English teacher identity measure: Development and validation in a Chinese EFL context

Abstract

With the emerging trend of researching teacher identity from using qualitative approach to adopting quantitative or mixed-methods approach, it is essential and significant to develop instruments for measuring this construct. This paper reports on the development and validation of a measurement tool, the English Teacher Identity Measure (ETIM). ETIM is designed to measure in-service English teachers’ identity levels. In accordance with DeVellis (2017) scale development guidelines, ETIM was developed following the procedure: (a) defining the construct and identifying indicators; (b) creating item pool; (c) seeking expert advice and validating the translation. The psychometric properties of the ETIM were explored on two separate participant groups. First, the initial 24-item ETIM was piloted online with 103 Chinese university EFL teachers for factor structure. The exploratory factor analysis (EFA) resulted in a 19-item scale with four factors: (a) future perspective, (b) career perception, (c) teacher belief, and (d) self-efficacy. Secondly, the 19-item ETIM was used to collect data from 331 university EFL teachers in a provincial capital city. The confirmatory composite analysis (CCA) using Smart PLS3.3.9 software confirmed the 19-item ETIM. The applications of the ETIM are discussed, and recommendations for future research are made in this paper.

Keywords:

• measurement tool
• teacher identity
• in-service English teacher
• scale development
• scale validation
• psychometric property

PUBLIC INTEREST STATEMENT

Teacher identity has gathered worldwide research interest over the past 30 years. However, to date, the definitions and indicators of this construct remain inconclusive. Despite the efforts to develop and validate instruments for measuring teacher identity, the attention to English language teachers is scarce. Based on the DSMRI framework and the review of prior definitions, teacher identity is defined as the extent to which a teacher identifies with the teaching profession as well as a teacher’s self-perception as an English teacher at present and in the future. An initial 24-item pool was developed. After exploratory factor analysis and confirmatory composite analysis, a 19-item scale was confirmed and validated. This study contributes to the existing literature by designing and validating a valid English Teacher Identity Measure (ETIM). Further application of this instrument into different settings with different samples is encouraged.

1. Introduction

Teacher identity (TI) has been recognized as a distinct and important research field in teacher education (Lentillon-Kaestner et al., 2018; Nguyen, 2016). Within the TI domain, the qualitative approach appears to be dominant (Hahl & Mikulec, 2018; Huang et al., 2019; Izadinia, 2013; Lutovac & Kaasila, 2018; Wang, 2021; Widodo et al., 2020; Xu et al., 2023; Xun & Zheng, 2015; Yang & Huang, 2022; Yuan, 2019; Zen et al., 2022). This line of inquiry focuses mainly on the formation and development process of TI. However, such studies are limited to small number of participants. In fact, the recent years has witnessed an emerging trend of researching teacher identity using non-qualitative approach (Hanna et al., 2022; Hong & Cross Francis, 2020; Li & Shao, 2018; Lin et al., 2022; Lu et al., 2022; Shi & Cheng, 2020; Tang, 2014; Wang, 2021; Zarrinabadi et al., 2023), in order to facilitate studies using quantitative or mixed-methods approach, it is imperative and significant to quantify teacher identity.

1.1. Prior definitions of teacher identity

Despite the consensus that TI is agentive, fluid, multifaceted, relation-embedded, and contextual (Beijaard et al., 2004; Huang & Wang, 2021; Moradkhani & Ebadijalal, 2021; S⊘reide, 2006; Solari & Martín Ortega, 2022; Swearingen, 2019), the definitions and components of TI remain inconclusive.

A wide variety of definitions have been documented in prior literature. For instance, Beijaard (1995) defined it as the various meanings attached to the teachers by themselves or by others. Beck and Kosnik (2014) theorized teacher identity as how teachers perceive themselves and their roles. It has also been referred to as the teacher’s positive evaluation of the profession (Wei et al., 2013), or the theories, attitudes and beliefs that teachers have about themselves which are developed throughout their teaching careers as they interact with others (Carlyon, 2016; Duchesne & McMaugh, 2016). The diverse definitions of TI are associated with the particular theoretical stand the researcher takes. For instance, Karaolis and Philippou (2019) adopted the psychological lens and defined teacher identity as the teachers’ perceptions of themselves at the present time. Enyedy et al. (2006) defined teacher identity by combining both psychological construct such as belief and sociological construct of practice.

