, engagement, [48] and language performance. The FLCA scale has been applied to Western subjects more predominantly, but it has been adapted to other cultural and societal contexts, for example, to Saudi Arabian learners of English [53]. Similarly, the FLE scale has been used in various contexts, including in higher education studies and on West Europeans, Asians, North Americans, South Americans, and Arabs [44], with more country-based research in the Iranian context [45,61], the Chinese context [62,93] in both secondary and tertiary systems, as well as the Eastern European higher education context (in Romania, for instance) [94,95].

Data Collection
An online survey was developed to investigate the behaviour of students towards the study of EFL and ESP during the COVID-19 pandemic period. A total of 394 students (out of 565 students who study English) enrolled in various academic programs in the first and second years of study at the University of Agricultural Science and Veterinary Medicine Cluj-Napoca, Romania, filled the questionnaire during the 2020-2021 academic year. Participants were informed about the purpose of the study before filling the questionnaire and agreed to participate under anonymity and protection of GDPR data.
The measures used in the questionnaire were based on the aforementioned Foreign Language Enjoyment (FLE) scale [91]. Six additional items were considered to reflect online teaching under the circumstances of the COVID-19 pandemic, developed based on discussions with university experts in education sciences about teaching methodology, as well as on the exiting literature on the online language teaching systems. The items referring to online teaching were formulated to depict students' opinions towards the online teaching system compared to not only the face-to-face teaching system in general, but also to specific aspects, such as teaching and assessment tools used during class, the easiness of interactions with teachers and peers, active participation during class, and technological challenges to participate in online classes. The measures were assessed on a five-point Likert-type scale which ranged from 1 (totally disagree) to 5 (totally agree).

Statistical Analyses
Following the description of the sample using descriptive statistics, a principal component analysis (PCA) using varimax rotation was undertaken on the items for data reduction [96,97]. Internal consistency was checked using Cronbach's alpha, knowing that a score of more than 0.6 is considered acceptable [96]. The validity of data and the adequacy of the sample size were checked prior to factor analysis with Bartlett's test of sphericity (BTS) (p-value less than 0.05 confirming that the sample is suitable for factor analysis) and respectively with the Kaiser-Meyer-Olkin (KMO) test (a value greater than 0.7 confirming the adequacy of data for factor analysis) [96,98]. Factors with eigenvalues greater than 1 and item loadings greater than 0.4 were retained for each factor grouping [96]. Furthermore, the PCA results were used to identify clusters of students expressing similar behaviours towards the learning process of English during the COVID-19 pandemic. Hierarchical cluster analysis using Ward's method was performed to estimate the number of clusters and to provide initial solutions for K-means clustering [96,99]. This enabled a final clustering solution to be reached. Data normality was checked using the Shapiro-Wilk test. Differences among groups were assessed using the Mann-Whitney U and Chi-square tests (a p-value of less than 0.05 was considered statistically significant). All analyses were performed using SPSS 20.0 software.

Sample Characteristics
For students belonging to the 1st and 2nd years of English as a Foreign Language and English for Specific Purposes, their curriculum comprises one compulsory foreign language in the first and second academic years with other study programmes covering a foreign language as a compulsory subject in the first year, but an elective one in the second. A proportion of approximately 75% of students chose English as their optional/elective subject to the detriment of the French and German languages as their other potential options. English is also the main foreign language studied in Romanian schools, with the first foreign language being taught from the preparatory grade (6 years of age) and the second being taught in the fifth grade (11 years of age) in state-funded education. Students' backgrounds and levels of English are heterogeneous within their respective groups, and their English-proficiency levels were self-perceived, according to the European Common Framework for Languages (A1 to C2), and are detailed in Table 1 along with other socio-demographic characteristics. Intermediate (B1) and Beginner (A1) students predominate with an almost equal share (98 and 94 students, respectively), followed closely by Intermediate + (B2) and then Pre-intermediate (A2) students. Advanced (C1) and Advanced + (C2) are the least numerous categories, with 44 and 3 students, respectively. Female students are preponderant (62.9%). The majority of students are up to 21 years old (92.9%). Regarding the place of residency, a slightly larger percentage of students reside in urban areas (57.9%).

