Connecting Language Proficiency to (Self-Reported) Teaching Ability: A Review and Analysis of Research

Abstract

This article provides a review and analysis of current research examining the connection between teacher language proficiency and their self-reported beliefs about their pedagogical abilities. Generally speaking, (English) language teachers require an advanced level of proficiency in order to be successful language teachers, but pedagogical skills are also necessary for effective instruction. However, examining the relationship between language proficiency and pedagogical skills is not straightforward given the nature of language teaching in which language is both the content and medium of instruction, issues with defining language proficiency (for teaching), and disagreement regarding effective teaching skills in different social contexts. One approach has been to focus on teachers’ self-efficacy beliefs about their abilities to enact specific classroom tasks in certain contexts and potential relationships with teachers’ reported language proficiency. The analysis provided in this article highlights how these constructs have been measured in the literature and if/how they are connected to one another across different studies. Overall, teachers’ self-efficacy beliefs regarding their pedagogical abilities do correlate with language proficiency, but results are at times weak and/or inconclusive with inconsistent results across studies and contexts.

Keywords

Teacher language proficiency teacher self-efficacy

Introduction

Content knowledge and pedagogical (content) knowledge are key components of a teacher’s knowledge base. With regards to language teaching, target language proficiency and teaching ability are the two most significant aspects of teacher competence. Proficiency in the target language allows teachers to maximize their language use in the classroom, which in turn provides opportunity for the required second/foreign language input crucial to second language learning (Turnbull, 2001). However, there is no agreed upon level of proficiency for language teachers. Different contexts have different definitions of proficiency and proficiency requirements are not the same across contexts (Dogancay-Aktuna and Hardman, 2012). Furthermore, high proficiency in a language alone does not make one an effective teacher. Teachers need to develop effective communication and pedagogical skills (Coelho, 2004; Faez, 2008). While there are several assertions on the importance of target language proficiency for language teaching (e.g. Faez, 2011a; Pasternak and Bailey, 2004; Salvatori, 2009), there is little known about their relationship. How does target language proficiency impact teachers’ instructional abilities in the classroom? This is not an easy investigation and there are several challenges associated with researching this relationship including:

Language teaching is different from other subject matter teaching in that language is both the content and medium of instruction.

Measuring teacher language proficiency and language teaching abilities is difficult and contentious due to validity issues, among other concerns.

Issues of ownership of English, English as a Lingua Franca, World Englishes, and Native/Non-native speaker¹ issues all further complicate the matter.

Looking at studies that focus on language teacher (self-) efficacy and language proficiency, this report provides a review and analysis of current research that has attempted to investigate the relationship between teachers’ confidence in their classroom abilities and their (self-reported) language proficiency.

Teacher (Self-) Efficacy

In general education, teacher efficacy research has generally followed two conceptual strands: 1) Locus of control which focuses generally on teachers’ abilities to affect outcomes, and 2) Sociocognitive perspective developed by Bandura (1977; 1986; 1997) which looks at teachers’ confidence in their abilities to enact certain tasks in specific contexts (Tschannen-Moran et al., 1998). While this division has been noted by researchers, the two concepts are often confused with one another causing uncertainty and methodological incongruence (Klassen et al., 2011). The issue appears, at times, that researchers claim to measure self-efficacy, but instead base items on teachers’ general abilities to impact the outcomes of events, which would pertain to teacher efficacy. Bandura (1977; 1986) initially outlined his sociocognitive self-efficacy perspective that pertained to a person’s ability to enact tasks, but many focused on the different locus of control construct. Bandura (1997) attempted to clear up the confusion by reaffirming his sociocognitively based self-efficacy and noting its difference to outcome expectancy, but studies still confuse the two concepts (Wyatt, 2014) leading authors to label efficacy as an ‘elusive construct’ (Tschannen-Moran and Woolfolk Hoy, 2001).

One of the major contributors to the confusion surrounding efficacy work is the use of the Teacher Efficacy Scale (TES) created by Gibson and Dembo (1984). Prior to the TES, numerous efficacy scales were created to investigate efficacy in general education, but the majority of the scales were never used outside of their original studies (Tschannen-Moran et al., 1998). The TES was different however and was adapted by numerous studies to measure teachers’ (self-) efficacy. The TES contains two subscales measuring: Personal Teaching Efficacy (PTE) and General Teaching Efficacy (GTE). Conceptual issues have been noted as while PTE (somewhat) measures self-efficacy, GTE is more focused on the locus of control construct and teachers’ abilities to influence outcomes. On its own, this is not an issue, but Gibson and Dembo (1984) claim to follow Bandura’s self-efficacy framework and many studies have used the TES as a measure of self-efficacy in the Bandurian sense, which is not appropriate (Klassen et al., 2011). This incongruence has led some authors to conclude that the results from studies that use the TES may be invalid (Dellinger et al., 2008; Klassen et al., 2011). Thus, while measuring General Teacher Efficacy is appropriate in its own right, misconstruing it with self-efficacy is conceptually inaccurate. This is significant because self-efficacy is a more powerful construct. People’s beliefs in their abilities to enact tasks are greater predictors of behaviour than their belief to enact outcomes (Bandura, 1997). Even if they desire a certain outcome, if they do not believe they can carry out the necessary task, it is likely that their behaviour will not drastically change (Bandura, 1997).