1.2. Previous TI measurements and indicators

When developing instruments, previous researchers have mainly identified TI indicators inductively from literature review. For instance, Karaolis and Philippou (2019) have identified seven factors from past literature: (1) self-esteem; (2) self-efficacy; (3) professional commitment; (4) job satisfaction; (5) work motivation; (6) task orientation; (7) future perspective. In a similar vein, by building on Lamote and Engels (2010) TI questionnaire, Wong and Liu (2022) drew on three of the four indicators to represent student teacher identity. Based on a review of TI instrument studies between 2000 and 2018, Hanna et al. (2019) summarized six major TI indicators: (1) self-image; (2) motivation; (3) commitment; (4) self-efficacy; (5) task perception; (6) job satisfaction. These factors were adapted by Ni et al. (2021) to develop a TI instrument for computer science teachers.

Despite the fact that job satisfaction has been identified as a TI indicator in several measurement instruments (Hanna et al., 2019; Su, 2016), this study argues that job satisfaction should be the outcome rather than the indicator of TI.

1.3. Research gap

Quite a number of measuring instruments of TI have been developed, some instruments are designed for teachers in general (Kao & Lin, 2015; Wei et al., 2013), while others are designed for teachers of a certain subject, such as computer science (Ni et al., 2021), football (Geng et al., 2017), physical education (Liu & Keating, 2022; Zhou et al., 2012), and mathematics (Karaolis & Philippou, 2019; Willis et al., 2021). However, most studies focused on student teachers, in-service teachers received less attention.

An English Teacher Identity Measure (ETIM) was needed for the following reasons: firstly, to the best of the researcher’s knowledge, only three instruments are designed specifically for English teachers: Mahmoodarabi et al. (2021) English Language Teacher Professional Identity Scale; Xun et al. (2014) Teacher Identity Inventory for EFL teachers; Tang (2013) College English Teachers’ Professional Identity Inventory. The instruments developed by non-Chinese researchers may not suit the Chinese context due to cultural differences. Secondly, Xun et al. (2014) ’s scale was for senior high school teachers. Although Tang (2013)’s inventory was for university English teachers, the measure cannot fully reflect the conceptualization of English Teacher Identity in this study.

2. Purpose of this study

Motivated by the above research gap, the main objective of this study is to develop and validate a measure of teacher identity specific to in-service English teachers. One of the vantage points of this measure is that it is based on a widely accepted TI framework.

2.1. Theoretical framework

In developing a new TI scale, the Dynamics Systems Model of Role Identity (DSMRI) was utilized to frame the TI indicators (Figure 1). The DSMRI Model is selected as the framework for the following reasons: firstly, it is a metatheoretical framework that answers the key identity questions, including who am I? What do I do? Why do I do? Secondly, it incorporates multiple theoretical perspectives (socio-psychological, sociocultural, socio-cognitive theories) on identity (Garner et al., 2016), thus providing a holistic explanation of identity. It also acknowledges the influence of social context and subject domain on teacher identity (Garner & Kaplan, 2018). Thirdly, it has been applied by Liu and Keating (2022) to inform the pre-service physical education teachers’ TI indicators.

Figure 1. DSMRI framework for measuring in-service English teachers’ identity.

2.2. Indicators of English teacher identity

This study takes Kaplan and Garner’s (2017) DSMRI model as its point of departure, the model depicts four components of the teacher identity: (1) Self-perceptions and self-definitions (self-attributes that someone perceives to be relevant to their functioning in the role of a teacher); (2) Ontological and epistemological belief (beliefs and perceptions regarding the nature of teaching, learning, and knowledge); (3) Purpose and goals (the purpose of actions and goals of education or a subject area); (4) Perceived action possibilities (intention and actual behaviors related to goals and purposes) (Kaplan & Garner, 2017).