Factors That Determine Students' Engagement While Learning EFL and ESP in the Online System
To address the first research question on the factors that determine students' engagement while learning EFL and ESP in online systems, a principal component analysis was carried out. The initial set of items was reduced until internal consistency was confirmed (Cronbach's alpha = 0.714). The validity of data and the adequacy of the sample size were also confirmed based on the BTS measure (χ 2 = 53107.695, df = 153, p < 0.001) and on the KMO measure (0.856). The principal component analysis with varimax rotation reveals four factors with eigenvalues greater than 1 and factor loadings greater than 0.4, explaining in total 64.31% of the variance ( Table 2). The Cronbach's alpha coefficient is above the recommended thresholds (ranged from 0.79 to 0.89), showing a good internal reliability in the case of each of the four factors. The factors were named "Anxiety", "Online system", "Involvement and enjoyment", and "Class environment and dynamics".
The first retained factor, "Anxiety", explains 27.5% of the total variance. The items grouped refer to emotions, such as panic, embarrassment, nervousness, anxiety, and underestimation, which students may experience during the learning process. The highest scores were found for the items related to level of nervousness while speaking (3.60 ± 1.203) and to the underestimation of one's own achievements compared with those of their classmates (3.48 ± 1.302), both with a tendency from neutral to agree. Panic was experienced when students were suddenly asked to intervene in class discussions (3.16 ± 1.333), and fear to volunteer to answer during the class (2.90 ± 1.307) registered neutral scores. The feeling of anxiety tends to be neutral (2.79 ± 1.339).
The second retained factor, "Online system", explains 17.9% of the total variance. This component gathers all items related to online teaching, thus making it possible to compare students' attitudes related to the process of online teaching versus classroom teaching based on the mean scores (Table 2). Students are neutral about the learning tools used online compared with the tools used in a classroom (3.22 ± 1.081) and about the assessment tools (3.15 ± 1.108) and interactions with teachers in the online classes (3.14 ± 1.138), with an overall neutral score when comparing the two learning systems (2.96 ± 1.139). The highest score was obtained for the item related to the technological challenges students experience in the online system, with some considering it to be an impediment and others not (3.49 ± 1.096), followed by easiness to participate and stay focused in online classes (3.41 ± 0.972) with a score ranging from neutral to agree. The "Involvement and enjoyment" factor explains 11.9% of the total variance and refers to items that describe how students see themselves and feel about participating in English classes. Students are content with how English classes are delivered (4.39 ± 0.741) and find the classes interesting (4.11 ± 0.908). A slightly agree score is found regarding whether they feel proud of their accomplishments (3.65 ± 0.969), but on the other hand, some recognize that their involvement is lower (3.39 ± 0.983).
The last factor, "Class environment and dynamics" explains 6.9% of the total variance and shows how students perceive English classes. The score on the item related to the perception on the atmosphere is high (4.43 ± 0.742). Slightly lower scores were found in the evaluation of colleagues (4.29 ± 0.759) and in the fun experienced during English classes expressed by laughing (3.84 ± 0.870).

Typology of Students
The second research question refers to identifying the typology of students based on the factors that influence their engagement while learning in the online system. In the initial step, the four factors obtained from PCA were used as a clustering basis in hierarchical cluster analysis using Ward's method to estimate the optimal number of clusters. In the second step, the K-means clustering method was used to obtain the final clusters. The twostep cluster analysis produced two clusters of students with similar behaviours towards the learning process of English as a foreign language during the COVID-19 pandemic. All four factors contribute to defining the two clusters (p < 0.001, Table 3). The mean scores of factors are different between the clusters (p < 0.001, Table 4).  Cluster 1 (moderately engaged students) includes 177 students (44.92% of the entire sample). Overall, students from this group gave more neutral answers. Although the class environment and dynamics are important to them, as they show appreciation towards colleagues and their overall environment, they express slight anxiety and nervousness during class when asked to actively participate. They also consider that online classes are delivered more or less the same as the face-to-face classes. Moreover, their involvement in class activities is lower than that of students belonging to Cluster 2, showing an overall passive behaviour. They enjoy class participation, but not as much as their colleagues from Cluster 2.