To alleviate the methodological issues created by the use of the TES, researchers at Ohio State University created the Ohio State Teacher Efficacy Scale, later referred to as simply the Teacher Sense of Efficacy Scale (TSES) (Tschannen-Moran and Woolfolk Hoy, 2001). The TSES was created for use in general education and, for many researchers (e.g. Dellinger et al., 2008), solved the conceptual issues surrounding self-efficacy and teacher efficacy (although see Wyatt, 2014). The scale is grounded in Bandura’s sociocognitive self-efficacy theory and has both a short form version consisting of 12 items, and a long form version with 24 items (Tschannen-Moran and Woolfolk Hoy, 2001). It consists of three subscales: 1) Efficacy for Instructional Practices, 2) Efficacy for Student Engagement, and 3) Efficacy for Classroom Management. Within these three factors, it provides simple questions for teachers to assess their self-efficacy regarding a specific ability. For example, from the Efficacy for Instructional Practices subfactor, one item asks: ‘To what extent can you craft good questions for your students?’ (Tschannen-Moran and Woolfolk Hoy, 2001: 800). Some studies alter these questions into statements, but the general content of the item remains the same. The authors use the terms teacher efficacy and self-efficacy interchangeably, which is somewhat misleading as the scale is focused on self-efficacy in the Bandurian sense, not general teaching efficacy, but the TSES quickly became a highly used efficacy scale.

The TSES has been widely accepted in efficacy research (Dellinger et al., 2008), but issues surrounding the specificity of the scale remained. In efficacy research, how specific to make survey items is an issue of continuing contention as researchers attempt to find the right balance. Bandura (2006) makes the point that scales that are too general do not provide adequate measurement of self-efficacy, but on the other hand, scales that are too focused will not have general use. Pajares (1996) notes that self-efficacy scales need to measure context specific tasks; general scales and even domain specific scales are often inadequate. This issue is prevalent when discussing how to use the TSES for other contexts. To alleviate this issue, researchers have created their own subject-specific scales for their own areas, for example the Science Teaching Efficacy Belief Instrument (STEBI) (Riggs and Enochs, 1990) or adapted the TSES for their own context matter. Many studies have adapted the TSES and the TES to investigate (English) language teachers because, until recently, no (English) language teaching specific scale existed.

Measuring Proficiency and (Self-) Efficacy

As efficacy research began in the foreign/second language teaching context, considering the importance of language proficiency, researchers began to investigate the relationship between (self-) efficacy and proficiency. However, the issue of how to measure proficiency and self-efficacy is still unclear. Self-efficacy research measures teachers’ efficacy beliefs about pedagogical capabilities in specific contexts (Dellinger et al., 2008; Tschannen-Moran and Woolfolk Hoy, 2001), but content knowledge such as language proficiency, should not be included (Choi and Lee, 2016). With this perspective, many studies measure teachers’ self-reported proficiency with separate scales (e.g. Chacon, 2005; Choi and Lee, 2016; Eslami and Fatahai, 2008). Some of these self-report proficiency scales are developed by the researcher (e.g. Chacon, 2002; 2005; Shim, 2001), while others adopt scales from other studies (e.g. Eslami and Fatahi, 2008), the most common of which appears to be the proficiency scale developed by Butler (2004). However, the line between distinguishing language as content knowledge and as a pedagogical skill/capability is not clear. Language is content in the language classroom, but it can also be the medium of instruction and can be deemed a pedagogical capability. This blurred division has led some studies to measure proficiency as part of self-efficacy. For example, Swanson (2010; 2012), on the Second/Foreign Language Teacher Efficacy Scale (S/FLTES), seemingly investigates language proficiency items under Content Knowledge. As an example, one item asks participants to assess their ability to ‘fully understand a movie that only uses the language(s) you teach?’ (Swanson, 2012: 87). Clearly, this can be construed as an assessment of listening proficiency. Choi and Lee (2016) argue that proficiency and efficacy are interdependent, but independent constructs, and thus argue against the inclusion of proficiency items on self-efficacy scales. However, drawing such clear distinction appears difficult. Still, most studies do differentiate between the two and use separate scales.

Furthermore, for the most part, studies use self-report measures to assess language proficiency. While these do not measure teachers’ actual proficiency, many researchers have noted that teachers’ assessments of their own proficiency are often close assessments of their actual ability (Blanche and Merino, 1989; LeBlanc and Painchaud, 1985; Ross, 1998). However, some researchers would dispute this and question the objectivity of self-assessments (Denies and Janssen, 2016). Teachers may compare themselves to native speaker models and feel inadequate in comparison, which could impact their language confidence and potentially how they assess their own proficiency (Hodgson, 2014). Yet, measuring actual proficiency is also fraught with difficulties. Elder (2001) notes the difficulty in measuring actual teacher proficiency and the potential issues (further elaborated on below). For this review, only one study was found that measured actual proficiency through a language test (Sabokrouh, 2014) and how it affected teachers (self-) efficacy regarding their classroom abilities and skills. The remaining studies all used self-reported proficiency assessments.

Methodology

This article embarked on a review of past literature regarding the relationship between (English) language teacher proficiency and teacher (self-) efficacy. As a starting point, an initial search for literature began with a keyword search using phrases including ‘efficacy’ or ‘self-efficacy’, ‘proficiency’, and ‘language teacher’. Various combinations were used on a university library site with access to major educational databases (e.g. ERIC, Proquest et al.). Journal articles, books and theses were all scanned and potentially relevant resources were noted after reading the titles and abstracts. To be included in the review, resources needed to look specifically at language teacher (self-) efficacy along with language proficiency in some capacity. There are numerous studies that focus on some aspect of (English) language teacher (self-) efficacy, and many studies address the issue of proficiency, but this review sought only studies that investigated the relationship between these two factors. Many studies specifically address proficiency using various scales, but some focus on comparing native/non-native speaking language teachers. Despite not specifically referring to proficiency, these studies were included because it was decided that their different scale of measurement (native and non-native) was a methodological choice, and still in-line with this review’s inclusion criteria of focusing on the interaction between teacher efficacy and proficiency. An initial group of studies was accumulated and the reference lists of these studies were reviewed to add any further resources that were not found during the library searches.

