Abstract
Aims and Objectives/Purpose/Research Questions:
Research on language learning motivation has mainly focused on the L2 classroom. However, to understand migrant language learners more fully, it is also important to consider how acculturation shapes their learning experiences. Acculturative stress may influence language anxiety (the fear of making mistakes), while learning motivation can play a key role in mediating this relationship. The present study investigates language learning motivation and acculturative stress among German learning migrants in Dresden, Germany. Research questions address whether acculturative stress and motivation can predict language anxiety, whether the perceived expectation to integrate predicts language anxiety, and whether German proficiency can predict language anxiety. It was of particular interest to uncover differential anxiety outcomes across public and classroom settings.
Design/Methodology/Approach:
To address these questions, the Assessment of Language Motivation and Acculturative Stress (ALMAS) was developed through exploratory and confirmatory factor analyses.
Data and Analysis:
German learning migrants (n = 148) ranging from A1 (beginner) to C2 (fluent) were recruited for optional and anonymous online participation. Data was collected from complete questionnaires and analyzed through ordinary least squares regression (OLS) and structural equation modeling (SEM).
Findings:
Analyses confirm that some acculturative stress dimensions (social isolation) predict language anxiety across contexts. Integration expectations emerged as a predictor of classroom language anxiety. German language proficiency offsets public language anxiety, but not classroom anxiety.
Originality:
These findings highlight the importance of addressing sociocultural themes in language education, particularly among migrants.
Implications:
Language proficiency is often discussed as a general remedy for anxiety and acculturative stress; however, these results demonstrate a more context-dependent connection. Moreover, these findings attest to the limitations of motivation (as extrinsic motivators such as societal pressure have adverse effects), and factors such as integrativeness and instrumentality had no significant offsetting effect on reducing anxiety.
Introduction
Language learning for sojourners and migrants is influenced by social input from the target linguistic community. The field of language learning was once studied from individual-cognitive perspectives, which confined research to the L2 (foreign language) classroom (Ushioda, 2012). Branching out, Gardner and Lambert (1959, 1972) speculated that L2 learning had important social and psychological dimensions that distinguished the motivation to learn a language from other kinds of learning motivation. In their view, L2 learners do more than simply acquire linguistic information; L2 learners try to identify with a target language community and adopt their distinctive linguistic styles (Gardner, 1979, p. 193). In this context, migrant language learners can be seen as making a deliberate effort to connect to (or identify with) a particular social group, thus building a bridge from the classroom to the L2 world.
Within the realm of second language acquisition (SLA), foreign language anxiety (FLA) emerged as an important construct of “self-perceptions, beliefs, feelings, and behaviors related to classroom language learning” (Horwitz et al., 1986; Wesely, 2012). The classroom context is undoubtedly relevant to exposing underlying learner processes, but it is nevertheless limiting and needs to be expanded beyond these confines (Sevinç & Dewaele, 2018). Both FLA and language learning motivation have their origins in the L2 classroom; however, as the world has globalized, it behooves research to further explore the acculturative effects on language learning.
Acculturative stress (AS) is defined as the reaction one may have while adapting to a new host culture (Berry, 1980). One of the most notable facets of AS is perceived discrimination from the host society (Sandhu & Asrabadi, 1994). As such, there is particular emphasis on how the host society allegedly perceives the newcomer. Language – being a fundamental agent of communication between parties – plays a role in either exacerbating or offsetting AS; it is established that lower language proficiency can predict higher stress (Lashari et al., 2023). Despite the connection between proficiency and AS, language motivation and language anxiety could contribute further insight into how acculturation impacts language learning.
While AS provides a look into how newcomers are allegedly perceived, language motivation (as outlined by Gardner’s Attitude/Motivation Test Battery, AMTB) emphasizes how the host society is perceived, and if there is an interest or desire for the newcomer to engage with the locals (Gardner & Lambert, 1959; Gardner & Lambert, 1972). Thus, it can provide more insight into the will to learn and communicate than measuring language proficiency alone. Language anxiety has been established as a part of Gardner’s seminal AMTB framework. Within the dynamic of language motivation and AS, there are competing themes that may impact an individual’s language competence: for example, it is less understood how AS impacts language anxiety, and it is less understood how the perceived expectation to integrate impacts language anxiety.