Cluster 2 (highly engaged students) includes 217 students (55.08% of the entire sample). These students gave more positive answers. Their active behaviour is the result of the high level of involvement and interest in participating in classroom activities. They also highly enjoy the classroom environment and consider their colleagues to be nice. Students enjoy actively participating in class and feel more confident while speaking during class, but they perceive a slight difference between online and face-to-face classes. Their level of anxiety and nervousness is much lower than those of their colleagues' belonging to Cluster 1.
When analysing socio-demographic characteristics, no significant differences were found between the two clusters (p > 0.05, Table 5). Thus, students' behaviours cannot be explained by gender, age, place of residency, year of study, nor by their level of knowledge. No differences were found between the two clusters in terms of the methods of learning EFL and ESP that are preferred (p > 0.05), which may reside in the fact that both clusters reveal moderate to high engagement in classroom activities and due to the plethora of preferred (inter) active and collaborative methods of learning ( Figure 1). As such, games and questionnaires/quizzes are at the top of their preferences, followed by conversations and video recordings. For students of ESP, it is surprising that written specialized texts fall at the bottom of the list. However, this can be explained by the fact that reading and writing tasks have been predominantly transferred to autonomous individual tasks on the Google Classroom platform, rather than used as in-class activities. These are followed by grammar exercises, which are rarely considered as dynamic and interactive as the previous methods selected, and they are most often approached in a contextualized manner in the ESP language classroom.
14, x FOR PEER REVIEW 11 of 20

Discussion
Incessant exploration for optimal ways to teach students (in the higher education context) is paramount, especially within today's world marked by the COVID-19 pandemic [14]. Regardless of whether the switch is temporary or not, or partial or not, university management, teachers, and students struggle to adapt to sudden changes, both aiming at a successful learning outcome. Even prior to the COVID-19 pandemic, literature studies have focused on the comparative efficiency of online learning and face-to-face learning in achieving proficiency. These efficient learning outcomes have been classified by Robinson and Hullinger [100] as student outcomes, test scores, and grades; student attitudes about learning; and overall student satisfaction with online learning. Most studies have resumed, though predominantly assessing test scores and grades, on the findings that the two systems of learning are not significantly different [101,102], or even that online learning leads to more performance and effectiveness in language learning [103,104]. Complementarily, this empirical analysis was designed to understand students' attitudes and behaviours in the COVID-19 pandemic higher education context, which are reflected in their level of engagement (with a predominant focus on emotional and social dimensions of engagement) while learning, alongside similar attempts in the language learning literature [8,12,76].