Some components and their specified meaning of the framework are not directly applicable to our research context; therefore, the elements of these four indicators are adapted (see the key adaptations in red color). First, considering the various English teaching reforms that are taking place in China, beliefs about reform are added. Second, English teacher’s identity is operationalized in a way that is specific to English teachers, which is not embodied in the framework.

3. Method

This study involves two distinct stages: scale development and scale validation.

3.1. Study procedure

Following DeVellis’s (2017) scale development guidelines, the first stage begins with literature search, which aims at defining the English teacher identity construct and identifying its indicators. After creating an item pool, a panel of experts were invited to discuss the relevance of items to the construct, length and redundancy of the initial item pool. Each phase is elaborated below. During the second stage (scale validation), the instrument’s psychometric properties were assessed using exploratory factor analysis (EFA), analysis of internal consistency, and confirmatory factor analysis. The procedure is illustrated in Figure 2. The psychometric properties of the ETIM were explored on two separate participant groups.

Figure 2. The study procedure.

3.1.1. Defining the construct and identifying indicators

Looking across the existing definitions of TI, some researchers focused on evaluation of the teaching profession only, while others only focused on the teachers’ self-perception. This study argues that both facets are indispensable for understanding teacher identity. This study defined TI as the extent to which a teacher identifies the teaching profession as well as the teachers’ perceptions of themselves as an English teacher at present and in the future. Firstly, the evaluation of the profession not only demonstrates the teacher’s personal values, but also reflects the motivation to become a teacher, which has been viewed as an indicator of teacher identity in some studies (Berger & Lê Van, 2018; Day, 2002). A person is unlikely to remain an English teacher if he does not identify with the teaching profession. On the other hand, the teacher’s perception of himself or herself is the core of TI, as has been concurred by Karaolis and Philippou (2019). The teacher’s perception of himself at present and future involves their perception of knowledge, skills, teaching belief and how they envision themselves in the future. In-service teachers with higher levels of identity are expected to hold firm belief, have clear personal goals and possess the sufficient knowledge and skills to perform everyday work.

As mentioned earlier, four indicators have been chosen for measuring the construct “English teacher identity”: teacher belief, self-efficacy as an English teacher, career perception, future perspectives. The four indicators were elaborated as follows:

3.1.1.1. Teacher belief

The importance of teacher belief is seen in previous literature; it has been highlighted as the building blocks of teacher identity (Beijaard & Meijer, 2017). Yazan (2018) noted that whether and how language teachers understand this nature of language is integral to their identities as language teachers, because this understanding guides what they see as important in language instruction and how they facilitate their students’ language development. In accordance with this statement, teacher belief is defined as the teacher’s belief about how English should be taught and learnt, and belief about how teachers react to the educational policy changes. Considering that in-service teachers in China are faced with English teaching reforms, their beliefs about changes should be included.

3.1.1.2. Future perspective

TI is not only the answer to who I am at the moment, but more importantly to what kind of person I want to become (Karaolis & Philippou, 2019; Kelchtermans, 2009). Future perspective in this study is linked with a teacher’s goal-setting, it reflects a teacher’s vision of himself and herself in the future, i.e. the teacher’s perception of himself as a teacher and researcher in the future. This indicator influences the teacher’s present actions, because the better future perspective a teacher holds, the more efforts he takes to achieve the aim, the higher the teacher identity level.

3.1.1.3. Career perception

Career perception reflects the teacher’s attitude towards the teaching profession. A positive perception of the career shows a higher sense of teacher professional identity.

3.1.1.4. Self-efficacy

Self-efficacy has been widely acknowledged as an indicator of TI (Canrinus et al., 2011; Hanna et al., 2019; Kelchtermans, 2009; Su, 2016; Wong & Liu, 2022). A strong sense of TI is strongly linked to self-efficacy. Menon (2020) found that sources of self-efficacy contributed to four dimensions of identity: institutional, nature, affinity and discourse identity. Self-efficacy is conceptualized in this study as in-service teacher’s ability in language proficiency, instructional strategies, classroom management, and student engagement.

