Pre-Service Teachers’ Beliefs, Practices, Emerging Ideologies about Multilingualism and Self-Efficacy Relative to Teaching Multilingual Learners

Abstract

Teacher beliefs have a powerful impact on the development of classroom instructional practices. This article reports the results of research that investigated the beliefs, practices, self-efficacy, and emerging ideologies of 268 pre-service teachers (PSTs) who were preparing for primary and secondary school contexts (Grades Kindergarten through 12; K-12), had just taken a course on teacher language awareness (TLA), and were midway through their teacher education program. Three different sources of numerical and non-numerical data were analyzed: (a) open-ended questions, (b) a language identification task, and (c) teacher-generated instructional materials. Four research questions focused on PSTs’ beliefs, perceptions, self-efficacy, and emerging ideologies about the challenges of teaching in a K-12 context in which structured English immersion (SEI) was the dominant model, working with English and multilingual learners (MLs), and developing TLA.

1. Introduction

For decades, issues related to language and cultural diversity in U.S. primary and secondary schools (Grades Kindergarten through 12; K-12) have been in the educational foreground for classroom teachers, administrators, researchers, and teacher educators. There are over 300 languages spoken by children in U.S. public schools and over five million children who speak a language at home that is different from the language of school (National Center for Educational Statistics (NCES) 2021; Riser-Kositsky 2019). These children are often referred to as English learners (ELs), which is the generic term most often used to refer to all learners in the U.S. K-12 context who are learning English as their second language. While EL is a useful term for reporting demographic information, determining the allocation of resources, and making policy decisions, it is less useful as a term for understanding learner differences and the complex process of language learning because it fails to recognize the diverse backgrounds and profiles of learners that could affect academic success in school.

A growing number of children who are referred to as ELs are, in fact, multilingual learners (MLs). In other words, they already speak or use more than one language as a result of their prior experiences in other contexts or the language(s) they use at home and in their communities outside of school (De Groot 2011). To recognize the important role that additional languages play in the acquisition of target languages and to focus attention on additive rather than subtractive views of bilingualism, the term MLs will be used in this paper to include all students who are learning English in school.

The rising number of MLs in U.S. public schools and the preparedness of teachers to educate MLs successfully is of crucial concern for university teacher educators and teacher education programs that provide courses to meet licensure requirements for K-12 teachers and prepare them for careers in teaching. Even though linguistic diversity and multilingualism are on the rise globally (Alisaari et al. 2019; Aronin and Singleton 2012; Conteh and Meier 2014), many K-12 practicing teachers in the U.S. public school context are monolinguals. As a result, they may find it difficult to develop self-efficacy as teachers of MLs and acquire the skills necessary for teaching effectively. Self-efficacy refers to the beliefs that teachers have in their capacity to act in the classroom in ways that promote learning (Bandura 1997; Cevdet 2011; Wyatt 2018).

Recent research has focused on critiquing teacher education, suggesting that models of teacher education offer ineffective and weak interventions and do not provide the deep specializations that teachers need for working in diverse contexts with MLs (see Cakcak 2016 for a critical review of teacher education models). While there may be some truth in what these all-encompassing critiques report, they do not provide empirically based guidance for determining how teacher education can be improved or what PSTs need in order to become effective teachers of MLs. There is evidence that teacher education can make a difference (Darling-Hammond et al. 2002), especially relative to teacher self-efficacy. To understand how to shape teacher education to serve the needs of pre-service teachers (PSTs) of MLs, it is important to understand what K-12 PSTs believe and how their beliefs, self-efficacy, and experiences are shaping them to become practicing teachers in contexts with MLs.

1.1. Teacher Cognition, Teacher Beliefs, and Teacher Identity

Teacher cognition (Borg 2006) focuses on understanding what teachers think and know and on the unobservable dimension of teachers’ mental lives (Freeman 2002). Rather than emphasize teaching behaviors and what teachers do as indicators of effective teaching, the focus of research on teacher cognition has evolved to include an examination of what teachers believe, what they think, what decisions they make, and why (Borg 2006). From this perspective, teaching is viewed as a complex mental activity that is guided by teacher cognition.

It is important to recognize that teaching is not only a cognitive endeavor but also a sociocultural one. As Barkhuizen (2016) states, teachers construct the “socio-cultural worlds in which they live their teacher and learner lives” (p. 3). The socially oriented conception of teaching has led to a growing body of research associated with the development of language teacher identity (Varghese et al. 2005), which can be conceptualized as teachers’ socially constructed understandings of self. It is assumed that teacher identity plays a key role in teachers’ decision-making processes and serves as an analytic lens that can be used to investigate teachers’ practices. The development of teacher identity is shaped by personal and professional histories and by the culture of the context in which teachers work and learn (e.g., the individual schools or districts—the geographical units for the local administration of elementary or secondary schools in the K-12 context). For PSTs in this study, the context in which they were learning included not only the courses in their teacher education program but also their field experiences in K-12 classrooms, which took place in conjunction with their university courses and their semester- or year-long practice teaching experiences at the conclusion of their formal course work.