The first research question of the study seeks to identify the factors that may influence student emotional and social engagement while learning English in the online system during the COVID-19 pandemic. Anxiety, the online system, involvement, enjoyment, class environment, and class dynamics were found as determining factors in the learning process. Previous research has also found anxiety, particularly Foreign Language Classroom Anxiety and Foreign Language Anxiety, as one of the major factors or barriers in the learning process of foreign languages [35,54,105], alongside the lack of personal contact and communication as well as unfavourable study environments at home [106] during the pandemic. The anxiety levels in the research group differs among students and clusters and was predominantly centred around speaking anxiety and panic and around lack of confidence and initiative in providing solicited or voluntary answers. This difference may be due to personality traits (extroverted students may be more likely to engage, whereas introverted ones might not), as the higher that the levels of introversion and anx-

Discussion
Incessant exploration for optimal ways to teach students (in the higher education context) is paramount, especially within today's world marked by the COVID-19 pandemic [14]. Regardless of whether the switch is temporary or not, or partial or not, university management, teachers, and students struggle to adapt to sudden changes, both aiming at a successful learning outcome. Even prior to the COVID-19 pandemic, literature studies have focused on the comparative efficiency of online learning and face-to-face learning in achieving proficiency. These efficient learning outcomes have been classified by Robinson and Hullinger [100] as student outcomes, test scores, and grades; student attitudes about learning; and overall student satisfaction with online learning. Most studies have resumed, though predominantly assessing test scores and grades, on the findings that the two systems of learning are not significantly different [101,102], or even that online learning leads to more performance and effectiveness in language learning [103,104]. Complementarily, this empirical analysis was designed to understand students' attitudes and behaviours in the COVID-19 pandemic higher education context, which are reflected in their level of engagement (with a predominant focus on emotional and social dimensions of engagement) while learning, alongside similar attempts in the language learning literature [8,12,76].
The first research question of the study seeks to identify the factors that may influence student emotional and social engagement while learning English in the online system during the COVID-19 pandemic. Anxiety, the online system, involvement, enjoyment, class environment, and class dynamics were found as determining factors in the learning process. Previous research has also found anxiety, particularly Foreign Language Classroom Anxiety and Foreign Language Anxiety, as one of the major factors or barriers in the learning process of foreign languages [35,54,105], alongside the lack of personal contact and communication as well as unfavourable study environments at home [106] during the pandemic. The anxiety levels in the research group differs among students and clusters and was predominantly centred around speaking anxiety and panic and around lack of confidence and initiative in providing solicited or voluntary answers. This difference may be due to personality traits (extroverted students may be more likely to engage, whereas introverted ones might not), as the higher that the levels of introversion and anxiety are in language learning, the lower that involvement in class and consistent oral production are [35], thus impacting learning efficacy. Similarly, English proficiency level, language, and socio-demographic background were paramount in assessing anxiety levels, as well as other socio-demographic variables [43].
Nevertheless, it is usually highly motivated students, in this case pertaining to the highly engaged cluster, who are less anxious and more confident and therefore more engaged in language learning classes, as was confirmed by previous studies [50]. It can be inferred that the more proficient in English a student is, the less anxious the student feels about English language use. In contrast, the more anxious and unsure of their abilities students are, the less they tend to communicate orally, and they are less likely to engage in (inter)active and collaborative activities in class, regardless of the learning system, either online or face-to-face, in order not to embarrass themselves in the classroom setting [107]. Additionally, they may be less likely to become more proficient in their language skills. Ultimately, however, certain studies have shown that there is no consistent link to be found between FLA and academic achievement [108].
Teaching methods play an important role in reducing anxiety levels by helping students feel more comfortable and by increasing their trust in their own accomplishments. This can be efficiently achieved by the use of online platforms, such as Google Meet and Classroom, as was the case in the current study or previous research [109]. Moreover, the use of web-based technologies encourages student autonomy and helps them to better control their anxiety [110].