This interplay is particularly relevant in regions that are hesitant to accept (or are generally critical of) migrants, as the social circumstances may not be conducive to successful foreign language learning (FLL) or engaging in communicative practice. The present study examines German language learners in Dresden, Germany. Dresden (the capital city of Saxony) and other swathes of former East Germany have become infamous for anti-immigrant sentiment. Stemming from a long history of isolation post–Second World War, Dresden is fundamentally one of the most “nativist” regions in Germany (Schaub et al., 2021). The resentment toward foreigners has culminated to such an extent that Dresden birthed the “PEGIDA” (Patriotic Europeans Against the Islamization of the West) movement in 2014, which takes a violent stance against immigration (Virchow, 2016). In the last decade alone, the populist party Alternative for Germany (AfD) has made alarming headway through its nativist platform; Dresden and the state of Saxony are among the most supportive of this party (Kaya, 2023). In sociocultural and pedagogical environments, there is a delicate balance between the motivation to learn German (interest in the language and to engage with locals) and AS factors (such as perceived discrimination). Understanding how these motivational and stress-related factors impact language anxiety is essential for explaining learner experiences and outcomes in SLA, a relationship that has been explored in a growing body of empirical research.
Empirical Background
Acculturative Stress and the Acculturative Stress Scale for International Students (ASSIS)
Stress is broadly defined as “individual’s psychological response to a situation in which there is something at stake for the individual and where the situation taxes or exceeds the individual’s capacity or resources” (Lazarus & Folkman, 1984; LePine et al., 2004). Within a cultural context, this “taxation” (whether psychological or material) may manifest through adapting to a host culture (Berry, 2006), as this process takes a toll on mental, emotional, cultural, and physical limits. The acculturation process is commonly defined as “culture change which results from continuous, firsthand contact between two distinct cultural groups” (Redfield et al., 1936) or individuals (Berry et al., 1987). Schmitz (1994) claimed that acculturation is not a passive experience, whereby individuals simply change just by existing in a new environment, but is rather an active process of directly dealing with cultural challenges – this interactive element facilitates a personal (and perhaps stressful) evolution of oneself. Although acculturation does not guarantee itself to be a stressful experience in every instance, there have been notable consequences on mental health (Berry et al., 1987; Hovey, 2000; Oh et al., 2002; Ortega et al., 2000; Thoman & Surís, 2004). Language permeates the acculturation process as it is often linked to the quality of health care access for migrants (Becerra et al., 2015; Deyo et al., 1985) and has been known to impact the educational aspirations of migrant children (Buttiler et al., 2023; Droessler et al., 2021).
Acculturative stress is a measurable construct. The Acculturative Stress Scale for International Students (ASSIS) was developed by Sandhu and Asrabadi (1994) as a tool for measuring the stress among international students in the United States but has long since been applied worldwide in a multitude of contexts; it is considered one of the most utilized measurement tools for AS (Çimşir & Kaynakçı, 2024). The original scale contains 36 items with seven factors. These are: perceived discrimination (8 items), homesickness (4 items), perceived hate/rejection (5 items), fear (4 items), stress due to change/culture shock (3 items), guilt (2 items), and “nonspecific” items (10 items). The responses are on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The total score of the scale varies from 36 to 180, with a higher score indicating a higher level of AS. The authors set 109, two standard deviations above the mean score (66.32), as an alert line for counseling and psychological intervention. Perceived discrimination has been proven to be one of the more impactful factors of AS (Ali et al., 2024; Duru & Poyrazli, 2011; Jung & Lee, 2015; Karuppan & Barari, 2010; Poyrazli & Lopez, 2007; Wei et al., 2008). Language proficiency has also been established as a predictor of AS (Lashari et al, 2023; Dwan O’Reilly, 2018).
Despite the role that language plays in the acculturation process, it is not specifically a factor in ASSIS. Rather, it is addressed in the seventh factor (nonspecific), in which there is only one item: “I feel nervous to communicate in English.” Researchers who aim to dissect the role of language within acculturation often consider proficiency; they adapt the ASSIS to account for language skills, such as Bai (2016), who created the ASSIS for Chinese Students (ASSCS), which includes proficiency as its own factor. Alternatively, researchers have implemented separate proficiency measurement tools in conjunction with ASSIS. Some of the most popular include: the standardized Test of English as a Foreign Language (TOEFL) (i.e., Pandya, 2021; Schmidt, 2017) and/or the Self-Reported Fluency English Survey (SRFES) (i.e., Jean-Paul, 2015; Tamang, 2019; Yu et al., 2014). The SRFES is a three-question survey developed by Yeh and Inose (2002). The SRFES asks:
How well do you speak [the language]?
How well do you understand [the language] when spoken to you?
How well do you read [the language]?
Similar tools include a five-question survey by Akhtar and Kröner-Herwig (2015), who investigated the relationship between German proficiency and acculturative stress. Participants were required to rate each item on a 5-point scale (1 = very uncomfortable/incompetent and 5 = very comfortable/competent). The raw score ranged from 5 to 25 (high scores indicate higher proficiency) (Akhtar & Kröner-Herwig, 2015). Some items included:
How comfortable/competent are you in speaking the German language with Germans?
How comfortable/competent are you in writing a paper in German?
As a result, much of what is known about language and acculturative stress comes from either standardized test scores, brief self-assessments, or a mix of both. Although language proficiency has been used to predict AS, there is little research acknowledging the role of motivation (Haviv Zehner, 2026) and little acknowledging the context-specific effects language proficiency and language motivation may have.
