Revisiting Integrative Motivation: Validation of International Posture among Malaysian Undergraduates

Gardner’s (1985) in his socio-educational model introduced the variable of ‘integrativeness’ as a construct that focuses on English as a Second Language (ESL) learners’ attitudes towards the native English community. Later, Yashima (2002) proposed ‘international posture’ as an affective construct that re-examined ‘integrativeness’. Research conducted to reach a common operationalisation of ‘international posture’ is very limited. Hence, our objective in this study is to determine the level of each measurement construct under the affective variable of International Posture as well as to validate the measurement constructs of this variable among Malaysian undergraduates. In this quantitative study, 540 undergraduate students from a public university were selected. These students responded to a questionnaire that contained the four measurement constructs. For analysing the data, we used Partial Least Square-Structural Equation Modelling (PLS-SEM). Our results have proven that the International Posture among Malaysian undergraduates is high, and the measurement constructs of this variable are valid and can be operationalised in the Malaysian English language learning context. The * Corresponding author, email: vahid@upm.edu.my Citation in APA style: Razak, F. N. A., Nimehchisalem, V., Tan, H., Lee, G. I., & Abdullah, A. N. (2022). Revisiting integrative motivation: Validation of International Posture among Malaysian undergraduates. Studies in English Language and Education, 9(1), 222-236. Received July 27, 2021; Revised October 6, 2021; Accepted November 30, 2021; Published Online January 17, 2022 https://doi.org/10.24815/siele.v9i1.21838 F. N. A. Razak, V. Nimehchisalem, H. Tan, G. I. Lee & A. N. Abdullah, Revisiting integrative motivation: Validation of international posture among Malaysian undergraduates | 223 scale validated in the study can contribute to future studies on this variable in the Malaysian English language learning context.


INTRODUCTION
International Posture is an affective variable studied by Yashima (2002) in the effort to contextualise MacIntyre et al.'s (1998) model into the context of Japanese English language learners concerning their motivation in foreign language (FL) learning. Among the Japanese foreign language learners, the English language represents the world beyond its native community, which is due to education and the exposure of media (Yashima, 2002). This view is highly related to the new status of the English language that has become the world's language as the non-native speakers of the language have outnumbered the native speakers (Bolton, 2002). Such a shift in viewing the English language necessitates revisiting the factors that internally motivate learners in different contexts to acquire the English language.
The readiness to participate in a target language conversation at a given time with a particular person(s) is called 'Willingness to Communicate' (WTC) (MacIntyre et al., 1998). In conceptualising WTC, MacIntyre et al. (1998) have incorporated the socio-educational model by Gardner (1985). As presented in Figure 1, Gardner's socio-educational model was presented in four components, that is (1) social milieu, individual differences, (3) language acquisition contexts, and (4) outcomes. Gardner's (1985) socio-educational model.

Figure 1.
In the component of individual differences, Gardner (1985) presented 'integrativeness' and 'attitude towards the learning situation' as the variables contributing to language learners' motivation. 'Integrativeness' specifically conceptualises language learners' desire to acquire another language to communicate with others in the native language community (Azar & Tanggaraju, 2020). However, for language learners today, the English language represents not only the native English community but also the English language users all over the world, including Asians and Africans (Lai, 2013;Rahman & Sahayu, 2020;Sa'adah et al., 2018;Sung, 2013). Hence, this leads to the re-conception of the fundamental theories related to motivational literature in the area of Second Language Acquisition (SLA). Yashima's (2002) affective variable, International Posture, has received the limelight in recent SLA studies. Originally, Yashima (2002) introduced International Posture to replace Gardner's (1985) notion of 'integrativeness' or the motivational desire of a language learner to communicate with individuals from another language community. This variable represents the broader context of English language users beyond the native community. It is divided into four dimensions, namely intercultural friendship orientation in learning english, interest in international vocation/activities, interest in foreign affairs, and intergroup approach avoidance tendency. According to Botes et al. (2020), International Posture is a valid variable for explaining motivation among English as a Foreign Language (EFL) learners, and it can also be a significant variable for explaining indirect relationships between EFL learners' motivation and language proficiency. Botes et al. (2020) also claimed that International Posture is under-researched and further studies are required for confirming the underlying design of this variable.
In relation to that, the current study acknowledges the status of English as the world language that it is not only owned and used by its native speakers. Hence, the language learners' motivation in learning the English language today is not restricted only to integration with its native speakers but to the international community as proposed by Yashima (2002). With this in mind, it is deemed necessary to revisit the construct of integrative motivation by Gardner (1985) by broadening the context of the community of English language speakers with the construct of International Posture by Yashima (2002).
Due to that, the current study aims to test the relevance of International Posture, as an affective variable among Malaysian undergraduates. Specifically, the current study was designed: (1) to identify the level of each measurement construct under the affective variable of International Posture among Malaysian undergraduates, and (2) to validate the measurement constructs of the affective variable of International Posture among Malaysian undergraduates. The results of the study contribute towards a better understanding of motivation in SLA settings and will elucidate how the underlying design of International Posture could be operationalised in the Malaysian English language learning context. On top of that, the current study also has significant theoretical contributions, as it introduces International Posture to the model of language learners' motivation in the current ESL context of Malaysia. It represents the shifted view on the English language that has brought changes to the factors that internally motivate language learners to learn the English language. In the context of the present study, English language speakers are not only from the native speakers' community, but the language is also used by non-native speakers to integrate with other native or non-native speakers from around the world.