1.2. Understanding Context

The K-12 context in which the PSTs in this study were preparing to work is culturally and linguistically diverse, with 120 different home languages spoken by children in the state and 80 in the urban district. Federal guidelines require individual states to implement programs to meet the needs of MLs, and local education agencies (e.g., schools) have a legal responsibility to ensure that all students have access to a quality education, including access to programs that are focused on English language development (ELD). In the K-12 context used in this study, ELD is most often designed and delivered within an educational model referred to as structured English immersion (SEI). In SEI, MLs are educated alongside majority-language speakers of English in both primary and secondary schools. There can be as many as 30 different home language backgrounds present in one school, so two-way immersion has not been viewed as a viable educational model for the education of MLs. Two-way immersion is a form of education in which students from two different language groups, each with a different home language, are taught literacy skills and grade level content in two languages, for example Spanish and English in the United States, so that both groups of students serve in the roles of language model and language learner at different times (for more information see Lindholm-Leary 2004; Lindholm-Leary and Hernandez 2011; Tedick and Wesely 2015).

For MLs, SEI has been criticized on the grounds that it leads to subtractive bilingualism. This criticism exists because in SEI the focus is solely on learning English, rather than on multilingualism—learning English in addition to recognizing, maintaining, and developing other languages in learners’ linguistic repertoires (Cummins 2017; García 2009; García and Wei 2013; Jessner 2008) through the use of multilingual pedagogies. Classes in SEI are taught by grade-level teachers in primary schools and subject-matter specialists in secondary schools in content areas such as history, language arts, math, social studies, and life and natural sciences (i.e., biology, chemistry, and physics) and not English language specialists. All teachers, by default, share the responsibility for the education of MLs because all teachers are likely to have MLs in their classes. Teacher education programs must, therefore, be designed to prepare all PSTs to work with MLs in this diverse context, regardless of whether they are mainstream teachers in primary schools, content-area specialists in secondary, or PSTs specializing in ELD. The teacher education program in the present study embedded courses specifically designed to prepare teachers for working with MLs within the requirements for licensure. Pre-service teachers could take additional courses beyond licensure requirements to obtain an English as a Second Language (ESL) Endorsement (i.e., a qualification for teaching MLs that is overseen by a State Board of Education and not the teacher education program).

It is important to point out that, at least in theory, SEI is meant to be fundamentally different from submersion, where MLs are placed in mainstream classes with little or no support services with the idea that they will simply pick up academic English and grade level content. According to the U.S. Department of Education (n.d.), all practicing teachers in SEI contexts are to have specialized training in ELD in order to meet the needs of MLs (Lillie et al. 2012). In practice, however, SEI may often be quite close to submersion relative to the learning experiences of MLs because developing a cadre of practicing mainstream and/or content area teachers with specialized training in ELD and multilingual pedagogies has proven to be a challenging endeavor. In addition, most practicing teachers are first socialized to teaching in their disciplines and develop professional teacher identities as either primary school teachers or disciplinary experts in secondary schools (Schleppegrell and O’Hallaron 2011) and not as expert teachers of MLs. In fact, many K-12 teachers draw the erroneous conclusion that teaching MLs effectively is just a matter of good teaching (Harper and de Jong 2004) even though there is evidence to show that MLs need targeted ELD to make proficiency gains in English (Saunders et al. 2013) and benefit from pedagogical approaches that draw on their diverse backgrounds and knowledge of other languages (Cenoz and Gorter 2021; García et al. 2016; Krultaz et al. 2022)—components of instruction that are often absent in ELD.

Targeting ELD within SEI is a complex process. In the K-12 contexts, SEI was superseded by an instructional model known as the pull-out model. In the pull-out model, MLs were taken out of mainstream classes to receive targeted ELD from English language specialists. According to critics of the pull-out model, this action denied MLs access to grade-level content and opportunities to develop subject matter expertise. While the claim about access seems logical to non-language specialists, such as parents or politicians, the issue of access is much more intricate. From the perspective of language proficiency, it is important to recognize that access to grade-level content is dependent on a learner’s level of language proficiency. If a learner’s level of language proficiency is quite low, it will be difficult for a learner to access content even if the learner is present in the classroom during instruction. In addition, Saunders et al. (2006) showed that MLs in stand-alone and pull-out programs made greater gains than learners in push-in programs (i.e., an ELD specialist enters a mainstream classroom to provide support for MLs). In pull-out programs, ELD specialists focused on language development activities 91% of the time. When used in conjunction with SEI, pull-out and push-in programs provide important benefits for MLs, but these models are not always available in contexts where SEI is the dominant instructional model.

1.3. Multilingualism and Multilingual Learners

In recent years, multilingualism has become more visible in political, social, and educational contexts (Aronin and Singleton 2012; Jessner 2008; May 2014, 2019; Ortega 2014). Contemporary multilingualism is no longer limited by specific geographical boundaries or to certain social echelons, or is it restricted to a particular modality, such as writing. Globalization, the use of digital technologies, and the increase in human migration (Aronin and Singleton 2008; Cenoz 2013; Christison and Murray 2022) have all contributed to its visibility. In addition, there has been an increased focus on multilingualism in applied linguistics as researchers have taken up the study of multilingualism from societal and individual perspectives (Kramsch 2010; Krultaz et al. 2022).

Methodologies that focus solely on the development of English (e.g., Echevarria et al. 2016) tend to discount the value of other languages and fail to recognize the concept of multilingualism at their core while multilingual pedagogies target the development of multiple languages and competences by making use of learners’ whole linguistic repertoires (De Angelis 2011). For example, pedagogical translanguaging (Cenoz and Gorter 2021) may include specific strategies for learning to write, which can transfer across languages (Cenoz and Gorter 2011). This type of pedagogy encourages MLs who have academic backgrounds in languages other than English to bring this knowledge to the task of learning English, thereby making it different from the subtractive model of bilingualism inherent in SEI. To address the needs of MLs most effectively, teachers and schools need to explore pedagogical strategies that consider MLs’ backgrounds and include them in instructional planning and curriculum design.