After these initial searches, a total of 13 studies were accumulated. All of the studies found were either quantitative (e.g. Eslami and Fatahi, 2008) or mixed methods (e.g. Lee, 2009) in nature. Similar to Basturkmen (2012), this review followed an inclusive approach as advocated by many research synthesists in the field of Applied Linguistics (Ortega, 2015; Plonsky, 2016). With such an approach, any study that fits the criteria for inclusion, in this case measuring the relationship between teacher (self-) efficacy and language proficiency, is included. In an inclusive approach, both published (e.g. journal articles) and unpublished works (e.g. doctoral dissertations) are included. Utilizing an inclusive approach, there is the possibility of including research that may be considered ‘low quality’, but in line with Norris and Ortega (2006), this review treats ‘methodological quality as an empirical matter’ (2006: 19). While concern over research quality is certainly justifiable, if judgment about study quality is left to only those who carry out the research review, certain hidden biases may result, which could equally impact the quality of the review in a negative way. Furthermore, in regards to published versus unpublished work, an unpublished work (e.g. dissertation) can be ‘of equal or better quality than many published ones’ (Norris and Ortega, 2006: 20). Considering the studies were all quantitative or mixed methods, the risk of publication bias was also considered as studies that do not report statistically significant results are often deemed not publishable (Norris and Ortega, 2006; Ortega, 2015). Thus, both published and unpublished work were considered and the quality of research was left to be analysed as part of the review. With this consideration, two of the initial 13 studies included master level theses, but these were excluded due to potential concerns regarding limited library resources, leaving the final number at 11 studies.

As mentioned, in total, 11 studies fit the above criteria and are used for this review. Six of these are doctoral dissertations (Chacon, 2002; Crook, 2016; Lee, 2009; Liaw, 2004; Praver, 2014; Shim, 2001), and six are from academic journals (Chacon, 2005; Choi and Lee, 2016; Eslami and Fatahi, 2008; Ghasemboland and Hashim, 2013; Sabokrouh, 2014; Yilmaz, 2011). The above list contains 12 citations, but because Chacon (2002; 2005) draw from the same data, they are considered as one study, thus making the final number 11. All of the studies focus on English language teaching contexts except for Liaw (2004) who focuses on foreign language (e.g. Chinese, Italian) teachers in the United States. Most studies specifically measure self-reported language proficiency with some form of a scale, but two studies (Liaw, 2004; Praver, 2014) compare native and non-native speaking teachers and do not use any specific proficiency scale. Sabokrouh (2014) is also unique as this study measures actual language proficiency instead of self-reported proficiency.

A handful of studies were found that investigated language teacher (self-) efficacy, but did not fit the narrow criteria to be included in the review. The work of Swanson (2010; 2012; 2014) falls under this category. Swanson (2010) developed the Foreign Language Teacher Efficacy Scale (FLTES), later adjusted to the Second/Foreign Language Teacher Efficacy Scale (S/FLTES) (Swanson, 2012). The creation of the S/FLTES (Swanson, 2010; 2012) was one of the first efforts to create an efficacy scale that was specifically aimed at foreign/second language teaching. The initial scale consisted of ten items across two factors: Teacher as Facilitator and Content Knowledge, but a third factor was added, Cultural Instruction, when it was realized that culture was a unique factor for language teaching. Swanson’s work was excluded because although it focuses on language teacher self-efficacy, it does not address any links with language proficiency. For example, looking at Spanish language teachers in the United States, Swanson (2010) notes a correlation between teacher self-efficacy to teach beginner level students and teacher attrition. Looking at foreign language teachers in the United States and French as a Second Language (FSL) teachers in Canada, Swanson (2012) notes a correlation between Efficacy for Classroom Management and Cultural Instruction and teacher attrition. Later, Swanson (2014) investigates the impact of teacher self-efficacy on student achievement. Focusing on Spanish teachers in the United States, Swanson divided teachers into groups based on their efficacy scores, and notes that students whose teachers had the highest efficacy scores had higher scores on the National Spanish Examination (Swanson, 2014). This serves as an initial study linking language teacher efficacy with student achievement. To reiterate, while these studies are insightful, they do not examine links between proficiency and self-efficacy, and were thus, excluded from the overall review.

Summary of Reviewed Articles

Table 1 (below) provides a general overview of the studies included in this review. The results show that the vast majority of studies show a positive correlation between language proficiency and teacher (self-) efficacy. Only one study (Shim, 2001) found (almost) no relationship between the two elements. Shim (2001) was one of the earliest studies to focus on (English) language teacher self-efficacy, and the first of the Ohio State theses (more on these below) to focus on English language teacher efficacy. As mentioned, the TSES was developed in a graduate seminar at Ohio State and reached publication in 2001 (Tschannen-Moran and Woolfolk Hoy, 2001), but Shim (2001) used the TES scale by Gibson and Dembo (1984). The issues with the TES have been discussed, but concerns with Shim’s (2001) study also pertain to scale specificity. None of the items on Shim’s (2001) survey were specific to English language teaching, which is one potential reason why few correlations were found between proficiency and self-efficacy (Lee, 2009).

Table 1.

Summary of Articles included in Review.