The transition to the online system plays a critical role in this new context determined by the COVID-19 pandemic, as many students have clearly been affected by the sudden change in terms of engagement, performance, and even in terms of well-being and mental health. Volodymyrivna et al. [106] point out that university students find it difficult to adapt to the online system due unfavourable study environments at home with different distractions, poor self-discipline, and demotivation. In some cases, foreign language higher education is cumbersome to access due to poor internet connection or technical problems [111][112][113][114], and, more predominantly, due to availability and sustainability of internet connection and tools to access the teaching media [115]. A study on higher education EFL learners in Algeria draws attention on the significance of technical problems leading to poor engagement of students and ultimate isolation [116]. It is thus important to mention that, in the present study, place of residency plays an important role in the online system, as 42.1% of students live in the rural area, and some areas encounter internet connection difficulties. At the national level there is a high percentage of households with internet connection (80.2% in 2021), but only 39.2% are situated in rural areas [117]. The quality of internet connection also differs among places of residency. Even though a study conducted by the Statista Research Department [118] regards the attendance of children to online classes in correlation with the quality of internet connection, it is most probable that this reflects the reality for university students as well; 19% from the rural areas and 10% from the urban areas had low connection or none at all. Barrot et al. [119] finds technological literacy and competency the least of concern, which supports the results in the present study. Student participation and engagement in class are not significantly hindered by technological setbacks. The focus falls on the impact of the COVID-19 pandemic on the quality of learning and on students' mental health, as confirmed by Copeland et al. [120], who identifies numerous persistent negative effects on student behavioural and emotional functioning during the pandemic. These findings converge towards the aforementioned considerations for Positive Psychology in language learning, where enjoyment in the foreign language setting, as expressed by students participating in the study, leads to engagement [44,77], performance, and ultimately, well-being [121].
Henceforth, in the case of the studied group, it should be pointed out that the interactive classroom approach stimulates student engagement, with games and quizzes, regardless of the teaching system. Certainly, flexibility and adaptations were necessary for the online system, but it seems that most students have managed to adapt to the sudden changes and features of the online system (such as Zoom, Google Meet/Classroom, and Kahoot!), and they enjoy the methods used. Other studies have reported a nostalgic preference for the printed materials used in class to take notes [122] or chiefly prefer PowerPoint presentations in the range of online resources that exert influence on stressors [95]. Stu-dent preference for collaborative and (inter)active methods shows, consistent with other studies [123], that students can still interact socially in the online system, regardless of distance, which is crucial for classroom and task engagement. It should not be disregarded that, although there is high enjoyment towards English language classes, students' levels of engagement may vary from individual to individual, subject to subject, day to day, and class to class [45].
Surprisingly, as was pointed out, students do not particularly find enjoyment in such teaching materials as ESP-specialized texts, contradicting previous research [90] stating that, if the content is familiar and relevant to the students' contexts, they may be more enjoyable and lead to more consistent engagement. One explanation may reside in the fact that authentic technical and medical texts are terminologically dense and demanding, requiring more language proficiency. It has been revealed that proficient students are more tolerant of ambiguity [51] and less anxious; therefore, they are more likely to engage. Since around 42% of respondents were self-perceived as beginners and pre-intermediates, it is perhaps natural for students to find technical and medical texts cumbersome to approach and to set them aside as the least preferred. Attempts were made to adapt materials onto PowerPoint presentations, games, and quizzes, and to transfer reading tasks to the Google Classroom platform alongside autonomous work mediated by the teachers and their consistent guidelines. Cognitive engagement should be fostered with ESP materials particularly incorporating assignments that require synthesis judgements on course-related materials [100]. This has ultimately been applied on the Google Classroom platform in the form of stand-alone academic summaries or in the form of argumentative, opinion, or descriptive essays on ESP subjects. Student engagement has been elicited by their active participation in the curriculum's design and delivery and by correlating subjects with students' own online PowerPoint presentations on ESP topics.
Another key factor is class environment and dynamics. Overall, students' positive evaluations as well as their enjoyment of class environments and peers may be explained by mixed and adaptive teaching approaches towards engaging students in interactive activities that can give a sense of community, belonging, friendliness, and safety, allowing students to express themselves openly and thereby contributing to the efficacy of the learning process. Highly engaged students scored significantly higher in relation to this factor, environment and dynamics, and class atmosphere and peers, showing that classroom enjoyment [44] and intergroup motivation [31] determine a higher level of engagement in classroom activities. Collaboration, communication, and engagement with peers may provide students with a sense of community and belonging, contributing to their overall openness in expression, learning efficacy and learning persistence, and ultimately lower university dropout rates, as research results have shown that dropout rates are higher in online settings [124,125].