Multilingual learners in the K-12 context exhibit many diverse profiles in terms of how they have acquired the languages in their linguistic repertoires and how they use them on a daily basis. Some multilinguals may have been exposed to two languages from birth and are learning English as a third language in school. Other MLs may have a primary language they speak at home, know a second language they learned in school as a result of migration, and be learning English in school. Some MLs may speak English at home when the family communicates as a group because the parents use different mother tongues when communicating individually with their children and extended family. The diverse experiences of MLs result in different levels of language proficiency and competences across language skills. To be a ML does not require perfect mastery of all the languages in one’s repertoire.

Views of multilingualism have been changing and evolving from atomistic views (Cenoz 2013) with a focus on understanding “one language only” and “one language at a time” (Wei 2011, p. 374) to holistic views with a focus on understanding how multiple languages are interconnected and used by MLs in “real life communication involving all the languages and multilingual discursive practices” (Cenoz 2013, p. 11). While holistic views of multilingualism are gaining currency in applied linguistics, atomistic views of language are widespread and still comprise the dominant view of language learning in the U.S. K-12 context.

1.4. Teacher Language Awareness

Shulman (1999) maintains that professional teachers need to be able to take thoughtful grounded actions in the classroom and that these actions must be based on a deep knowledge of subject matter. The subject matter for teachers in SEI contexts includes disciplinary content knowledge as well as knowledge of language. While knowledge of language can be construed broadly to include many different foci such as language as it is used in a specific context, the relationship between language and culture, emerging patterns of discourse, language change and variation, and the relationship between language and power (Arndt et al. 2000; Mooney and Evans 2018; Rojo 2016), the research presented in this article takes a narrower focus and concentrates on PSTs’ knowledge and understandings of language systems, particularly knowledge of morphology (i.e., words and vocabulary) and syntax (i.e., grammar). These systems are at the heart of the language acquisition process for MLs; therefore, it stands to reason that if teachers are to help learners achieve both language and content goals, teacher language awareness (TLA) must be included in the core of subject-matter knowledge of K-12 teachers (Andrews 2007).

Teacher language awareness can be defined as a teacher’s ability to use, analyze, and teach language, and it can be conceptualized into three domains (Andrews 2003; Edge 1988; Lindahl 2018; Wright and Bolitho 1993). The first domain is the user domain. This domain focuses on the language teacher as proficient user of language in a variety of contexts. The analyst domain focuses on developing teachers’ understandings of language systems (e.g., the systems of sounds, words, or sentences). It also includes an introduction to multilingualism, second language acquisition, and the potential ways in which languages beyond one’s home language can be learned. The teacher domain is associated with pedagogical content knowledge (PCK) (Park and Oliver 2008; Shulman 1986), a term used in education to describe the interconnected nature of knowledge—content knowledge and the knowledge of pedagogy that are used to teach specific subject matter. For teachers in K-12 contexts who work with MLs, PCK must also include the knowledge of language that enables teachers to think about language as if they were MLs so that they can identify specific features of language that may pose difficulties for MLs, adapt particular content information in their disciplines (e.g., math, language arts, history, social sciences), and convey this information to MLs through pedagogical means.

1.5. Research Questions

Four research questions guided the current study:

• What are PSTs’ beliefs and perceptions about the challenges of teaching in the K-12 context?
• What are PSTs’ beliefs, perceptions, and emerging ideologies about multilingualism, and their self-efficacy for educating MLs in K-12 contexts?
• How do PSTs frame their identities as teachers midway through their teacher education program?
• What level of TLA do PSTs exhibit after the TLA course and midway through their teacher education program?

2. Materials and Methods

2.1. Research Design

The research design that was chosen for this study allowed the voices of PSTs to be center stage and to be consistent with approaches used for practicing teachers—approaches that have focused on trying to understand what teachers need to know and be able to do (Sharkey and Johnson 2003) in a given context. The research adopted a Grounded Theory approach to analyzing data from the open-ended questions (Creswell 2015; Glaser and Strauss 1967; Milliken 2010). This approach to data analysis allowed the researcher to interpret data and generate themes that were grounded in the data (Braun and Clarke 2006; Hsieh and Shannon 2005) and flexible so that imaginative theorizing could take place (Charmaz 2014). This approach to design permitted the researcher to avoid making apriori assumptions and adopt a neutral view of PSTs’ knowledge, beliefs, and skills within the context of their teacher education program (Simmons 2006). The research was designed around the collection of three sources of numerical and non-numerical data—open ended questions, a language identification task to measure TLA, and teacher-generated instructional materials. These multiple data sources interacted with one another and were used to illuminate, confirm, or dispute what was learned from one data source based on the analysis of another one; for example, how the results from the open-ended questions were informed by the other two sources. By drawing on different sources of data (i.e., triangulating data) both patterns and inconsistencies in the data could be identified.

2.2. Context for the Research

The TLA course in which data were collected for this study was taught by the same teacher for each of the six semesters. It was a required course for obtaining K-12 licensure and also for the ESL Endorsement. There are always limits on the number of credit hours that teacher education programs can require of PSTs. As a result, this TLA course was the only course among the required courses that focused on the development of TLA, specifically what knowledge and skills teachers need to have related to language for working effectively with MLs. The course placed multilingualism at its core and concentrated on helping PSTs develop a broad understanding of how multilingual learners use their language systems rather than solely developing expertise in teaching English. The course centered on two of the three domains of TLA (Andrews 2003; Edge 1988; Wright and Bolitho 1993), the analyst domain and the teacher domain.