Article	Context	Participants	Measure of Proficiency	Measure of Skills (Self-efficacy)	Reported Results
Shim (2001)	South Korea	106 Korean EFL middle school and high school teachers.	18-item survey created by the researcher measuring: Listening, reading, speaking and writing.	TES – Reduced ten item version.	No statistically significant correlation existed between PTE or GTE and all four language skills. Only listening and reading had mild correlations with PTE.
Chacon (2002; 2005)	Venezuela	100 Venezuelan EFL middle school teachers.	16-item survey drawn from literature and researcher experience measuring: Listening, reading, speaking writing and sociolinguistic/cultural knowledge.	TSES-12 item (adapted for EFL context) STEBI (13 items adapted) 2 RAND items	Self-reported English language proficiency was positively correlated with teachers’ self-efficacy.
Liaw (2004)	United States	104 native and non-native language teaching assistants (Spanish, Chinese, Japanese, German, French and Italian) at Midwestern American university.	No measure of proficiency used. Makes distinction between native and non-native teachers for comparison.	Items drawn from TSES and STEBI (Classroom Management items removed because they do not apply to university setting). Items also drawn from literature, prior teaching experience and Chacon (2002).	NESTs felt capable teaching high level reading and writing classes, while NNESTs felt less capable. NESTs also more efficacious to teach speaking, colloquial language and knowledge related to culture.
Eslami and Fatahi (2008)	Iran	40 Iranian EFL high school teachers.	12-item survey from Butler (2004) and Chacon (2005).	TSES – 12- item (adapted and translated in Farsi).	Teachers’ perceived efficacy was positively correlated with self-reported English proficiency.
Lee (2009)	South Korea	1327 Korean EFL elementary school teachers.	Four language domains drawn from Butler (2004) measuring: Listening, reading, oral fluency and writing.	TSES – 12-item (adapted to fit Korean context) six new items under Oral English Use added to make an 18-item scale. PTE – five items adapted from Hoy and Woolfolk Hoy (1993)	Teachers’ self-reported level of English proficiency was a significant predictor of their teaching efficacy beliefs.
Yilmaz (2011)	Turkey	54 Turkish EFL teachers.	12 statements from Butler (2004) and Chacon (2005).	TSES – 12- item (adapted for Turkish context).	Teachers’ efficacy was positively correlated with their self-reported English proficiency.
Ghasemboland and Hashim (2013)	Middle East (country not specified)	187 English teachers in English language centres in one Middle East country (not specified).	21-item survey drawing on Chacon (2002; 2005) and Shim (2001).	TSES – 12- item (adapted for context).	Teachers perceived efficacy was positively and significantly correlated with self-reported English proficiency.
Praver (2014)	Japan	392 EFL teachers in Japanese university.	No specific measurement for proficiency. A comparison of native and non-native speaker teachers.	TSES – 24- item (adapted for context) New efficacy factor: Dealing With Superiors.	Native English language teachers perceived themselves to be more efficacious than Japanese English teachers across all four self-efficacy variables.
Sabokrouh (2014)	Iran	68 Iranian EFL private school teachers.	TOEFL test – PBT version. Listening and writing were removed. Only grammar and reading were included.	TSES – 18 items drawn from 24- item scale.	A significant, but weak, relationship between English language proficiency and teacher self-efficacy.
Choi and Lee (2016)	South Korea	167 Korean EFL high school teachers	Butler (2004) is modified and measured: Listening, speaking, reading, writing, vocabulary, pronunciation and grammar. Asked teachers to self-report proficiency and minimum level they believe is needed to teach in Korean context. The gap between these two levels is operationalized as proficiency.	Adapted 21- item scale drawn from TSES and Dellinger et. al (2008). Also create one item.	Authors note proficiency and efficacy had an effect on English use in the classroom. Note a 4.44 threshold level out of six. When self-efficacy values were at, or above, this level, then proficiency served as a predictor of English use.
Crook (2016)	Thailand	257 Thai teachers in elementary and high schools.	Chacon (2005) proficiency scale	TSES – 12-item short form (translated into Thai).	Proficiency and self-efficacy were significantly correlated.

Noting Shim (2001), a closer look at the doctoral dissertations that emerged from Ohio State University (e.g. Chacon, 2002; Lee, 2009; Liaw, 2004) is useful. The dissertations are interesting as they stem from the same conceptual base at Ohio State, where the TSES was created and the studies show a methodological progression as to how proficiency and self-efficacy are measured. As mentioned, Shim’s (2001) study was the earliest of these dissertations and used the general, and conceptually incongruent, TES. Chacon (2002) was next and used the TSES, but only altered the scale slightly for the Venezuelan EFL context leaving the TSES largely intact for her study. Liaw (2004) is the next dissertation and shows some progression in terms of scale development as this study is one of the first attempts to include language teaching specific efficacy items. Along with including some modified items from Chacon (2002), Liaw draws on past literature and teaching experience to create some efficacy items for the study. Finally, Lee (2009) shows an even greater methodological progression as this dissertation includes a new efficacy factor, Oral English Use. Scanning the Ohio State dissertations, the scales gradually go from very general (Shim, 2001) to more (English) language teacher relevant with Lee’s (2009) study.

Liaw (2004) and Praver (2014) are somewhat comparable as they focus on nativeship and its correlations with self-efficacy. Looking at foreign language instructors at an American university, Liaw (2004) found native-speakers to be more efficacious in their abilities to teach high level reading and writing classes, while non-native speaking teachers felt less capable. Native-speakers also rated themselves higher on ability to teach speaking, colloquial language and knowledge related to culture compared to non-natives. In Japan, the native-English speaking teachers reported higher levels of self-efficacy compared with non-native English-speaking Japanese instructors (Praver, 2014). However, cultural issues are noted as a potential reason for this, specifically Japanese humility. Self-efficacy scales are often phrased with ‘I can’ statements, and Praver (2014) argues that Japanese collectivist culture and modesty likely impacted self-efficacy scores somewhat. For all studies in this review, the teaching context also needs to be added as a consideration.