Motivation
Motivation has been deemed an essential factor in the sustainable success of FLL (Dörnyei, 1998; Lăpădat & Lăpădat, 2023). The motivation to learn has been defined as the “direction, intensity, and persistence of learning-directed behavior” (Colquitt et al., 2000). Although ASSIS literature occasionally mentions language learning motivation, literature of the past decade has seldom measured language motivation in relation to AS (Haviv Zehner, 2026). Other aspects of motivation – including the motivation to study abroad (e.g., Mucsi, 2021) or academic motivation (e.g., Wypych, 2020) are more fully accounted for. Despite this, language learning remains a relevant theme, both independently and in relation to AS. Motivational aspects are central to understanding stress, as motivation often occurs with the stressor (Lazarus & Folkman, 1985); although stress is not always considered as an antecedent of motivation to learn, the transactional theory of stress provides support for this relationship (Lazarus & Folkman, 1985). Stress and motivation are not one construct, but they can be seen as two distinct and separate constructs that influence each other to varying degrees. To this extent, one should not conflate them but rather acknowledge the potential of one predicting the other.
Early and seminal studies on language motivation focused on Canadians learning French (Gardner & Smythe, 1975), but these studies failed to acknowledge the role of discrimination or prejudice, or how these negative social perceptions stifle the willingness to adopt the language. More recent literature focuses on motivation to learn English as a foreign language (EFL) (Shaaban & Ghaith, 2000), because English is now the main language of instruction for international students around the world (He & Chiang, 2016; Le Ha, 2009). FLL involves “more than simply learning skills, or grammar rules, it involves an alteration in self-image, the adoption of new social and cultural behaviors [. . .]” (Gardner et al., 1985). To this end, focus should shift away from skill and more toward interest in the language, as well as interest in the linguistic community itself (Gardner et al., 1985). Measurement tools have been based on this theory. The most widely referred to is the AMTB (Dörnyei, 1998; Gardner & Lambert, 1972; Gardner & Wagner, 2005); it is the cornerstone of Gardner’s socioeducational model. The primary justification for focusing on the socioeducational model is that ithas been established in decades’ worth of literature and remains frequently used in the current state of research, often confirming its continued utility (i.e., Rock & Danaee, 2025). When compared to other dominating theories in language learning, such as Ryan and Deci’s (2024) self-determination theory or expectancy/value theory (Atkinson, 2021), the socioeducational model remains distinct in its emphasis on the sociocultural component (Izhar et al., 2025) and is for the purposes of the present study the most relevant, as this study is concerned exclusively with how the acculturative experience interacts with motivators and may induce language anxiety.
The AMTB (Table 1) focuses on five underlying constructs: integrativeness (the general attitude toward the foreign language community), attitudes toward the learning situation (implying that the learning experience influences willingness to learn to the language), language anxiety (fear of using the language, akin to fear of making mistakes), instrumentality (whether the language is actually useful for practical reasons), and motivation (desires or reasons for learning and attitude toward the language itself). According to MacIntyre et al (2009), the integrative motivation from Gardner et al.’s (1985) socioeducational model is central to the study of motivation. Although the model has also been criticized for lacking development and remaining rather stagnant over time (Taie & Afshari, 2015), recent literature continues to use this model (i.e., Harrison & Rodriguez, 2023) to support the ongoing connection between motivation and proficiency.
Constructs and Scales From the AMTB, After Gardner (2009).
Language Anxiety
While all factors play a role in the language motivation framework, language anxiety is of particular interest for its potentially adverse effect on motivation and its potential to exacerbate AS. Language anxiety is a negative emotional response associated with using, learning, or understanding language (Argaman & Abu-Rabia, 2002; Daly & Stafford, 1984; Horwitz, 2001; Pizzie & Kraemer, 2019). Horwitz et al. (1986) were the first to conceptualize language anxiety as its own distinct phenomenon (Agbalizu, 2006). Ayers and Gray (2006) attribute the fear of making mistakes to low self-confidence and motivation; Babu (2010) also draws a connection between this fear and the demotivation to communicate, believing that “inadequate assistance” from educators has resulted in a failure to boost motivation among language learners. This echoes the early findings of Horwitz et al. (1986, 2001), who had demonstrated that the fear of making mistakes can have a considerably negative impact on the efficiency of language learning and ultimately affects learners’ willingness to engage in communicative practice (Arnold, 2000). Thus, there is an inherent connection between language anxiety, motivation, and the learner experience.