The course occurred midway through the teacher education program and included a 16-hour field experience in a classroom context with an experienced site teacher who was required to have an ESL Endorsement. The learner population at the sites comprised at least 30% MLs. During the 16-hour field experience, PSTs participated in a variety of instructional activities from tutoring a small group of learners to making presentations to the entire class, as determined by the site teacher.

2.3. Participants

The PSTs who took the TLA course were preparing to teach in the K-12 context. The course was required for licensure and ESL Endorsement. It was taught by the same teacher in each of the six different 16-week semesters. During the six semesters, a total of 268 PSTs participated in the course. The majority of the PSTs (n = 196 or 73%) were seeking licensure for primary school or special education (Grades K-6). Primary school teachers are responsible for teaching math, science, social studies, history, and language arts at grade level. Seventy-three PSTs (27%) were seeking licensure for secondary school contexts in a specific content area, for example English language arts, foreign language teaching, health, history, math, social studies, or the sciences (e.g., biology, physics, and chemistry) and the ESL Endorsement. Seventy-three percent (n = 196) of participants considered themselves to be monolingual speakers of English even though eighty-three percent (n = 222) indicated that they had studied at least one foreign language (i.e., a language other than English) at some point in their educational history. For primary school teachers, 47 (24%) were males and 148 (76%) were females. In the secondary school context, 43 (59%) were male and 30 (41%) were female. All participants were between the ages of 19 and 32, with a mean age of 21.5 years, and were novice teachers meaning that none had obtained licensure, nor had they assumed full responsibility for classroom instruction in the K-12 context, although they had varying degrees of experience as teacher aids and tutors.

2.4. Data Collection

To understand the effects that a course on TLA may have on PSTs’ beliefs, perceptions, self-efficacy, and emerging ideologies about teaching MLs in the K-12 context, their understandings of multilingualism, and their developing identities as teachers, as well as the development of TLA, three types of data were collected at the conclusion of each TLA course: (a) written reflections in response to five open-ended questions, (b) information from a language identification task, and (c) a teacher-generated instructional materials task.

2.4.1. Open-Ended Questions

The PSTs were asked to respond to five open-ended questions in writing. The questions were designed to capture the perspectives of novice teachers:

• What worries you most about teaching in the K-12 context?
• What worries you most about teaching MLs in the K-12 context?
• What do you still want to know about teaching MLs?
• What do you think might be the most important concept you learned in the course? Why?
• In terms of teacher language awareness, how well prepared do you think you are for teaching MLs in the K-12 contexts? Place yourself in one of three categories. Please feel free to explain why you placed yourself in the category you did:
  • Category 1: Well-prepared. I would feel confident in accepting a job next week.
  • Category 2: Somewhat prepared. I would need considerable support if were to accept a job next week.
  • Category 3: Not prepared. I would not be prepared to accept a job next week.

Data from open-ended questions were collected during the last 30 min of the final class each semester. All PSTs were given the questions at the same time. If they finished the task early, they were asked to sit quietly, reflect, read, or work on written assignments so that all PSTs in the course would have an equal opportunity to reflect and write for 30 min without the distractions that occur when individual teachers leave at different times.

2.4.2. Language Identification and Teacher-Generated Materials Tasks

For the language identification and teacher-generated instructional materials tasks PSTs chose either a science or social studies text. The science text was about the fundamental properties of volcanoes and volcanic eruptions, and the social studies text was about how Hurricane Maria in 2017 had affected Puerto Rico. Both example texts were ones that could be included in Middle School (Grades 6–8) curricula (see ReadWorks n.d. for an example of the science text). These texts were chosen because they were texts that could be used in lower secondary or upper primary school grades, so the reading level of the texts would be familiar to both primary and secondary PSTs. They were given access to the texts a week in advance of data collection. Data were collected for both tasks in one two-hour online session during the last week of instruction.

In the language identification task, PSTs were given a list of 20 different features of language (see Appendix A) and asked to search the text they had chosen to find examples for 10 of the 20 features. The language features were chosen because they focused on teachers’ broad understandings of how relationships among different parts of language are connected (e.g., words representing different lexical categories, types of phrases and clauses, or types of sentences). These features of language were explicitly taught in the course modules in class and supplemented with online activities and tasks that PSTs could explore on their own in order to gain the expertise they felt they needed. In the language identification task, it was assumed that teachers would choose the features of language in which they were most confident and produce a result that was typical of their best work. Having the text in advance allowed teachers to become familiar with the text and select features of language that would be the best fit for the text.

In the teacher-generated materials task, teachers were asked to create an instructional activity for teaching both the content and language of the text they had chosen. There were three components to the task: (a) identifying the language and content, (b) describing the specific instructional activity so that it would be clear to MLs, and (c) explaining why the activity would be useful for them. The two TLA tasks were chosen because they would be ones that PSTs would likely encounter during their first year as teachers of MLs.

2.5. Data Analysis

Data for the open-ended questions were analyzed using the constant comparative method (Corbin and Strauss 2008; Kolb 2012). Following protocols for this method, data were analyzed for all three types of data as soon as they were collected so that systematic coding and analyses could be combined in an ongoing process that allowed the researcher to reinforce theory generation through the six cycles (i.e., the six semesters) of theoretical sampling.