Comparison of the other studies was not as straightforward as they used either different analysis or different scales. For example, Choi and Lee (2016) do not use self-reported proficiency alone like most studies, but rather, use the gap between teachers’ self-reported language proficiency and the minimum level they believe is needed to effectively teach English in the Korean context. They report that minimum threshold levels exist for both language proficiency and self-efficacy and only beyond minimum threshold levels, there is a positive association between the two attributes. They note a significant positive correlation between the language proficiency gap and self-efficacy beliefs about pedagogical capabilities, but a weak one (0.18). Sabokrouh (2014) only measures grammar and reading proficiency, but uses a measure of actual proficiency instead of self-reported proficiency. Similar to Crook (2016), she measures an overall score of proficiency with efficacy, noting a significant but very low correlation (0.16). Finally, although Crook (2016) uses the TSES, individual subscale data pertaining to the TSES subscales and individual language skills (e.g. listening etc.) is not reported. However, the study does report an overall positive and significant relationship between teacher self-efficacy and proficiency (0.39). Thus, while the studies mentioned are insightful, the methodological differences make comparison difficult.

One advantage of numerous studies using the TSES is that it allows for simple comparison. Table 2 shows the correlation results for self-reported proficiency in four language subskills (listening, reading, writing, and speaking) with the different TSES efficacy subscales (Classroom Management, Instructional Strategies, and Student Engagement) from five studies (Chacon, 2005; Eslami and Fatahi, 2008; Ghasemboland and Hashim, 2013; Lee, 2009; Yilmaz, 2011). While many of the studies use the TSES, these five studies draw correlations between at least the above mentioned four language skills and the individual TSES subscales. Any correlations below 0.20 were not included in Table 2. Depending on the definition, correlations between 0.20–0.35 are considered as ‘very slight relationships’ (Cohen et al., 2011: 636), or generally, a threshold of 0.30 is a weak correlation. As can be seen from Table 2, many of the studies report very weak correlations (e.g. Chacon, 2005; Eslami and Fatahi, 2008). None of the correlations in Chacon (2005) or Eslami and Fatahi (2008) go over 0.40, and Chacon (2005) reports no significant correlational data. Lee’s (2009) correlations are somewhat stronger, but still are only in the mid-range. Yilmaz (2011) reports some significant correlations between Classroom Management and listening and writing, along with Instructional Strategies and listening and writing, but the correlations are low and very weak correlations are reported on the other factors. Ghasemboland and Hashim (2013) report the strongest correlations. Across all four language skills and the three efficacy subscales, they report strong and significant correlations, but they offer little indication as to why this may be. Thus, the results appear somewhat inconclusive, and although there are some positive correlations, they are often weak and not statistically significant.

Table 2.

Correlations between TSES Factors and Language Skills*.

Language Subskill	Classroom Management	Student Engagement	Instructional Strategies
Speaking	0.31* (Eslami and Fatahi, 2008) 0.32 (Lee, 2009) 0.83** (Ghasemboland and Hashim, 2013)	0.20 (Yilmaz, 2011) 0.23 (Eslami and Fatahi, 2008) 0.34 (Chacon, 2005) 0.37 (Lee, 2009) 0.81** (Ghasemboland and Hashim, 2013)	0.39 (Chacon, 2005) 0.39* (Eslami and Fatahi, 2008) 0.49 (Lee, 2009) 0.78** (Ghasemboland and Hashim, 2013)
Listening	0.20 (Eslami and Fatahi, 2008) 0.34* (Yilmaz, 2011) 0.35 (Lee, 2009) 0.82** (Ghasemboland and Hashim, 2013)	0.22 (Yilmaz, 2011) 0.29 (Chacon, 2005) 0.37 (Lee, 2009) 0.84** (Ghasemboland and Hashim, 2013)	0.30* (Eslami and Fatahi, 2008) 0.34 (Chacon, 2005) 0.39* (Yilmaz, 2011) 0.49 (Lee, 2009) 0.80** (Ghasemboland and Hashim, 2013)
Reading	0.20 (Eslami and Fatahi, 2008) 0.20 (Yilmaz, 2011) 0.31 (Lee, 2009) 0.80** (Ghasemboland and Hashim, 2013)	0.26 (Chacon, 2005) 0.31 (Lee, 2009) 0.83** (Ghasemboland and Hashim, 2013)	0.22 (Eslami and Fatahi, 2008) 0.25 (Yilmaz, 2011) 0.41 (Chacon, 2005) 0.44 (Lee, 2009) 0.82** (Ghasemboland and Hashim, 2013)
Writing	0.23 (Chacon, 2005) 0.27 (Lee, 2009) 0.28* (Yilmaz, 2011) 0.89** (Ghasemboland and Hashim, 2013)	0.30 (Chacon, 2005) 0.30 (Lee, 2009) 0.92** (Ghasemboland and Hashim, 2013)	0.35* (Yilmaz, 2011) 0.38 (Chacon, 2005) 0.38* (Eslami and Fatahi, 2008) 0.44 (Lee, 2009) 0.90** (Ghasemboland and Hashim, 2013)

= Significant at the 0.05 level.

= Significant at the 0.01 level.

NOTE: Correlations are listed in ascending order.

Only five studies provided appropriate data for Table 2 (Chacon, 2005; Eslami and Fatahi, 2008; Ghasemboland and Hashim, 2013; Lee, 2009; Yilmaz, 2011).