3.1.2. Creating item pool

The items were either self-designed or adapted from various available instruments, including Wei et al. (2013) Primary, Middle and High School Teacher Professional Identity Scale, Tang (2013) College English Teachers’ Professional Identity Inventory, Xun et al. (2014) Teacher Identity Inventory for EFL teachers, Choi and Lee (2016) The Self-efficacy Survey, and Tschannen-Moran and Hoy (2001) The Ohio State Teacher Efficacy Scale.

For the indicator teacher belief, items 1–3 were concerned with teacher’s belief about English teaching reform, items 4–5 were about teacher’s belief about learning method and the emphasis that students should take a dominant role in learning. For career perception, two of the items were extracted from Wei et al. (2013) Primary, Middle and High School Teacher Professional Identity Scale, and two items were self-designed. The future perspective items are self-designed. Example items of the indicator future perspective include “I expect to become a teacher with strong academic research ability.” Efficacy items for language proficiency (SE1 and SE2) are added into the self-efficacy indicator, because the language teacher’s identity is associated with his or her language background and language proficiency (Pennington & Richards, 2016). Furthermore, China, as a foreign language context, provides few opportunities for language learners to use English for daily social interactions. English education in China mostly relies on classroom instruction. Therefore, the teachers’ efficacy for engaging students in English learning in EFL context is significant. Other self-efficacy items were adapted from two existing self-efficacy scales by Choi and Lee (2016) and Tschannen-Moran and Hoy (2001). SE3–5 were self-efficacy items for instructional strategies, SE 6–8 were self-efficacy items for classroom management, SE9–10 were self-efficacy items for student engagement.

Based on the identified four factors, a 30-item pool was created. The response format for these items utilized a 5-point Likert scale ranging from (1) strongly disagree to (5) strongly agree.

3.1.3 Seeking expert advice and validating the translation

In order to examine the clarity and content validity of each item, the initial pool of items was reviewed by three professionals holding PhD in education. After consulting experts, four items were removed for redundancy and two were deleted for irrelevance. Hence, 24 items were retained (Table 1). The questionnaire was then translated into Chinese. To ensure the quality of translation, two experts in translation were invited to back-translate and proofread the items.

Table 1. Preliminary item pool

	Item
FP1	I hope to excel in teaching English-related courses.
FP2	I expect to improve myself professionally.
FP3	I expect to be promoted in my university.
FP4	I expect to become a university English teacher with strong academic capability.
FP5	I strive to improve my teaching by carrying out academic research.
CP1	Teaching is a well-respected profession.
CP2	Teaching is very important for the development of students.
CP3	Teachers’ work is important for the development of human society.
CP4	The teaching profession is one of the most important professions in social division of labor.
TB1	Teachers should be equipped with the basic information technology skills for English teaching.
TB2	Teachers should continually update their professional knowledge to keep up with English reform.
TB3	Teachers should be attentive to teaching-related training opportunities to improve their teaching.
TB4	Teaching methods are very important to students’ English language learning.
TB5	Students should take dominant role in learning, so I attach great importance to training students’ learning methods.
SE1	I am fluent in spoken English language.
SE2	I am proficient in written English language.
SE3	I can communicate to students the specific learning objectives and outcomes of the lesson.
SE4	I can craft a variety of good questions throughout the lesson.
SE5	I can respond to difficult questions from my students.
SE6	I can establish a classroom management system with each group of students.
SE7	I can effectively use allocated time for various teaching activities.
SE8	I can maintain a classroom environment in which students work cooperatively.
SE9	I can motivate students who show low interest in learning English.
SE10	I can maintain high levels of student engagement in learning tasks.

3.2. Data collection and analysis

Approval from the Institutional Review Board of the first author’s affiliated university was obtained before collecting data. All participants were briefed on the aim of the study and were informed that they could withdraw from this research at any time if they wished to.