All data were anonymized for participants who had given consent. During this process, names were removed from each piece of data and replaced with a number, for example 57, and a letter associated with each of the three types of data. The letter “a” was used for open-ended questions, “b” for language identification, and “c” for instructional materials. Data were identified in a spreadsheet, with the participants’ numbers and the letter associated with the type of data, for example “57b” referred to Participant 57 for the language identification task.

2.5.1. Open-Ended Questions

To answer Research Questions 1, 2, and 3, data from open-ended questions were analyzed using an open-coding process proposed by Russel (2000; see, for example, pp. 443–44). In this process, the researcher read through the data numerous times to identify potential categories or themes. As themes emerged, data were pulled into the categories. Finally, data from the categories were compared and linked so that relations among categories could be used to build theoretical understandings or positions. These initial processes were repeated each time data were collected. Through each iteration of coding, new relationships emerged, and data were repositioned and recategorized. The categories that remained stable throughout the six points of data collection were ultimately identified as themes, and examples of participants’ language from the data were attached to these themes. The themes represented the primary beliefs and emerging ideologies of the participants and offered glimpses into the development of teacher identity.

2.5.2. Language Identification and Instructional Materials Tasks

To answer Research Question 4, each piece of data from the language identification task was evaluated and assigned a numerical number: 0 = incorrect and 1 = correct. For example, if the participant selected “an adjective” from the list of language features and if the sentence the participant selected contained an adjective, a score of “1” would be assigned. A total of 10 points was possible.

The number of times participants selected specific language features was also tracked, and a rank ordering of language features was developed. These rankings were meant to assist the researcher in determining the types of language features in which the PSTs exhibited the most confidence.

Data from the instructional materials task were analyzed according to the rubric in

Table 1. Rubric for evaluating teacher-generated instructional materials.

Indicator	Scoring 0 = Not Present or Inaccurate; 1 = Moderately Effective; 2 = Effective
	0	1	2
Identification of specific language and content
Description of the learning activity
Explanation of how the activity would be useful for MLs
Total score ______

. Each sample was given a score (i.e., 0, 1, or 2) on each of the indicators. A score of 0 meant that the indicator was not present or in some cases that it was inaccurate. A score of 1 meant that the indicator was moderately effective; for example, the learning activity was not described so that it could be clearly understood. A score of 2 meant that the indicator was effective; for example, the PST was able to explain why the learning activity would be useful for MLs. A total score was calculated. The scores for each participant were then entered into a spreadsheet. A total of six points was possible.

3. Results

Results are organized and presented as they relate to each type of data that was collected—open-ended questions, the language-identification task, and the teacher-generated instructional materials tasks. For the open-ended questions, the results are centered on the themes or categories that emerged from the interpretation of the data. Example responses from the open-ended questions are selected to support each of the themes that emerged. For the language identification task, the results focus on TLA and on how well PSTs are able to identify features of language in the texts. Data are also presented in terms of the degree of sophistication represented by the feature, for example an adjective or a complex sentence. For the instructional materials tasks, the results are presented based on the indicators in Table 1. These indicators target PCK and how well teachers are able to identify the language demands for MLs that are inherent in the text and create instructional materials.

3.1. Open-Ended Questions

The themes that emerged from the coding processes for the data for open-ended Questions 1 and 2 are summarized in

Table 2. Themes from open-ended Questions 1 and 2.

Open-Ended Questions	Themes
Question 1: What still worries you about teaching in the K12 context?	Theme 1: Managing and balancing life Theme 2: Finding support Theme 3: SEI context Theme 4: Classroom management
Question 2: What still worries you about teaching MLs in the K-12 context?	Theme 1: MLs’ learning needs Theme 2: Diverse teachers’ roles and responsibilities and emerging teacher identities Theme 3: Instructional preparation and planning

.

Four themes were identified in response to the first open-ended question. These themes represent the challenges that PSTs believed they were going face in working in the K-12 context, in other words, working with MLs and home language speakers of English in the same classroom. Theme 1 focused on general challenges in managing and balancing their lives as teachers. These worries are captured in these example data:

• “I think I will have trouble keeping up with everything—preparation, record keeping, correcting papers, attending trainings, and meeting parents” (33a).
• “So much seems to be expected of public-school teachers. Now that I know more from my field experience, I am worried about whether I can be successful” (16a).
• “I worry about not having enough planning and preparation time” (24a).
• “I worry about being able to keep up and about being able to balance life” (42a).
• “To be able to keep all balls in the air as I juggle teaching, grading, classroom management, etc., etc., etc. What are the secrets of getting it all done while staying cool and collected” (178a)?

Theme 2 can be characterized as challenges that PSTs believed they were going to face as teachers of MLs relative to finding support within the school system:

• “I worry about having a supportive principal. My site teacher said she was lucky having her principal, but not all principals are supportive” (61a).
• “If I cannot reach the MLs in my classes, who can help me” (40a)?
• “I worry about accidentally saying something to ELs or MLs and getting in trouble with parents or ending up in some type of litigation” (221a).
• “I want to be a good teacher, but I am worried about making a mistake with the MLs and losing my job. Who would help me” (51a)?

Theme 3 spotlights the challenges that PSTs thought they were going to face working in SEI classrooms with MLs:

• “With learners from so many different backgrounds and levels of language proficiency in one class, am I going to get my points across” (2a)?
• “I worry about modifying my lessons for MLs and boring the other students” (56a).
• “I don’t want to bore my native English-speaking students by ‘dummy down’ lessons for MLs and ELs” (48a).
• “I won’t be able to accommodate the students who need more time without letting the advanced kids get bored” (38a).