The second research question focuses on identifying the typology of students based on the factors that influence their engagement while learning in the online system. The two clusters that were identified differ by the anxiety students express during class, perceptions of the online system, levels of involvement and enjoyment, as well as class environments and dynamics. Cluster 1-the 'Moderately engaged students'-expressed a rather passive behaviour with higher anxiety and lower involvement in class activities. Their behaviour appears to be the same regardless of the teaching system, either online or face-to-face, as they expressed a neutral answer when comparing the two teaching systems. This finding may be explained by the various emotions that students express. Studies conducted worldwide following the onset of the pandemic and its education model have revealed that social and emotional engagement is poor in the online system and that students feel isolated and emotionally disengaged [116] due to technical problems, isolation, and a lack of teaching and learning strategies. As was proven time and time again by research in the field, emotions such as enjoyment, anxiety, confidence, helplessness, and boredom during class are emotions that have a direct impact on engagement, as well as on learning behaviour and achievements [126]. However, the neutral perception on the online system found by this study alongside the face-to-face system is consistent with the literature's findings, where the online system is perceived as neutral and interest was reported to remain either unaltered [127] or an enjoyable, participatory [128], and positive learning environment, which fosters creativity, innovation, communication, and collaboration [114]. On one hand, learning performance is reported to be great, with language proficiency being acquired successfully in the online system [114]. On the other hand, students in higher education English classes do not feel that they have improved their language skills or progressed in language learning by studying in the online system [122]. Teacher-student dynamics, monitoring students' learning processes, and ultimately a formal assessment can certainly help the teacher to find and/or adapt tools and methods to increase students' interest and determination to become more active during class and to increase their levels of confidence in personal achievements as well as their language proficiency.
The second 'Highly engaged student' cluster was slightly more consistent than the first cluster. The fact that they are more actively involved in the entire learning process is denoted by their higher scores. In terms of the teaching system, they agree to some extent that the face-to-face system is different, even though they appreciate class dynamics and the class environment to a high extent. This is a nod towards other findings that suggest language students, although highly engaged, perceive the online system as not-so-effective and rigorous [10]. It is rather natural in adult learning and in higher education systems that resilience and motivation towards achieving one's academic and professional goals lead to adaptation and engagement in online classes, as well. However, certain studies have revealed that adult learners, possessing the evaluation ability and the previous consistent experience of the face-to-face system, perceive online classes as effective but consider that they cannot replace face-to-face ones [122], or they feel more confident to study languages face-to-face [129]. This can be explained by the methods used in the face-to-face setting versus the online system. For instance, the methods may be the same, but interactive activities such as speaking tasks and educational games that foster engagement and social participation are more accessible and natural face-to-face, as they foster communities of engagement in learning. Additionally, they may seem to yield immediate results, as there is no social distancing, and the learner can receive immediate feedback from the teacher and the peers. This is of the utmost relevance in speaking tasks, where technical difficulties may hinder oral expression and may be a drawback in students' confident oral expression and in improving their speaking skills [129]. Manners to overcome speaking anxiety may include ungraded speaking tests [130]; explaining the importance and relevance of speaking tasks in the professional real-life context, especially with ESP; challenging students individually; providing students with reflection or research time to prepare their answers; and considering interesting tasks/topics for speaking that may provide students with motivation to engage regardless of their level of English, especially in such situations as the present one, where groups are heterogenous in terms of language proficiency and where the more proficient learners are the ones to naturally engage in the online system, possibly overshadowing their less-proficient peers. Ultimately, as certain studies have shown regardless of the affective dimensions involved, both explicit correction and metalinguistic feedback [86] are to be desired by students from the teachers towards language improvement and stimulating engagement.