Exploratory factor analysis (EFA) was performed on the first participant group. The sampling adequacy of factor analysis was determined by O’Rourke and Hatcher (2013) suggestion of having “five times the number of variables being analyzed” (p. 73), therefore, the minimum sample size for factor analysis in this study is 100. The Chinese version of the Initial 24-item ETIM was posted on the Wenjuanxing platform (a popular platform in China for creating and posting questionnaires). The participants were recruited by convenience sampling. The link of the survey was sent via Wechat (a popular social media platform) or email to the 130 participants from different Higher Education Institutions (HEIs) across five provinces in China, including Anhui, Hubei, Hunan, Guangdong, Shanxi Provinces. The selection of various institutions and different regions was done purposively to ensure diversity in responses. 103 questionnaires were returned during a span of three months. These 103 participants were not approached for the subsequent data collection for scale validation.

Confirmatory composite analysis (CCA) was conducted on the second participant group recruited by snowball sampling. The link of the online survey (see Appendix A for the Chinese version of the 19-item ETIM) was posted in the Wechat groups of university EFL teachers with the aid of the deans of the English Department in a provincial capital city in China, 331 questionnaires were returned during a span of four months.

The data collected was analyzed using SPSS 27 and Smart PLS 3.3.9. To reveal the underlying structure among the items in the initial ETIM, EFA using principal component analysis with varimax rotation was conducted, followed by the reliability analysis using SPSS 27. After confirming the factor structure of the ETIM, the factor structure obtained from EFA was tested with CCA to test whether the structure was consistent with the data. More evidence for the validity of the ETIM (e.g. convergent, discriminant validity) was examined.

3.3. Participants

The participants of this study comprised a total of 434 university EFL teachers. The first participant group consisted of 103 EFL teachers (Table 2). Female (n = 76) teachers outnumbered male (n = 27) teachers. The participants born in the 1960s to 1990s were all involved. Over half of them were born in the 80s (62.1%) and were lecturers (68%). Eighteen of them were teaching assistants, 70 were lecturers, 14 were associate professors and one of them was professor. Sixty-nine teachers were serving in public HEIs and 34 in private HEIs.

Table 2. Participant profile for EFA (N = 103)

Variables		Frequency N	Percentage %
Gender	Male	27	26.2
	female	76	73.8
Year of birth	90s	25	24.3
	80s	64	62.1
	70s	10	9.7
	60s	4	3.9
Academic title	Teaching assistant	18	17.5
	Lecturer	70	68
	Associate professor	14	13.6
	Professor	1	1
Type of institution	Public HEI	69	67
	Private HEI	34	33

CCA was conducted on the second participant group that comprised 331 EFL teachers (55 male and 276 female teachers) from a provincial city in central China (Table 3). There were more teachers from private (180) than public (151) universities. Over half of the participants (57.7%) were born in the 1980s. Lecturers (45.9%) and associate professors (37.2%) make up the majority of the population. Forty of them were teaching assistants, 152 were lecturers, 123 were associate professors and 16 of them were professors. Approximately 45% of them (N = 151) were serving in public HEIs and 55% (N = 180) in private HEIs.

Table 3. Participant profile for CCA (N = 331)

Variables		Frequency N	Percentage %
Gender	Male	55	16.6
	Female	276	83.4
Year of birth	90s	55	16.6
	80s	191	57.7
	70s	56	16.9
	60s	23	6.9
	50s	6	1.8
Academic title	Teaching assistant	40	12.1
	Lecturer	152	45.9
	Associate professor	123	37.2
	Professor	16	4.8
Type of institution	Public HEI	151	45.6
	Private HEI	180	54.4

4. Results

This section presents the results from EFA. The convergent, discriminant validity of the ETIM are also explored.

4.1. Data examination

There were no missing values in the data, so a total of 103 complete cases were used for EFA. The normality of the data was examined through inspection of skewness and kurtosis (Table 4). From the findings, the skewness (−.905) and kurtosis (2.913) values are within the recommended cutoffs of |3| and |10| for skewness and kurtosis, respectively (Kline, 2005). This ensures the normality of the data.