Theme 4 focuses on the challenges of classroom management:

• “I worry about out-of-control students” (17a).
• “I don’t think that I am at all prepared to handle behavioral issues” (147a).
• “I really hope that my dedication as a teacher and my patience and kindness [will] create an environment without these issues.” (92a).

Three themes were identified in response to the second open-ended question. Theme 1 focused on PSTs worries about meeting MLs’ needs:

• “I worry about being able to understand the MLs in my classes, especially when they ask questions. I know from experience that going to a place where no one understands you can be frustrating” (26a).
• “I worry about not being able to meet MLs’ needs because I am not multilingual. Can I do that” (38a)?
• “I am not multilingual, so I am worried about meeting the specific needs of ML students. Seeing things from their view (26a).
• “How do I help them make use of their full linguistic repertoires” (28a)?
• “I find talking to ELs and MLs intimidating” (32a).
• “I cannot understand them and feel embarrassed” (33a).

Theme 2 concentrated on how PSTs’ lack of understanding of their diverse roles and responsibilities affected the development of their identities as teachers of both language and content:

• “Am I a social studies teacher or a teacher of English and MLs” (45a)?
• “My job is teaching history, right” (127a)?
• I am not an English teacher” (142a).
• “I plan to be a chemistry and math teacher in secondary. Next fall, I will begin my practice teaching. In this class, I learned that I would likely have ELs and MLs in my math and chemistry classes. It becomes my responsibility to teach them. But, I am not an English teacher; I am a math and chemistry teacher” (112a).

The third theme targets PSTs’ concerns about instructional planning:

• “I am worried about planning lessons and helping MLs learn new material” (67a).
• “When unexpected things happen in the classroom while teaching, I worry about how to adjust my teaching and getting behind” (26a).
• “I worry if I will know when to recycle materials” (42a).
• “I worry that I am unable to design a course that is very effective” (17a).
• “How do I get ML students to higher reading, writing, speaking, and listening levels? Not sure” (71a).
• “I only know English. How do I talk to MLs” (65a)?

The third open-ended question asked participants what they still wanted to know about teaching and working with MLs. Only 30 mentioned that they wanted to know more about specific concepts or instructional strategies (e.g., pedagogical translanguaging or differentiation). The remaining 238 responses were all of a general nature:

• “creating a curriculum” (10a);
• “how to get MLs excited about learning” (6a);
• “how to know if MLs understand what I say” (13a);
• “working with a language barrier” (19a);
• “finding a way to explain content” (34a).

The fourth open-ended question asked participants to identify a new concept or a skill they had learned in the course that they believed would be helpful in working with MLs. The most frequently mentioned was the concept of pedagogical translanguaging, with 182 responses (68%). The second most frequently mentioned was the stages of second language acquisition, with 51 responses. There were also 154 participants (57%) who identified concepts associated with language and teacher language awareness; for example, more about the sound systems of the languages their students speak or more about how words are formed in other languages. Many of the concepts listed were general ones about teaching and not specific to the TLA course.

The final open-ended question concentrated on PSTs’ perceptions of their TLA preparedness. One hundred forty-two participants (53%) placed themselves in Category 1, believing they were well prepared for teaching in terms of TLA while fifty participants (19%) placed themselves in Category 2 and stated that they were not ready for teaching. The remaining 76 PSTs (28%) in Category 3 believed that they were not prepared to teach MLs and would need considerable support. Two-hundred thirty-nine (89%) of the PSTs noted that the TLA course was the first course they had taken that focused on language.

3.2. Language Identification and Instructional Materials

For the language identification task, the highest score was 90% with 9 out of 10 language features identified correctly; the lowest was 50% with 5 out of 10 correctly identified. The mean score was 6.2/10 (62%). The most frequently selected language features were adjectives, verbs, and prepositional phrases. For the participants who chose these language features, 100% identified them accurately in the text. The next most frequently chosen language features were simple sentences, coordinating conjunctions, and verb forms used as nouns (e.g., a gerund). Eighty-four percent of participants were able to identify a simple sentence correctly. Ninety-six percent identified a coordinating conjunction correctly, yet only fifty-two percent could identify a compound sentence. Forty-two percent identified a verb form used as a noun correctly, in this case a gerund. Words that were incorrectly identified as gerunds were -ing words that were used as verbs in the sentence and not nouns. Only 40% of the participants who chose a restricted relative clause were able to identify one accurately in the text. Seventy-seven percent of the participants who chose a passive sentence as the language feature identified it accurately. For those participants who chose a subordinating conjunction, 66% were able to identify one accurately in the text, but only 38% could identify a complex sentence. All language features were chosen at least once.

Based on the rubric, there were six points possible for the teacher-generated materials. Twenty-seven percent of the participants scored 5 or 6. Forty-six percent received scores of 3 or 4. In addition, 27% received scores of 1 or 2.

4. Discussion

4.1. Research Question 1

Research Question 1 focused on the PSTs’ beliefs and perceptions concerning the challenges they would face in teaching MLs in the K-12 context. The most prominent challenges for PSTs were (a) limited planning time, and (b) managing the diversity of tasks (open-ended Question 1, Theme 2, and open-ended Question 2, Theme 2). The identification of these challenges suggests that PSTs are experiencing a certain degree of angst about their future as teachers in the K-12 context. None of these challenges is surprising as anxiety is to be expected, particularly given that PSTs’ lack experience in teaching. Their field experience in an SEI classroom with at least 30% MLs was likely their first opportunity to experience what their lives as teachers would be like in this context (Pillen et al. 2013). Their ability to identify these challenges showed a high degree of situational perceptiveness. Exposing PSTs to the real-world of K-12 classrooms seemed to be an important factor in their development as teachers of MLs. It was helping them bridge the gap between the theories of teaching to which they had been exposed in their teacher education courses and the realities of educational practices in the context of classrooms with MLs. Even though anxiety is to be expected, the challenges that PSTs articulated need to be carefully evaluated by teacher educators and teacher education programs to determine if they are being explicitly and adequately addressed in their coursework. For example, where in the teacher education program do PSTs have an opportunity to develop strategies for managing time, planning for instruction, and dealing with non-teaching related tasks?