Certain pedagogical implications may derive from the overview on student-perceived enjoyment and on connected emotional and social engagement in the online English language classroom environment. This may yield results in selecting appropriate (inter)active methods to elicit FLE and classroom engagement towards language proficiency, as student attitudes towards the language classroom and methods are of the essence. The general nature of this study on student attitudes and preferences has tackled predominantly emotional and social engagement and thus can be a modest steppingstone for teachers to assess more specific behavioural and cognitive task engagement. As such, (inter)active speaking tasks, for instance, in the form of modelling job interviews, followed by explicit corrective feedback provided to students on their performances may efficiently stimulate engagement.
Their pedagogical potential may be augmented by a survey to measure task engagement and give rise to further language teaching initiatives.

Conclusions
Undoubtedly, this study implies that a teacher's presence is key for support and involvement in fostering interactive engagement communities and therefore for developing students' language skills. If the teacher's role is played, there are consistent teacher-student dynamics, and if there is familiarity with peers, anxiety is limited [61]. It is not only one's emotional and personality traits that come into play, but certainly also the social relationships and climate in the classroom that yield confidence and engagement [62]. Common experiences, interpersonal and intergroup motivation, shared academic and professional goals, and even a community of shared "pandemic" anxieties and a joint wish to overcome adversities and prevail may lead to additional efforts being made towards seeing the learning process through. This study on students' behaviours, attitudes, and preferences is a consideration towards students as central partners in the education process, as their enjoyment and engagement are intertwined and lead to a less stressful environment, especially during challenging times and ultimately leading to learning proficiency. The identification of the typology of students may help teachers to understand the factors that determine their behaviour during the learning process and to pay special attention to the elements that shape this behaviour.
This study also emphasizes the need for educators and policy makers to be encouraged to focus more on what the student prefers and enjoys, as well as on their professed expectations, as academic performance is highly dependent on enjoyment. Learning methods should predominantly be evaluated in relation to their success in stimulating student involvement and engagement, while eliminating possible negative feelings and stressors that can affect their well-being, especially during challenging times. Comprehensiveness, multidisciplinarity, and integrativeness of English language knowledge and skills in the overall academic, professional, and personal development of future engineers and veterinarians are fundamental for the sustainability of a university's academic offer, as well as for a country's valued workforce and economy.
This study is not without limitations. No formal assessment was conducted on students to be correlated; therefore, the real learning outcome and efficiency of the online system were not considered in the overall analysis. This is certainly a subject to be approached in the future. However, it should be performed cautiously to avoid biased answers, as grades are of the utmost importance for students to access scholarships in the Romanian system of higher education. Moreover, there should be careful extrapolation of findings to the higher education context, as university study programmes (in sciences and arts) require not only common skills but also the development of specific skills (technical, practical abilities) that may ultimately be relevant for student foreign language classroom engagement. The teacher's role in classroom dynamics can also be further explored, as the literature converges significantly towards the teacher [46] as the aggregate of classroom dynamics. An added value to this research may be the inclusion of specialists in an interdisciplinary team that can provide their expertise in investigating students' psychological profiles and their implications for classroom engagement, as well as the impact of the online learning system on students' well-being. Additionally, engagement may be considered difficult to assess in an overall comprehensive manner, as it may vary on an individual basis, as well as on a day to day, class to class, and subject to subject basis. Foreign language learning experience [45] is relevant in assessing foreign language enjoyment and anxiety; therefore, more investigation of a qualitative nature into prior language experience can also be important.
Further research implications may transcend the assessment of emotional and social engagement as it transpires from the current study and focuses on connecting these types of engagements to particular task engagement and to the investigation of the cognitive and behavioural engagement therein. This direction may complement the present research, as these last two types of engagement are reliable predictors for language learning efficacy and may ultimately be correlated with students' evaluation results, as well as with their self-perceived progress and ultimately their academic achievement. Institutional Review Board Statement: Ethical review and approval were waived for this study, due to the fact that participation was voluntary and that all data were anonymous.
Informed Consent Statement: Informed consent was obtained from all participants in the study.

Data Availability Statement:
The data presented in this study are available upon request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.