Discussion

Ten out of 11 studies reviewed in this article reported a positive correlation between language proficiency and teachers’ confidence in their classroom abilities. Based on these studies, in most cases, teachers who perceived themselves to have higher target language proficiency also had more confidence in their instructional abilities. In looking closer at the data of the studies, rather than what authors have reported, the correlations reported in most of the studies (e.g. Chacon, 2005; Eslami and Fatahi, 2008; Lee, 2009; Yilmaz, 2011) are not particularly strong; in fact, they are moderate/low (between 0.2–0.4). A high correlation (between 0.7–0.9) was reported in just one study i.e. Ghasemboland and Hashim’s (2013). There are also few studies that report that the correlation is statistically significant (Eslami and Fatahi, 2008; Ghasemboland and Hashim, 2013, Yilmaz, 2011) and this information is missing from other studies (e.g. Chacon, 2005). As noted before, this may partly be due to the measures used to assess teachers’ language proficiency and teaching efficacy. Only one study used an actual language proficiency test (Sabokrouh, 2014), and there were no studies that used a (validated) performance-based (Douglas, 2000; Elder, 2001) measure of language proficiency. The other studies all used self-reported proficiency measures.

Elder (2001) refers to the ‘blurred boundaries’ in attempting to assess language proficiency of teachers and highlights challenges associated with characterizing teacher proficiency as:

Specificity (defining the domain of teacher proficiency and separating it from other areas of professional competence)

Authenticity (eliciting real teacher-like language)

Inseparability (the role of non-linguistic factors such as strategic competence or teaching skills).

The studies reviewed have predominantly used a self-reported scale of teachers’ abilities in the four language skills (listening, reading, speaking and writing), and while others may gather information regarding other language skills, proficiency is operationalized as an overall score (Crook, 2016) or as a gap between reported and minimum levels (Choi and Lee, 2016). More importantly, due to construct validity issues, the use (and relevance) of general teaching efficacy scales such as TSES for measuring teacher efficacy in language teaching is problematic. The TSES is a validated and widely used measure of teacher efficacy for mainstream classrooms. The three factors in TSES include Efficacy for Instructional Strategies, Classroom Management and Student Engagement. There is more to language teaching than these factors combined. The intricacies of language teaching, such as teaching pronunciation and vocabulary or providing form-focused instruction and corrective feedback, for example, are not captured by general teaching efficacy scales. In some studies, there have been attempts to adapt the TSES for use in specific contexts. For example, Chacon (2002; 2005) altered some statements on the TSES to make them more specific to the Venezuelan ELT context. Liaw (2004) drew on literature and teaching experience to add new items alongside the TSES scale. Lee (2009) and Praver (2014) added new factors to the TSES to make the scale more specific to the Korean and Japanese contexts where their studies were conducted. These have been worthwhile attempts, but they do not address the task-specific measures needed of self-efficacy scales (Pajares, 1996), but rather convert a general scale to a domain specific scale that still lacks task specificity. The complexities of language teaching are still not investigated with these broader scales.

There have been recent attempts to address this issue. Because the TSES is a general education scale, researchers have sought to create (English) language teacher specific scales that focus on the specific skills and knowledge required of (English) language teachers. Swanson’s (2010; 2012) S/FLTES has already been discussed. More recently, Akbari and Tavassoli (2014) created a 32-item English language teacher efficacy scale that is partially influenced by the TSES. It differs from other language teacher efficacy scales in that it uses vignettes drawn from Ashton, Olejnik, Crocker and McAuliffe’s (1982) work to present the items. Generally, the vignette approach is not as ideal as ‘can do’ statements proposed by Bandura (2006) which can tap into one’s confidence in their abilities to accomplish a certain task. Also, the vignettes and the scale items appear specific to the Iranian context and may not align with areas recognized for teaching English in other contexts. While the authors outline the scale development and validation, no studies have been found that actually use the scale. Rezvani and Amiri (2013) also created an ELT-specific efficacy instrument in Iran, and Shim (2006) for the Korean context, but again, no studies were found to have used these scales either.

Efficacy research is predominantly quantitative and dominated by self-reports using Likert-scale items. The development of validated instruments for assessing English proficiency and English language teaching efficacy are certainly steps in the right direction but at best, the accuracy of beliefs, elicited through self-reported instruments is a major concern (See Wyatt, 2014 for a similar discussion). For various reasons, teachers may over or under-estimate their abilities in language or teaching skills. Triangulation, using mixed-methods, including in-depth interviews and observations, can provide new insights and help with interpretation of quantitative data. In a qualitatively oriented study, Faez (2010) reported that some NNESTs stated that language proficiency was a major concern for them and believed that their level of language proficiency affected their instructional ability in the classroom. The purpose of this study was not to compare the relationship between language proficiency and instructional ability, therefore it is not included in this review, but instead the purpose is to examine the experiences of teachers from diverse linguistic and cultural backgrounds as they prepare to teach in diverse multilingual classrooms. However, in light of the inconsistent findings of some of the studies reviewed in this article, this study shows the significance of qualitative data coupled with quantitative findings.

Qualitative data can also help add new information regarding contextual issues. As noted at the beginning, there is no agreed upon standard of proficiency and each context requires different types of proficiency in order to teach (Dogancay-Aktuna and Hardman, 2012). For example, Lee (2009) noted weak correlations between speaking and Classroom Management, but the interview data revealed that much of classroom management was carried out in Korean and did not require English. Also using interviews, Chacon (2005) found the Venezuelan EFL teachers in her study did not feel proficient enough to use communicative language teaching strategies, instead opting for more grammar-based methods that emphasized grammatical accuracy over fluency. These studies show that incorporating qualitative data can reveal the contextual intricacies not found in quantitative data and provide a more nuanced look at how proficiency and teachers’ beliefs in their own ability may interact, and even impact classroom behaviour.