Pre-service teachers also expressed worries over issues related to classroom management (open-ended Question 1, Theme 4), particularly how they would deal with disruptive behavior. In considering these data, it is important for teacher educators to note that the classroom management issues that PSTs experienced likely stemmed from the fact that they did not feel capable of managing disruptions. They needed tools for classroom management, more time observing experienced teachers as they used these tools, and experience in implementing the tools (Theelen et al. 2020). The question for teacher educators and teacher education programs is to determine how to introduce, practice, and model the use of these tools.

Data from these PSTs also suggested that they were experiencing challenges relative to understanding the politics of public school (open-ended Question 1, Theme 2); they seemed to worry about the parents of MLs and about becoming involved in some type of conflict. While there may be a basis for these worries in the reality of public education, there are also many support systems available for teachers. Teacher educators need to be cognizant of where these support systems are introduced in the curriculum and how they are reinforced to PSTs. What was clear from these data was that PSTs likely did not understand that there were support systems available to them. They needed numerous positive experiences during their teacher education programs, interacting with parents and diverse communities, as well as more direct experience in classrooms with MLs to develop self-efficacy.

4.2. Research Question 2

Research Question 2 concentrated on PSTs’ beliefs, perceptions, emerging ideologies about multilingualism, and self-efficacy for educating MLs in the K-12 context. The classroom context, resulting from the SEI model, was seen as perhaps the greatest challenge for PSTs in working with MLs. The overwhelming majority of PSTs expressed concern about their abilities to teach native speakers of English alongside MLs in one classroom (open-ended Question 1, Theme 3; open-ended Question 2, Theme 1). Pre-service teachers’ perceptions of teaching in this context described the options for teaching from limited perspectives, for example stating that teaching native English speakers effectively would mean that MLs would not understand and adapting their lessons to accommodate MLs would mean that native English speakers would become bored. This emerging ideology that separated learners and learning into two distinct groups within the SEI classroom was concerning. Culturally and linguistically responsive and relevant teaching (Ladson-Billings 1995) and pedagogical translanguaging (Cenoz and Gorter 2021) are based on approaches to teaching in which all learners experience academic success and teachers see the benefits of linguistic and cultural diversity for all learners. To develop a cadre of PSTs who embrace diversity and are able to develop self-efficacy for working in the SEI context, teacher education programs in local contexts must wrestle with how to socialize PSTs to their roles as both content and language experts within their disciplines (open-ended Question 2, Theme 2). For most teacher education programs, achieving this goal will likely result in the need to place a greater emphasis on facilitating PSTs’ understandings of the complexities of language learning and expanding their knowledge of multilingualism and their skills relative to TLA.

The data collected in this study offered further insights into PSTs’ beliefs, perceptions, self-efficacy, and emerging ideologies about multilingualism and MLs (open-ended Question 2, Theme 2). Seventy-three percent of the PSTs self-identified as monolinguals even though eighty-eight percent indicated that they had studied a foreign language at some point in their educational history, likely in secondary school as admission to the university requires at least two years of a foreign language or an equivalent. They could have identified as bilinguals or as emerging multilinguals, which were concepts introduced and discussed in the TLA course. That the majority PSTs self-identified as monolinguals regardless of their experiences with languages other than English needs to be explored further. The reasons could be related to beliefs about their levels of language proficiency (Cook and Bassetti 2011). In other words, the PSTs might have believed that they would need to develop higher levels of language proficiency before they could consider themselves to be multilinguals. Not identifying as multilinguals could also be related to their beliefs about multilingualism and whether they thought it was necessary to use more than one language in their daily lives in school or with friends or family to be considered multilingual. Regardless of the reasons, the PSTs in this study were articulating an emerging ideology in which they were seeing themselves as separate or different from MLs. Monolingualism, as with bi- and multilingualism, is a manifestation of an individual’s background and experiences. Whether teachers self-identify as monolinguals or conceptualize their identities relative to languages in other ways, the responsibility for the teacher education program is to support all PSTs in developing self-efficacy, especially for working in SEI context with MLs.

These data show that PSTs are advancing in their understanding of teaching, as they articulated concerns about teaching that could only arise from experiences in real world contexts, for example concerns about how to respond to teaching in situ while also following a set instructional plan with specific goals and objectives or how to know when concepts should be recycled. When asked about the specific concepts they learned in the course that would be most useful for them in teaching, well over half of the teachers recognized the need for more exposure to TLA concepts, and some expressed desires to know more about pedagogical translanguaging (open-ended Question 2, Theme 3; open-ended Question 5). What was generally concerning about their responses was that very few PSTs were able to articulate specific concepts when asked what they still wanted to know about teaching in the SEI context. Most of the PSTs were still conceptualizing teaching in the abstract. The pre-service teachers’ self-efficacy for working with MLs midway through the course was concerning, with only a little over half indicating that they were well prepared and about one-fifth indicating that they were not prepared at all. In addition, 89% of the PSTs indicated that the TLA was the first course they had taken that focused on language, and it would likely be the only one most of them would take within their formal teacher education program as no other courses on TLA were required.