The fears associated with language anxiety include apprehension about using the language and fear of making mistakes (Goh & Burns, 2012). These are regarded as strong barriers that squander the linguistic performance of language learners; fear of making mistakes is akin to fear of failure, evaluation, and embarrassment (Salihoglu, 2024). Horwitz (2010) notes the potential of group membership to provoke anxiety: when speaking in a new language, some individuals – especially those with a good target language accent – may perceive themselves and be perceived by others as being “too foreign” for their native group. MacIntyre and Gardner (1991) proposed a different explanation for the development of language anxiety. They saw language anxiety as a “learned emotional response” resulting from repeated negative experiences associated with the foreign language.
There have been several (often conflicting) findings related to language anxiety. Most studies examining language skill and anxiety report a negative relationship (Montemitro et al., 2021), including a decrease in one’s self-perceived proficiency (MacIntyre et al., 1997). A consistent finding is that anxious learners do not communicate as often as more relaxed learners (Anggraini, 2022; Kleinmann, 1977; MacIntyre & Gardner, 1991; MacIntyre & Charos, 1996). In regard to the L2 classroom, earlier findings concluded that anxiety can have a negative impact on students’ attitudes and motivation toward language study (Phillips, 1991) and can negatively impact final grades (Jee, 2012), and that this tendency holds regardless of what the foreign language is (Saito et al., 1999). Significant negative correlations have been found between language classroom anxiety and general language-use anxiety (Perales & Cenoz, 2002). More recent studies continue to find an inverse relationship between motivation and language anxiety across various languages, including among English learners (i.e., Liu & Huang, 2011; Pei, 2023), and an inverse relationship between language anxiety and German proficiency among German learning Jordanians (Rabadi & Rabadi, 2022).
However, as both language anxiety and language learning motivation have their origins in the L2 classroom, literature representing a concrete L2 society and its effect on anxiety is more challenging. For example, Spitalli (2000) found a negative relationship between anxiety and high school students’ attitudes toward other cultures. This suggests that integrative motivation potentially alleviates anxiety, but (considering the study was done in the classroom), it lacked an objective target L2 community for the learners to aspire toward. The concept of the “L2 Self” (as famously proposed by Dörnyei, 2009) suggests that a target L2 community need not be present for the learner to develop integrative motivations, as learners can still develop an alternative language identity and imagine their future selves within a cohort of L2 speakers. The effects of the imagined L2 self on language anxiety are nonunanimous: for Taiwanese English learners, a stronger personal desire to become proficient in English in the future results in higher levels of anxiety (Wang, 2009). This result is at odds with other findings, such as Papi (2010), who found that the L2 self plays a significant role in reducing Iranian learners’ English anxiety.
One proposed explanation for the positive correlation between anxiety and heightened will for proficiency includes Schumann’s (1986) acculturation theory, in which language learners resist the target language culture to preserve their own cultural identity. Other literature pertaining to language anxiety and acculturation includes Liu (2025), who found a positive association between anxiety and social integration among elderly migrants in China who had relocated to a new Chinese region. Rose (2010) had investigated English learners within the L2 classroom, although the learners themselves were Spanish-speaking migrants living in the United States (among the L2 society). The results found no correlation between anxiety and AS, but this may have been because the language anxiety questionnaire was geared toward classroom anxiety (Foreign Language Classroom Anxiety Scale; Horwitz et al., 1986) rather than societal use anxiety.
Essentially, one perspective finds that heightened anxiety results in lower motivation (including reduced integrative interest in the language community and culture), while another perspective finds that a heightened sense of integrative orientation (or sense of one’s own future proficiency) may mediate anxiety. The role of proficiency is also wavering, as an increase in proficiency does not necessarily imply the motivation or willingness to communicate outside the classroom (Albarracin et al., 2019; Strong, 1984). It is essential to extend literature beyond the L2 classroom and into the lived reality of language learners within the L2 society, particularly regarding migrants (Kukulska-Hulme & Lee, 2019; Morrice et al., 2021; Perry & Moses, 2019).
Stress, Motivation, and Anxiety: The Intersection of Integration
Germany is undoubtedly one of the most significant locations for migration (Green, 2013; Martin, 2002). Despite this, it was a famous motto in the late 20th century that Germany was “not a country of migration” (Martin, 2002). Migration and social integration are often themes treated with skepticism rather than enthusiasm (i.e., Deylen et al., 2026; Heckmann, 2016). Such social topics have been the platform of right-wing extremism in Germany (Kononova, 2024) and potentially contribute to an unwelcoming environment, which is unconducive to the sustainable learning of German. Native Germans consider learning German to be an essential expectation for integration and for eventually obtaining citizenship (Baumann et al., 2025). The expectations from others, however, can serve as a potent source of extrinsic motivation (Sugden, 2000; Troyer & Younts, 1997). Nowadays, language learning is more than a matter of will but of obligation.