Table 4. Descriptive statistics of the ETIM

			Statistic	Std.Error
TI	Mean		4.0767	.03109
	95% Confidence Interval for Mean	Lower bound	4.0155
		Upper bound	4.1379
	5% Trimmed Mean		4.1051
	Median		4.0278
	Variance		.320
	Std. Deviation		.56567
	Minimum		1.25
	Maximum		5.00
	Range		3.75
	Interquartile Range		.72
	Skewness		−.905	.134
	Kurtosis		2.913	.267

4.2. Exploratory factor analysis

According to Ong and Puteh (2017), the SPSS statistical package is most appropriate when the research aim is to refine and validate the variable items. The EFA involves five stages: checking whether data is suitable for factor analysis, selecting factor extraction method, determining number of factors and factor rotation method, interpretation and labeling (Williams et al., 2010). Comrey and Lee (2009) offered guidelines on judging the factor loading value: Loadings equal or higher than 0.32 were considered poor, loadings equal or higher than 0.45 reasonable, loadings equal or higher than 0.55 as good, loading equal or higher than 0.63 as very good and loadings equal or higher than 0.71 were considered to be excellent. Therefore, in this study, the factor loading cutoff value of 0.7 was adopted. The final factor structures were finalized by looking into the following criteria: (i) KMO value above 0.7, (ii) Kaiser criterion and scree plot, (iii) total variance explained above 60%, and (iv) factor loading above 0.7.

A principal component analysis (PCA) was conducted on the 24 items with orthogonal rotation (varimax). As seen in Table 5, the Kaiser—Meyer–Olkin measure for the sampling adequacy was .928, Bartlett’s test of sphericity (df = 276, p < .001) indicating that the data was appropriate for conducting exploratory factor analysis.

Table 5. The results of KMO and Bartlett’s test of sphericity

KMO Measure of Sampling Adequacy		.928
Bartlett’s Test of Sphericity	Approx. Chi-Square	2144.370
	df	276
	Sig	<.001

An initial analysis was run to obtain eigenvalues for each factor in the data (Table 6). Four factors had eigenvalues over Kaiser’s criterion of 1 and in combination explained 73.399% of the variance. The rotated component matrix was analyzed. Within the matrix, items with factor loadings above 0.7 were retained, after removing the items less than 0.7 one by one, for example, item “Teaching is a well-respected profession” (the sample item of career perception) which had a loading of .564 was deleted. Each time an item was removed, the exploratory factor analysis was run again, the process ended when the rotated component matrix showed a cleanly distributed pattern.

Table 6. Eigenvalue and total variance explained

Component	Extraction Sums of Squared Loadings			Rotation Sums of Squared Loadings
	Total	% of variance	Cumulative %	Total	% of variance	Cumulative %
1	10.485	55.183	55.183	6.116	32.188	32.188
2	1.886	9.929	65.112	3.532	18.592	50.780
3	1.234	6.495	71.607	2.565	13.498	64.278
4	1.027	5.405	77.012	2.419	12.734	77.012

On the basis of EFA, we confirmed a scale comprising 19 items that explained 77.012% of the total variance. The scree plot (Figure 3) showed inflexions that would justify retaining four factors. These factors were retained because of the convergence of the scree plot and Kaiser’s criterion, the literature also justified the 4-factor solution. The items that cluster on the same factor suggest that factor 1 represents self-efficacy (9 items), factor 2 represents teacher belief (4 items), factor 3 career perception (3 items) and factor 4 future perspective (3 items). After rotation, the four factors explained 32.188%, 18.592%, 13.498%, and 12.734% of total variance respectively (refer to Table 6). The findings of EFA representing the respective factor loadings of 19 items along with their respective dimensions are provided in Table 7.

Figure 3. Scree plot.