LITERATURE REVIEW
The term International Posture is coined by Yashima (2002) to explain the insufficiency of positive feelings that can lead towards integrative motivation among Japanese English language learners due to minimal to no direct contact with the English language L2 community (Lockley, 2015;Yashima, 2002). Discussion on International Posture is made in response to the framework of L2 WTC by MacIntyre et al. (1998) to contextualise L2 WTC into the Japanese English language young learners. International Posture has been defined as having a non-ethnocentric attitude toward different cultures, an interest in foreign affairs, an interest in traveling to other countries to study or work, and readiness to communicate with intercultural partners. Based on this definition, it can be understood that both instrumental (functional reasons of achieving language mastery such as passing exam and career development, other than the integration goal) and integrative orientations are addressed in the notion of International Posture or also known as attitudes towards the international community. Yashima (2002) supports the claim made by Dörnyei (1990) that language learners' attitudes towards the English language community are largely created through education and media exposure. Thus, this favourable attitude towards the international English communities varies among individuals in which one learner might have a more positive attitude in comparison with others. Therefore, it is believed that proper measurement of this variable would allow a better understanding of an individual internal motivation in integrating with international speakers of English. Yashima (2002) operationalised International Posture by breaking it into the following domains: interest in foreign affairs, interest in international vocation/activities, intercultural friendship orientation in learning english, and intergroup approach avoidance tendency, as presented in Figure 2.  According to Yashima (2002), all the measurement constructs of International Posture in the study were significant at p <.01 ( Figure 1) with the weight (the β values of indicators yield on the variable) of the dimensions towards the construct varying from moderate to strong (IFA, .36; AAT, .77; IVA,.78; IFO,.85). However, according to Botes et al. (2020), the operationalisation of the variable International Posture has experienced a few changes through Yashima's studies in 2002Yashima's studies in , 2004Yashima's studies in and 2009. These changes are essentially made to address different types of language learners and their views for optimum language learning. Yashima et al. (2004) replicated Yashima's (2002) study by focusing on younger language learners. This study was conducted to include the element of L2 behaviour or frequency of communication into the previous model. Thus, they selected their sample among high schoolers who had a native English speaker as their EFL and homeroom teacher. Their other group of samples were high school students who were involved in a study-abroad programme in the United States of America. In this study, the variable of International Posture was measured through (1) interest in international vocation/activities, (2) approach-avoidance tendency, and (3) interest in foreign affairs. The measurement constructs for the variable International Posture used in the study by Yashima et al. (2004) are presented in Figure 3.  (Yashima et al., 2004) As shown in Figure 3, the weights of the dimensions towards the construct of International Posture are stated as .43 for Interest in Foreign Affairs, .64 for Approach Avoidance, and .75 for International Vocation. However, International Friendship Orientation was omitted from the measurement construct as the focus in the study shifted to the participants that had bigger opportunities in L2 communication. Hence, communication in the English language did not only focus on international friendship.
In 2009, in an effort to bring forward the concept of Ideal L2 Self, Yashima conducted another study. The concept, Ideal L2 Self, functions as an encouragement for the language learners' L2 related actions and conducts. Hence, a more meaningful language learning situation can be created. Therefore, the scale of International Posture was revised again in this study and a new sub-scale was added. Yashima (2009) named the new sub-scale as "Having things to communicate to the world" and proposed four items to measure this sub-scale. An example of one of the items is "I have issues to F. N. A. Razak, V. Nimehchisalem, H. Tan, G. I. Lee & A. N. Abdullah, Revisiting (2009) is presented in Figure 4 below. These three studies by Yashima and colleagues in 2002Yashima and colleagues in , 2004Yashima and colleagues in , and 2009 have operationalised International Posture in determining its relationship in enhancing language learners' WTC in English, specifically in Japan. Other than that, Structural Equation Modelling (SEM) was utilised as the method of data analysis in all these three studies by Yashima. In support of this construct and its vast potential in explaining language learners' WTC, the researchers believed that it is significant to test the validity of this variable in the Malaysian English language learning context as presented in sub-sections 4.2 and 4.3.