There are both compulsory and voluntary language training courses for migrants, which are included in several state-funded efforts to support integration (Schroeder, 2024). Germany uses the Common European Reference Framework for Languages (CEFR), which is a Europe-wide framework for language education, running from A1/A2 (beginner) to B1/B2 (intermediate) and then C1/C2, with C2 (native-like speaker) (Council of Europe, 2018; Verhelst et al., 2009). The B2 is commonly deemed the desirable standard for migrant learners, as proof of B2 proficiency is often required for employment, university enrollment, or higher visa status (including citizenship) (Svoboda & Zagar-Sostaric, 2019). The compulsory nature of German language learning is juxtaposed by the assumption that migrants indeed have an interest in learning German (whether that be for instrumental, intrinsic, or integrative purposes).
It has been shown that these extensive measures to facilitate and/or expedite linguistic prowess in migrants have had adverse effects (Schulte, 2024). Moreover, learning German for increased job market access does not necessarily equate to social acceptance in the interpersonal sense, as there have been several examples of minority populations who achieved fluency in the host population but still cannot escape discrimination or a lingering sense of segregation, such as Hispanic migrants in the United States (Bednarz, 2017). It stands to reason that acculturative stressors, language motivation (including the perceptions migrants have toward Germans and interest in the German language), and language anxiety (the fear of making mistakes) play distinct and interwoven roles within the constellation of integration.
Problem Statement and the Present Study
Acculturative stress and language motivation are transactional themes. Both original frameworks acknowledge the role of the host society: the ASSIS focuses on how the host society sees the migrant (i.e., perceived discrimination), whereas the AMTB focuses on how the migrant sees the host society (i.e., integrativeness). However, there are numerous gaps to address:
First, AS literature focuses predominantly on international students at host universities. This echoes into a similar trend within the fields of language motivation and language anxiety, as these constructs have their origin in the L2 classroom. However, the deliberately temporary circumstance for those studying abroad is perhaps one reason why language motivation is seldom measured in tandem with AS and why the motivation to learn a language is foregone for academic motivation. In addition, international students are often taught in English as an international language of instruction, whereas the host language may not be English. For researchers interested in the effects of language beyond campus, this could mean contingently studying the effects of at least two languages (within the classroom and within society). For these reasons, more research on migrant language learners needs to be conducted to understand the acculturative effects on language anxiety and motivation from within the L2 classroom and among the representative L2 community.
Second, there is a lack of literature comparing predictors of anxiety across settings, as most literature tends to focus on either the classroom or the public sphere. In a similar light, proficiency tends to be discussed as a general cure-all for AS and language anxiety, although this relationship could be more nuanced. Considering the theoretical and empirical background surrounding motivation, language anxiety, and AS, the following research questions will be addressed:
H1 supposes that AS will predict language anxiety (all dimensions and in public and classroom contexts). H2 follows that the perceived expectation to integrate will also predict language anxiety (in both classroom and public settings). H3 predicts that proficiency will negatively predict anxiety in both contexts. Finally, H4 posits that motivation will negatively predict language anxiety (providing an offsetting effect).
Methodology
The present study conducts a questionnaire on AS in tandem with a survey on language motivation. However, one cannot simply combine the ASSIS and AMTB – both surveys need to be modified to enhance the respondent experience and tailored to fit the social context. Therefore, this section is ultimately divided into several components: the design process for a questionnaire with stress and motivation components (the Assessment of Language Motivation and Acculturative Stress, ALMAS), exploratory and confirmatory factor analyses, survey administration, and predictive/explanatory analyses (via OLS and SEM) .
ALMAS Questionnaire Design
Adapting ASSIS
As mentioned by Sandhu and Asrabadi (1994), factors Fear and Stress Due to Change yielded low variation compared to other factors in the original ASSIS. Bai (2016) yielded similar results and had omitted them from the results. The original factors of Fear and Stress Due to Change were therefore also omitted from the present study for the following reasons: one, previous literature shows that including them does enhance the survey. Second, a shorter survey would improve participant experience. Additional changes have also been made in this study: many of the original items in Factor 1 (perceived discrimination) are redundant with the items in Factor 3 (perceived hate/rejection); therefore, Factor 3 was omitted for survey brevity. The original “nonspecific” factor was renamed “Social Isolation” to be more descriptive, as it contains the following items:
Ultimately, ALMAS includes four AS factors (perceived discrimination (5 items), homesickness (4 items), guilt (2 items), and social isolation (6 items) – 17 items total. Using a sample of 254 participants, a confirmatory factor analysis (CFA) was conducted to examine the factor structure of the revised questionnaire, comprising four latent factors: social isolation, perceived discrimination, homesickness, and guilt. It should be clear that although ALMAS represents stress and motivation in one survey with two divided sections, stress and motivation are separate constructs and therefore require separate factor analyses. For this reason, 254 participants were sufficient to carry out a factor analysis on the 17 stress-related questions. A widely accepted ratio is 10 cases per indicator variable (Mundfrom et al., 2005; Nunnally & Bernstein, 1967).