The other issue, not particularly addressed in the studies reviewed in this report, but worth noting, are issues pertaining to the use of English as a Lingua Franca and/or as an International Language. While a discussion on the origin and controversy around these terms is beyond the scope of this review (see Faez and Taylor, 2017), it is perhaps important to note that in a world in which multiligualism is the norm and its users employ many different mother tongues (L1s) and a variety of ‘Englishes’, measures of language proficiency and teacher efficacy should reflect the context-specific norms and standards.

Conclusion

Language proficiency is generally perceived to be one of the most important qualifications of a language teacher by experts and non-experts. The review of studies presented in this article lends support to this position. The analysis showed that generally there was a positive correlation between teachers’ self-reported language proficiency and instructional abilities. One caveat to this conclusion, however, is that the correlations were generally not strong and not all studies reported whether they were significant or not. This article discussed several issues with the measures teachers used to appraise their proficiency and instructional ability. Converting general efficacy scales (like the TSES) for use in (English) language teaching contexts has its limitations and future studies will need to consider the specific tasks required of (English) language teachers. In combination with different measures of proficiency, both actual and self-reported, such research could add further details about the interaction of language teacher proficiency and teachers’ (self-reported) pedagogical abilities.

Footnotes

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Akbari

Tavassoli

(2014) Developing an ELT context-specific teacher efficacy instrument. RELC Journal 45(1): 27–50.

Ashton

Olejnik

Crocker

McAuliffe

(1982) Measurement problems in the study of teachers’ sense of efficacy. Paper presented at the Annual Meeting of the American Educational Research Association, New York.

Bandura

(1977) Self-efficacy: toward a unifying theory of behavioral change. Psychological Review 84(2): 191–215.

Bandura

(1986) Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice-Hall Inc.

Bandura

(1997) Self-efficacy: The Exercise of Control. New York: W.H. Freeman & Company.

Bandura

(2006) Guide for constructing self-efficacy scales. In: Pajares

Urdan

(eds) Self-efficacy Beliefs of Adolescents. Greenwich, CT: Information Age Publishing, 307–337.

Basturkmen

(2012) Review of research into the correspondence between language teachers’ stated beliefs and practices. System 40(2): 282–295.

Blanche

Merino

(1989) Self-assessment of foreign language skills: implications for teachers and researchers. Language Learning 39(3): 313–40.

Butler

(2004) What level of English proficiency do elementary school teachers need to attain to teach EFL? Case studies from Korea, Taiwan, and Japan. TESOL Quarterly 38(2): 245–78.

10.

Chacon

(2002) Teachers’ sense of efficacy and selected characteristics of selected English as a foreign language Venezuelan middle school teachers. Unpublished doctoral dissertation, Ohio State University, Columbus, Ohio, USA.

11.

Chacon

(2005) Teachers’ perceived efficacy among English as a foreign language teachers in middle schools in Venezuela. Teaching and Teacher Education 21(3): 257–272.

12.

Choi

Lee

(2016) Investigating the relationship of target language proficiency and self-efficacy among nonnative EFL teachers. System 58: 49–63.

13.

Coelho

(2004) Adding English: A Guide to Teaching in Multilingual Schools. Toronto: Pippin.

14.

Cohen

Manion

Morrison

(2011) Research Methods in Education (7th ed). London: Routledge.

15.

Crook

(2016) The predictive relationship between specific teacher characteristics and the perceived sense of teacher self-efficacy of non-native English speaking teachers of English as a foreign language in rural Thailand. Unpublished doctoral dissertation, Liberty University, Lynchburg, VA, USA.

16.

Dellinger

Bobbett

Olivier

Ellet

(2008) Measuring teachers’ self-efficacy beliefs: Development and use of the TEBS-Self. Teaching and Teacher Education 24(3): 751–766.

17.

Denies

Janssen

(2016) Country and gender differences in the functioning of CEFR-based can-do statements as a tool for self-assessing English proficiency. Language Assessment Quarterly 13(3): 251–76.

18.

Dogancay-Aktuna

Hardman

(2012) Teacher education for EIL: working toward a situated meta-praxis. In: Matsuda

(ed.) Principles and Practices of Teaching English as An International Language. Bristol, CT: Multilingual Matters.

19.

Douglas

(2000) Assessing Languages for Specific Purposes: Theory and Practice. Cambridge: Cambridge University Press.

20.

Elder

(2001) Assessing the language proficiency of teachers: are there any border controls? Language Testing 18(2): 149–70.

21.

Eslami

Fatahi

(2008) Teachers’ sense of self-efficacy, English proficiency, and instructional strategies: a study of non-native EFL teachers in Iran. TESL-EJ 11(4): 1–19.

22.

Faez

(2008) Support for English language learners in kindergarten through twelfth grade classrooms: are internationally educated teachers better positioned? In: Mantero

Miller

Watzke

(eds) Readings in Language Studies, Volume 1: Language Across Disciplinary Boundaries. St. Louis, MO: International Society for Language Studies, 579–96.

23.

Faez

(2010) Linguistic and cultural adaptation of internationally educated teacher candidates. Canadian Journal of Educational Administration and Policy Special Issue(100): 1–20.

24.

Faez

(2011a) Developing the knowledge base of ESL and FSL teachers for K-12 programs in Canada: points of departure. Canadian Journal of Applied Linguistics 14(1): 48–82.

25.