4.3. Research Question 3

Research Question 3 was aimed at understanding PSTs’ emerging identities as teachers midway through their teacher education program. Teacher professional identity is defined as the beliefs and values teachers hold about being teachers. These beliefs and values contribute to a coherent sense of who they are as teachers and how they engage with MLs and in the profession of language teaching. Pre-service teachers’ identities are evolving as they are shaped and reshaped by their experiences. Data from the PSTs in this study showed that PSTs recognized that their emerging identities as mainstream and content area teachers were sometimes in conflict with their identities as teachers of MLs (Open-ended Question 2, Theme 2). They were beginning to recognize the diverse roles and responsibilities they would need to assume as language experts in their disciplines and as teachers of MLs. Midway through their teacher education program, these PSTs were struggling to find a teacher identity that was inclusive of these diverse roles and responsibilities. The shaping and reshaping nature of teacher identity for PSTs in response to their experiences during their teacher education program was expected. However, teacher educators must recognize that for PSTs to develop identities that are inclusive of their roles and responsibilities as language experts in their disciplines, socialization must be integral to all courses and experiences within the teacher education program from the beginning. A major question for teacher educators who are preparing teachers to work in this context is how to accomplish this feat.

4.4. Research Question 4

Research Question 4 targeted the level of TLA that PSTs were able to achieve at the conclusion of the TLA course and midway through the teacher education program. The researcher made every effort to structure the language feature identification task in a way that would allow PSTs to submit their best work. They had access to the text in advance and could choose the language features in which they had the most confidence. The strategy seemed to work, as PTS exhibited a high level of accuracy for the features that were chosen the most frequently, such as adjectives, verbs, and prepositional phrases. Nevertheless, given the structure of the task, the overall scores for PSTs were low with a mean of only 6.5/10 and with some scores as low as 5.0/10.

It is interesting to note that while PSTs seemed to have developed a system for recognizing lexical categories, such as adjectives, verbs, conjunctions, and adverbs, they were struggling overall with constituent structure and making sense of how English is organized hierarchically. For example, they could identify a coordinating conjunction correctly in English but could not easily recognize a compound sentence. The same relationship was noticed with subordinating conjunctions and complex sentences. Pre-service teachers who struggle with these types of tasks will likely not be able to explain these relationships among language features satisfactorily to MLs or be able create instructional materials to meet MLs’ language demands.

The teacher-generated materials task that PSTs were asked to complete is one that is common for most classroom teachers and one in which PSTs in SEI contexts will engage on a regular basis when they begin teaching. Only about a quarter of the PSTs could be thought to have the necessary skills to complete such a task effectively midway through their teacher education program. It was clear from these results that PSTs still needed to develop TLA and also needed a stronger understanding of the relationship between theory and practice to execute PCK effectively. For the PSTs who scored five instead of six on this task, the missed point was most often because there was an insufficient or incorrect explanation concerning why or how the activity would be useful for MLs. Almost half of the PSTs had difficulty describing a specific activity. Instead, they described generic procedures, such as having students work in groups, complete a worksheet, or review vocabulary. Pre-service teachers who scored a 1 or 2 on the task also struggled with identifying specific language and content that would be difficult for MLs.

5. Conclusions

The research reported in this article focused on the voices of PSTs and on their perceptions of the issues and challenges that faced them as they prepared to teach MLs in the K-12 context. Of particular interest was PSTs’ self-efficacy for working with MLs and the development of knowledge and skills that underpin TLA. The results from the open-ended questions showed that many of the issues and challenges that PSTs identified were similar to the issues and challenges of novice teachers (Farrell 2012; Veenman 1984) in other contexts, for example, how to find balance in their personal and professional lives and how to manage diverse roles and responsibilities. The pre-service teachers in this study experienced tensions relative to the development of their identities as teachers.

The multiple data sources in this study also shed light on the development of TLA and PCK for PSTs. For the PSTs in this study, the issues and challenges pertained specifically to working in SEI and how to deliver instruction effectively in the classroom to both MLs and home language speakers of English, who were often monolinguals. Pre-service teachers were experiencing stress as they tried to meet both the language and content demands of diverse groups of learners.

What we know from these data is that the performances on the language feature and teacher-generated materials tasks, which are typical tasks required of teachers in SEI contexts with MLs, were too low for PSTs to be successful in preparing instructional materials for real world classrooms. Multilingual learners need teachers with high levels of TLA, so it is understandable that the PSTs in this study lacked self-efficacy. Teachers’ perceptions of their TLA mirrored the results from the task-based data, demonstrating that PSTs were accurate in their perceptions of their skills and knowledge of TLA. Most PSTs (89%) stated that the class was the first one that focused on TLA. Even though the results of the PSTs’ performances on the language feature and teacher-generated materials tasks were overall too low for them to be considered effective, the performances may represent what can reasonably be expected of PSTs with no prior experience in TLA. Relative to the design and development of teacher education courses for the multilingual world of the 21st century, it is necessary for teacher educators and teacher education programs to consider how to create programs wherein PSTs can develop high levels of TLA, along with PCK, and improve their self-efficacy for working with MLs. The results from this study show that, without additional work on developing TLA beyond one course, the PSTs in this study would exit the teacher education program and obtain licensure with the level of TLA that they exhibited midway through their program, which is a level of expertise in TLA that is insufficient to meet the language demands of today’s K-12 classrooms.