Table 7. Factor loadings

ETIM items	Factor loading
	1	2	3	4
Factor 1: Self-efficacy
I can effectively use allocated time for various teaching activities.	.821
I can establish a classroom management system with each group of students.	.821
I can respond to difficult questions from my students.	.813
I can communicate to students the specific learning objectives and outcomes of the lesson.	.772
I am proficient in written English language.	.767
I can craft a variety of good questions throughout the lesson.	.759
I can maintain a classroom environment in which students work cooperatively.	.753
I can motivate students who show low interest in learning English.	.739
I am fluent in spoken English language.	.716
Factor 2: Teacher belief
English teachers should continually update their professional knowledge to keep up with College English reform.		.817
English teachers should be attentive to various teaching-related training opportunities to improve their teaching.		.802
English teachers should be equipped with the basic information technology skills for English teaching.		.765
English teachers should conduct their courses effectively.		.730
Factor 3: Career Perception
The teaching profession is one of the most important professions in social division of labor.			.831
Teachers’ work is important for the development of human society.			.806
Teachers’ work is very important in the growth and development of students.			.762
Factor 4: Future perspective
I expect to become a teacher with strong academic capability.				.772
I expect to be promoted in my university.				.756
I strive to improve my teaching by carrying out academic research.				.709

4.3. Reliability analysis

Cronbach Alpha coefficient was used to measure internal consistency; the most agreed minimum coefficient value was 0.7 (Taherdoost, 2016). The reliability test was conducted on the four indicators of TI. Cronbach’s alpha values were acceptable for each dimension from 0.798 to 0.953 (Table 8).

Table 8. Reliability test result

Factor	Cronbach alpha value
Self-efficacy	.953
Teacher belief	.908
Career perception	.899
Future perspective	.798

4.4. Confirmatory composite analysis

To test how well the four-factor model fit, a CCA was conducted. The measurement model assessment step in Partial Least Squares-Structural Equation Modeling (PLS-SEM) was referred to as CCA, which is a systematic methodological process for confirming measurement theory (Hair et al., 2020).

4.4.1. Model fit

According to Dash and Paul (2021), the model fit can be assessed by Standardized root mean square residual (SRMR). The acceptable value range of SRMR is between 0 and 1. The lower the value is, the better the model fit. CCA resulted in satisfactory indices (SRMR = 0.139) indicating that the four-factor model, obtained in EFA, was best of the fit.

4.4.2. Reliability analysis

As indicated in Figure 4, all the item loadings of the four factors of TI are above 0.708, indicating that item reliability was acceptable, internal consistency reliability was assessed by examining composite reliability, values between 0.7 and 0.9 are good, values above .95 are problematic (Hair et al., 2019). As seen in Table 9, the reliability is good.

Figure 4. Analysis of the four indicators of TI.

Table 9. Cronbach alpha and composite reliability

Indicators	Number of items	Cronbach alpha	Composite reliability
Teacher belief	4	0.918	0.944
Self-efficacy	9	0.928	0.940
Future perspective	3	0.874	0.922
Career perception	3	0.9	0.938

4.4.3. Convergent and discriminant validity

To confirm convergent validity, the average variance extracted (AVE) is assessed, the value at o.5 or higher is acceptable. As seen in Table 10, AVE for each of the four dimensions is higher than 0.5.

Table 10. Convergent validity

Indicators	Number of Items	Average variance extracted
Teacher belief	4	0.804
Self-efficacy	9	0.636
Future perspective	3	0.798
Career perception	3	0.835

The heterotrait-monotrait ratio of correlation (HTMT) was proposed as a new approach to assess discriminant validity in PLS-SEM (Henseler et al., 2015). As seen from Table 11, the HTMT value is less than the threshold value of 0.9, indicating that the discriminant validity is present.

Table 11. Discriminant validity

	Career perception	Future perspective	Self-efficacy	Teacher belief
Career perception	0.896
Future perspective	0.315	0.893
Self-efficacy	0.613	0.580	0.798
Teacher-belief	0.836	0.415	0.673	0.897

5. Discussion and conclusion

The topic of TI has gained considerable attention from across the world. Unfortunately, till date, the definitions and indicators of TI have not been agreed upon. Recent years have seen increasing scholarly interest in measuring this construct.