All items loaded significantly on their respective factors, with unstandardized estimates ranging from 0.530 to 1.453. Loadings were strong for the “Social Isolation” factor, with Sense of Community showing the highest loading (1.427, SE = 0.252, p < .001) and Cultural Values the lowest (0.530, SE = 0.163, p = .001). The perceived discrimination factor also demonstrated robust loadings, with Biased Against Me showing the highest loading (1.453, SE = 0.186, p < .001). Homesickness items loaded significantly on their factor, with Unfamiliar Surroundings exhibiting the strongest loading (1.331, SE = 0.147, p < .001) and Leaving Relatives the lowest (0.802, SE = 0.114, p < .001). Finally, the guilt factor showed substantial loadings, with Living Different Lifestyle loading at 1.071 (SE = 0.141, p < .001). These results (Appendix, Item A) indicate that the hypothesized four-factor model is acceptable and that each observed variable reliably represents its intended latent construction. As the AS component includes sensitive questions (pertaining to race, discrimination, and potentially private feelings), it was placed after the motivation component to encourage participation and to prevent participant intimidation (Yan, 2021).
Adapting AMTB
Regarding the motivation questionnaire, the five factors from Gardner were adapted as a general framework for the survey: motivation, integrativeness, attitude toward the learning situation, learning anxiety, and instrumentality. Seventeen original motivation items were created for ALMAS using the AMTB framework. Many items were inspired by the Foreign Language Learning Motivation Questionnaire for Filipinos (FLLMQ-F) (Gonzáles, 2010), specifically the desire for professional enhancement, desire to communicate with foreigners, personal interest in [German], and desire to integrate oneself. Considering the context in which ALMAS is being implemented (in Germany, with German language learners), these questions are adapted accordingly to suit the circumstances. The initial result is an 16-item questionnaire utilizing the same Likert-type scale, 1 (strongly disagree) and 5 (strongly agree) (the same scale as in the AS component).
Exploratory Factor Analysis
Given that this motivation questionnaire lacked the established foundation of its acculturative counterpart, an exploratory factor analysis (EFA) was used to investigate the factor structure, using data from 170 participants, which is considered an acceptable amount of data for an 16-question survey (Mundfrom et al., 2005; Nunnally & Bernstein, 1967).
Tests of uniqueness and communality indicated that four motivation variables were too unique to form a cohesive factor and were therefore removed from the factor analysis: Perceived Expectation Integrate, Important to Adapt, Experience Impacts Will, and Accepting Feedback. The test of item uniqueness/communality can be found in the Appendix (Item B). The perceived expectation to integrate is of distinct consideration as it approaches the concept of integrative motivation from extrinsic angles. Whereas integrative motivation probes whether a learner is fundamentally interested in a linguistic community, the perceived societal expectation that a learner ought to be interested may provide additional insight.
The EFA factors accounted for 41% of the total variance (Appendix, Item C). Factor 1 essentially represents Gardner’s “Motivation” factor, which included interest in German and attitude toward learning as subcategories: Will to Learn German, Interest in German, Positive Attitude Learning, Perception German Community, and Positive Learning Experience. Factor 2 comprised instrumentality items German Helps Integrate, Improves Daily Life, and My Goals. Factor 3 captured language anxiety items (Fear Mistakes in Classroom and Fear Mistakes in Public), and Factor 4 reflected Integrativeness (Foreigners are Welcome and City Reputation). Factor loadings were generally strong (λ ⩾ 0.54), with minimal cross-loadings, and the Kaiser–Meyer–Olkin measure indicated sampling adequacy (KMO = 0.665). This structure demonstrated an interpretable pattern consistent with theoretical expectations of motivation subdomains, although the questionnaire itself is relatively brief.
Confirmatory Factor Analysis
A CFA was conducted on a larger sample of 254 participants to test the factor structure of the motivation questionnaire. The CFAs are generally considered more sample-demanding than EFAs, because they involve the estimation and evaluation of a hypothesized measurement model, including latent variable relationships and model fit indices, which are sensitive to sample size (Brown, 2015). The CFA largely confirmed a four-factor structure. Standardized factor loadings were mostly strong (λ = 0.48–0.81) and significant. Factor correlations were moderate, supporting the conceptual distinction between subdomains. Model fit indices indicated acceptable fit: CFI = 0.91, TLI = 0.88, RMSEA = 0.074. Overall, the CFA supported the validity of the four-factor model. The CFA can be found in the Appendix (Item D).
Language and Demographic Questions
German Level
The convenience of implementing ALMAS at language centers is that every student has already been placed by trained educators into their corresponding language levels. This is a more objective approach to classifying aptitude compared to self-assessment. Each student is enrolled in a course ranging from A1 (Beginner) to C2 (Advanced), as based on the CEFR. In ALMAS’s language and demographic components, respondents are asked the following:
Present Language Level (A1 C2 – European Framework Scale)
How long have you been learning German?