Faez

(2011b) Are you a native speaker of English? Moving beyond a simplistic dichotomy. Critical Inquiry in Language Studies 8(4): 378–99. doi.org/10.1080/15427587.2011.615708

26.

Faez

(2011c) Reconceptualizing the native/nonnative speaker dichotomy. Journal of Language, Identity and Education 10(4): 231–49.

27.

Faez

Taylor

(2017) Language planning. In: Liontas

(ed.) The TESOL Encyclopedia of English Language Teaching. In: Abrar-ul-Hassan

(ed.) Teaching English as an International Language. Alexandria, VA: TESOL International Association.

28.

Ghasemboland

Hashim

(2013) Teachers’ self-efficacy beliefs and their English language proficiency: a study of non-native EFL teachers in selected language centers. Procedia – Social and Behavioral Sciences 103: 890–99.

29.

Gibson

Dembo

(1984) Teacher efficacy: a construct validation. Journal of Educational Psychology 76(4): 569–82.

30.

Hodgson

(2014) Mismatch: globalization and native speaker models of linguistic competence. RELC Journal 45(2): 113–34.

31.

Hoy

Woolfolk Hoy

(1993) Teachers’ sense of efficacy and the organizational health of schools. The Elementary School Journal 93(4): 356–72.

32.

Klassen

Tze

VMC

Betts

Gordon

(2011) Teacher efficacy research 1998–2009: signs of progress or unfulfilled promise? Educational Psychological Review 23(1): 21–43.

33.

LeBlanc

Painchaud

(1985) Self-assessment as a second language placement instrument. TESOL Quarterly 19(4): 673–87.

34.

Lee

(2009) Teachers’ sense of efficacy in teaching English, perceived English language proficiency, and attitudes toward the English language: a case study of Korean public elementary school teachers. Unpublished doctoral dissertation, Ohio State University, Columbus, Ohio, USA.

35.

Liaw

(2004) ‘How are they different?’ A comparative study of native and non-native foreign language teaching assistants regarding selected characteristics: teacher efficacy, approach to language teaching/learning, teaching strategies and perception of nativeship. Unpublished doctoral dissertation, Ohio State University, Columbus, Ohio, USA.

36.

Norris

Ortega

(2006) The value and practice of research synthesis for language learning and teaching. In: Norris

Ortega

(eds) Synthesizing Research on Language Learning. Amsterdam, NE: John Benjamins Publishing Company, 3–50.

37.

Ortega

(2015) Research synthesis. In: Paltridge

Phakiti

(eds) Research Methods in Applied Linguistics: A Practical Resource. London: Bloomsbury, 225–44.

38.

Pajares

(1996) Self-efficacy beliefs in academic settings. Review of Educational Research 66(4): 543–78.

39.

Pasternak

Bailey

(2004) Preparing nonnative and native English-speaking teachers: issues of professionalism and proficiency. In: Kamhi-Stein

(ed.) Learning and Teaching from Experience: Perspectives on Nonnative English-speaking Professionals. Ann Arbor, MI: The University of Michigan Press, 155–75.

40.

Plonsky

(2016). Applying meta-analysis in instructed L2 research. Presentation at Second Language Research Forum (SLRF), Columbia University, New York, USA.

41.

Praver

(2014) Japanese university English language teachers’ self-efficacy beliefs: a mixed methods approach. Unpublished doctoral dissertation, Temple University, Philadelphia, PA, USA.

42.

Rezvani

Amiri

(2013) English language teachers’ efficacy beliefs: validation of the instrument. International Journal of Applied Linguistics & English Literature 2(4): 14–26.

43.

Ross

(1998) Self-assessment in second language testing: a meta-analysis and analysis of experiential factors. Language Testing 15(1): 1–20.

44.

Sabokrouh

(2014) The effect of EFL teachers’ attitude toward English language and English language proficiency on their sense of efficacy. English Language Teaching 7(1): 66–74.

45.

Salvatori

(2009) A Canadian perspective on language teacher education: challenges and opportunities. The Modern Language Journal 93(2): 287–91.

46.

Shim

(2001) The teacher efficacy beliefs of Korean teachers of English as a foreign language. Unpublished doctoral dissertation, Ohio State University, Columbus, Ohio, USA.

47.

Shim

(2006) Validation of dimensions of teacher efficacy beliefs specific to nonnative Korean teachers of English. Studies on English Language and Literature 32(2): 207–24.

48.

Swanson

(2010) Teacher efficacy and attrition: helping students at introductory levels of language instruction appears critical. Hispania 93(2): 305–21.

49.

Swanson

(2012) Second/foreign language teacher efficacy and its relationship to professional attrition. The Canadian Modern Language Review 68(1): 78–101.

50.

Swanson

(2014) The power of belief: Spanish teachers’ sense of efficacy and student performance on the National Spanish Examinations. Hispania 97(1): 5–20.

51.

Tschannen-Moran

Woolfolk Hoy

(2001) Teacher efficacy: capturing an elusive construct. Teaching and Teacher Education 17: 783–805.

52.

Tschannen-Moran

Woolfolk Hoy

Hoy

(1998) Teacher efficacy: its meaning and measure. Review of Educational Research 68(2): 202–48.

53.

Turnbull

(2001) There is a role for the L1 in second and foreign language teaching, but…. Canadian Modern Language Review 57(4): 531–40.

54.

Wyatt

(2014) Towards a re-conceptualization of teachers’ self-efficacy beliefs: tackling enduring problems with the quantitative research and moving on. International Journal of Research & Method in Education 37(2): 166–89.

55.

Yilmaz

(2011) Teachers’ perceptions of self-efficacy, English proficiency, and instructional strategies. Social Behavior and Personality 39(1): 91–100.