This study was undertaken to develop and validate a tool for measuring in-service English teachers’ identity. Results from the EFA and CCA show that the measurement tool was valid. The EFA procedure resulted in the removal of five items, these items were eliminated for not reaching the threshold value of 0.7 based on the data obtained from the pilot study. The CCA then confirmed that the 19-item ETIM meets the threshold value for reliability, convergent validity, and discriminant validity. Hence, the ETIM was proved to be psychometrically sound.

This study extends extant research on TI in several ways. First, this new instrument to measure TI augments the understanding of TI. Second, teacher belief and future perspective, as two of the relatively new indicators of TI have been validated by this study. We argue that teacher belief and future perspective are important aspects of teacher identity. Third, in practice, future researchers can use this instrument to measure English teacher identity. Since this instrument was developed in EFL setting in China, this study recommends applying ETIM in different research settings such as ESL settings. Application of this instrument in other EFL settings in countries other than China is also encouraged. This instrument was tested among the sample of university EFL teachers in China, it is suggested that the instrument is used for teachers teaching at different levels, such as preschool, primary, secondary school, and vocational college teachers.

6. Limitations and recommendations

Several limitations should be addressed. Firstly, sourced data from university EFL teachers in a provincial city in China context, the sample is unique because most participants were female teachers. Future research could include a replication of the study with the inclusion of different contexts and the solicitation of teachers with different socio-demographic factors. Secondly, the sampling technique is convenience sampling and snowball sampling. Therefore, the results should be interpreted with caution as their generalizability may be limited. Third, ETIM only captured a few facets of the construct “teacher identity” and may not be applicable to teachers teaching other subjects. Future studies could delve into other disciplines and take into consideration the characteristics exclusive to the specific discipline.

It has been observed that quite a few studies selected existing yet recurrent indicators to develop TI measurement tools, which may be a safer option. However, since researchers advocate using different theories or theoretical framework to shed light on the concept, and teacher identity can be operationalized varyingly based on differing theories or theoretical frameworks, and it is encouraged that future researchers validate “new” TI indicators.

Acknowledgments

We wish to acknowledge all the 434 Chinese university EFL teachers who joined this research, special thanks to those deans and friends who helped the first author warmly in collecting data. Our heartfelt thanks also go to the journal administrator, editor, and two anonymous reviewers for their hard work.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Yiqian Yan

Yiqian Yan is an EFL teacher in China. She obtained her PhD in Education from SEGi University, Malaysia in 2023. Her research interest is teacher development.

Madhubala Bava Harji

Madhubala Bava Harji is the acting dean of Faculty of Education, Languages, Psychology & Music, SEGi University, Malaysia. Her research interests include assessment, mobile learning, innovative teaching and learning approaches, literacy development, and academic integrity.

Sathiamoorthy Kannan

Sathiamoorthy Kannan is a senior lecturer at SEGi University, Malaysia. His specialized areas include educational leadership, curriculum leadership, change leadership, technology leadership, design thinking, quantitative research methods, assessment and evaluation, management of innovation, and quantitative data analysis using SPSS and Smart PLS.

Appendix A.

The Chinese version of the 19-item ETIM

Table

维度 1: 自我效能感
1	我能为各种教学活动有效分配时间。
2	我能制定班级制度，管理好各组学生
3	学生提出比较难的问题时，我也能应对。
4	我能告诉学生每节课的具体学习目标和成果。
5	我能熟练地用英语书写。
6	我能针对每节课设计出好的问题。
7	我能创造班级环境，使学生之间互相配合。
8	我能够激励英语学习兴趣不高的学生。
9	我的英语口语流利。
维度 2: 教师信念
10	大学英语教师应不断更新自己的专业知识。
11	大学英语教师应利用各种培训机会提高教学水平。
12	大学英语教师应具备一定的信息技术应用能力，以满足新形势对英语教学的需求。
13	教师应当确保教学效果是有效的。
维度 3: 职业感知
14	教师职业是社会分工中最重要的职业之一。
15	教师的工作对人类社会发展有重要作用。
16	教师的工作对促进学生的成长与发展很重要。
维度 4: 未来展望
17	我努力成为科研能力强的大学英语教师。
18	我努力在单位获得晋升的机会。
19	我尽力通过做科研提升自身的教学水平。