How long have you been in Germany?
English Level
General Demographic Questions: Age, Gender Identity, Where Are You From?
The full ALMAS scale is available in the Appendix (Item E).
Recruiting Participants and Administering ALMAS
Two different language schools in Dresden (Kästner Kolleg and the Goethe Institute) agreed to participate. The director of each school was given a link to the online survey, which included a cover page (a description of the survey). The cover page explicitly states that participation is optional and that the survey is anonymous. This information was reiterated to the directors via email. Once the survey had been reviewed by the director(s), the institutes then confirmed their willingness to participate.
Recruiting took place in two phases: in the first phase, the staff (director and teachers) emailed the entire cohort of students from all language levels and invited them to take the survey (providing the same link and same informative cover page). There are advantages to this approach: considering the invitation comes from trusted staff with whom the students are familiar with, the students are likely to feel more comfortable participating. Moreover, it is much easier for the researcher to widely distribute access to the study.
The second phase took place 1–2 weeks following the email, in which the school invited the researcher to visit the classrooms and introduce the study. This served as a post-email reminder. The QR codes were handed out to willing students for each course, and the researcher was dismissed from campus. The analytical sample included only participants who met the inclusion criterion (i.e., German learning migrants in Dresden and fully completed surveys).
Analyses
The analysis proceeded in two steps. First, an ordinary least squares (OLS) regression model examined predictors of language anxiety. Language anxiety was operationalized as a composite factor based on two indicators (fear of making mistakes in the classroom and fear of making mistakes in public), capturing a basic tendency toward anxiety across language-use contexts. Predictors included AS factors (perceived discrimination, social isolation, homesickness, and guilt), motivation factors (intrinsic, instrumental, and integrative), and German language level (GL). Age, sex, region of origin, and single-item Perceived Expectation to Integrate (motivation variable) were included as covariates. The sample size included 148 participating German learning migrants who exclusively live and study German in Dresden. This sample size is sufficient, as there were 12 predictors for this regression, and the general rule of thumb suggests a minimum of 10 participants per predictor (Green, 1991), it also adheres to Tabachnick and Fidell (2025)’s guideline of N ⩾ 50 + 8 m.
In a second step, a structural equation modeling (SEM) approach was applied to disaggregate this general construct into its constituent contexts. In this multivariate SEM, classroom anxiety and public anxiety were modeled simultaneously as separate but correlated outcomes, allowing for the estimation of context-specific effects while accounting for shared variance between the two anxiety variables. This approach enabled a comparison of whether predictors operate uniformly across contexts or exhibit differential effects depending on the situational domain of language anxiety, as these nuances may disappear when treating the variables as a complete factor.
Results
OLS
An OLS regression (Table 2) was conducted to examine predictors of language anxiety (public and classroom fear of mistakes). The overall model was statistically significant, F(12, 148) = 5.62, p < .001, explaining 38.9% of the variance in language anxiety (R² = .389, adjusted R² = .319). Among the predictors, social isolation (B = 0.577, β = 0.548, p < .001), guilt (B = 0.298, β = 0.363, p = .001), GL (B = −0.204, β = −0.243, p = .007), and sex (B = −0.381, β = −0.190, p = .016) were statistically significant predictors of language anxiety. Whereas AS factors (social isolation and guilt) were associated with higher anxiety, higher German proficiency was associated with lower anxiety.
Multiple Regression Predicting Language Anxiety.
In contrast, motivation factors (intrinsic motivation, instrumentality, and integrativeness) did not significantly predict language anxiety (all p > .05). Although integrativeness shows a negative coefficient (B = −0.134) consistent with offsetting potential, this effect is not statistically significant. The only exception is the motivation item Perceived Expectation to Integrate (B = 0.200, β = 0.193, p = .040), which demonstrates a positive (inducing) effect on anxiety.
Considering that the perceived expectation to integrate can predict language anxiety as a factor in the OLS, it is worthwhile investigating this effect across settings (i.e., public and classroom language anxiety). The SEM (Table 3) indicated that Perceived Expectation to Integrate has a strong and statistically significant positive association with classroom language anxiety (B = 0.431, p < .001), whereas its association with public language anxiety (B = 0.153, p = .150) was not significant. Social isolation significantly predicts anxiety in the classroom (B = 0.431, p = .0017) and in public (B = 0.590, p < .001), which is consistent with OLS identifying it as a strong predictor. Although perceived discrimination did not emerge as a significant predictor in the OLS with language anxiety as a comprehensive factor, it did emerge in the SEM as a significant negative predictor of classroom anxiety (B = −0.311, p = .006). Guilt was a significant OLS predictor, and SEM confirms it is a strongly significant positive predictor of classroom anxiety (B = 0.321, p < .001). While GL predicted language anxiety in the aggregate, SEM indicates that it is a significant negative predictor of public anxiety (B = −0.120, p = .047), but not classroom anxiety (B = −0.026, p = .668). These are not contradictions, but rather, the SEM is extending the initial OLS findings; this shows that the predictive relationship with language anxiety is not uniform, but context-dependent.
Structural Equation Model Predicting Classroom and Public Language Anxiety.
Note. Standardized errors (SE) are shown in parentheses.
p < .10. *p < .05. **p < .01. ***p < .001.
These results provide more nuance into the contextualized effects of stress and the inability of motivation factors to predictably offset anxiety in either context (as they failed to predict anxiety both in the OLS and in the SEM). The only exception to this was the motivation variable Perceived Expectation to Integrate, which serves as an example of extrinsic (and adverse) motivation. Moreover, it provides more detailed insight into the nature of proficiency and its relationship to anxiety.
Addressing Research Questions
Discussion
Connection to Prior Literature
Provided that acculturative stressors such as feelings of guilt, social isolation, and perceived discrimination are predictors of classroom language anxiety, this study upholds prior literature that attests to the importance of supportive educators (Babu, 2010; Gray, 2006). As the role of German learning is so central to integration, it is important to consider how the perceived expectation to integrate induces language anxiety (the fear of making mistakes) within the classroom and how this effect seems magnified within the L2 classroom rather than in public. Addressing sociocultural expectations in the classroom can help learners to cope with this perceived pressure.
There is a long-standing debate as to whether integrative motivation can offset language anxiety. Whereas prior findings maintain that anxiety can sink language motivation (Phillips, 1991), this is contingent on what form of motivation, as this study found a positive (inducing) relationship between integration expectations and classroom fear of making mistakes. Aside from this source of extrinsic motivation, the other motivators had no significant effect. Prior findings suggest an inverse relationship between language anxiety and German proficiency (i.e., Rabadi & Rabadi, 2022). However, the present findings suggest that the connection between proficiency and anxiety is more nuanced; greater proficiency appears to have an offsetting effect in public, but not necessarily in the classroom. The qualifying effects of proficiency’s offsetting power are compounded by the stronger effects of AS on language anxiety, which often permeate both contexts.
Other findings have found an inverse relationship between cultural attitudes (or integrative orientation) and anxiety (Spitalli, 2000). In this study, integrative motivation did not emerge as a significant mediator of anxiety in either context. However, it is relevant to note that the effects of social isolation (marked by feelings of intimidation or fear of discrimination by the host society) did have significant effects in both classroom and public settings. Therefore, the results indicate that negative perceptions of the host society influence anxiety.
Limitations
There were considerable limitations to this study. First, language anxiety was represented by a limited number of items (the fear of making mistakes) rather than with a comprehensive scale; the goal was to predict the fear of making mistakes rather than to comprehensively measure various dimensions of the anxiety construct. Nevertheless, this poses a limitation. In addition, it appears that an item about contact with locals would have enhanced the analysis. For example, it would have been beneficial to know how often, to what extent, or in which contexts, migrant German learners interact with locals. Second, there are challenges related to data collection: these language centers, often being private, often have relatively small class sizes; it may take several semesters of high participation to achieve larger data sets and proportional samples from each language level.
The ALMAS has proven itself to be a viable tool with potential for improvement. The interplay between AS and language motivation offers multiple angles from which this dynamic can be analyzed. This study, therefore, not only contributes a new way of applying motivation but also provides insight into how language anxiety impacts German language learners. Sociocultural insight should be considered when studying language learning, as it appears anxiety and integrative expectations are heightened in the classroom, although elements of social isolation potentially create a rift the public sphere. Moreover, as this study explored the effects of motivation and acculturative stress on language anxiety, it behooves researchers to also explore the dynamic of motivation and acculturative stress directly.
Footnotes
Appendix
Model Fit Indices.
| Fit index | Value |
|---|---|
| Chi-square | 113.620 |
| df | 48.000 |
| CFI | 0.910 |
| TLI | 0.876 |
| RMSEA | 0.074 |
Note. All factor loadings are standardized. Loadings ⩾ 0.40 are considered meaningful. Factor correlations are shown above. Model fit indices indicate acceptable fit (CFI ⩾ 0.90, TLI ≈ 0.876, RMSEA ⩽ 0.08). N = 254.
Acknowledgements
Many thanks to Professor Catherine Caldwell-Harris (Boston University) and Professor Susanne Narciss (TU Dresden) for their guidance, insight, and support.
Ethical Considerations
Ethical approval for this study was obtained from the Ethics Commission of TU Dresden and Office of Data Protection at TU Dresden. Official letters of confirmation from both departments are immediately available upon request.
Funding
The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by Ernst Ludwig Ehrlich Studienwerk (ELES): ELES Research Fellow Funding Code: PF731.
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.