METHOD
The current study is a quantitative study utilising Structural Equation Modelling (SEM) as the statistical analysis method. The participants of the study were 540 Malaysian undergraduates from a public university. The researcher used proportional quota sampling to identify the participants in the study. A questionnaire with 5-scale answers ranging from totally agree to totally disagree was employed in the current study to measure the four sub-scales of International Posture. The four dimensions are (1) intercultural Friendship Orientation (IFO), (2) Interest in International Vocation/Activities (IIV), (3) Intergroup Approach-Avoidance Tendency (IAT), and (4) Interest in Foreign Affairs (IFA), adapted from Yashima (2002). These dimensions are referred to as IFO, IIV, IAT and IFA respectively in the current study. Each item is labelled as IN_POSTURE_1 until IN_POSTURE_21. The sub-scale of 'Having Things to Communicate to the World' from Yashima (2009) is not included in the study as the English language is used more frequently in Malaysia, where English has the status of L2, as compared to Japan.
Data were collected through the distribution of questionnaires to 600 undergraduates studying in a public university. The data analysis process started with transferring the raw data into SPSS (version 25). Then, the data were examined. This examination stage is crucial, especially when the study involves the use of SEM, as the data are assumed to be error-free after the descriptive statistical analysis. Subsequently, the researchers screened the data for suspicious response patterns and missing values through a complete case analysis. The outliers were discarded before examining the normality of data distribution. A final 540 samples were selected before the data were computed into scores. Finally, the data were transferred into SMART-PLS, a statistical software program enabling the Partial Least Square-Structure Equation Modelling (PLS-SEM) analysis. We chose PLS-SEM as compared to Covariance Based-Structural Equation Modelling (CB-SEM) as the model tested in the study because it has a combination of reflective (indicators are reflected by a latent variable) and formative measurement models (latent variable measured by the indicators). When the model is a combination of reflective and formative measurement models, PLS-SEM is likely to be the ideal choice (Memon et al., 2017).

RESULTS AND DISCUSSIONS
The results are presented in three sub-sections. The first section examines the level of each construct under the affective variable of International Posture. The second sub-section presents the validity and reliability of the indicators (items) through Confirmatory Factor Analysis (CFA) of reflective measurement model assessments, namely (1) internal consistency, (2) convergent validity and (3) discriminant validity. The third sub-section explores the CFA of the formative measurement model which includes three assessment steps, namely (1) assessing convergent validity, (2) addressing collinearity issues, and (3) assessing the relevance and significance of the formative indicators.

Level of each Construct under the Affective Variable of International Posture
International Posture was assessed under four constructs, that is (1) IFO, (2) IAT, (3) IIV (4) IFA as presented in Figure 5. This figure indicates that IFO has the highest mean (4.3), whereas the mean of IAT is 3.9. Meanwhile, the mean for both IIV and IFA is 3.8. Conclusively, all International Posture dimensions have high mean values, exceeding 3.5. This result proves that Malaysian undergraduates have a high interest in engaging with the world outside Malaysia, especially in the aspect of IFO that mainly explored the undergraduates' interest in making friends with foreigners and exploring their cultures. Therefore, it is clear that Malaysian undergraduates' interest to learn the English language is not concentrated on integrating with the native speakers of English as suggested by Gardner (1985). Instead, their target of 'integration' is the international community.

Figure 5. Level of each construct under International Posture
The next step is to confirm the validity and reliability of these constructs towards the affective variable, International Posture.

Validation of Reflective Measurement Model
International Posture is a Higher Component Model (HCM). The assessment of the reflective measurement model for this HCM in this sub-section is crucial to ensure that each indicator (item) in the questionnaire measures the lower order constructs that it represents. Each construct was measured through 4 to 5 indicators in the questionnaire. Details of these indicators can be referred to in the Appendix. These indicators are reflective in nature and thus represented in the reflective measurement model discussed below. The reflective measurement model involves the assessment of the validity and reliability of the indicators (items) (see Table 1). Three types of validity assessments were conducted under the reflective measurement model, i.e. (1) internal consistency, (2) convergent validity, and (3) discriminant validity. Internal consistency test is conducted through the value of composite reliability (CR). Convergent validity is the degree to which indicators of a specific construct converge or share a high proportion of variance in common (Hair et al., 2010, p. 771). In accordance with Hair et al. (2017), factor loadings and average variance extracted (AVE) were used to assess the convergent validity of the indicators in the current study.
As shown in Table 1, loadings which exceeded the recommended value of 0.708 (Hair et al., 2017) are retained. Similarly, some items with low loadings are deleted*. Even though the loading score of item IN_POSTURE_9 is only 0.650, it is retained as the sum of loadings resulted in high loading scores leading to AVE scores greater than 0.5 as recommended by Byrne (2016). Consequently, all the four dimensions satisfied the threshold values for composite reliability (CR) and AVE, whereby all the CRs are 3.5 3.6 3.7 3.8 3.9 4 4.1 4.2 4.3 4.4

Intercultural Friendship Orientations
Intergroup Approach-Avoidance Tendencies International Vocation Language and Education, 9(1), 222-236, 2022 greater than 0.7, and all the AVEs are more than 0.5 after the item deletion process (Hair et al., 2017). Hence, the reliability and validity of the measurement model in the current study are confirmed. Next, the establishment of discriminant validity using the Heterotrait-Monotrait Ratio of Correlations (HTMT) technique is shown in Table 2. Discriminant validity is established when the values are lower than the threshold value of HTMT.85 suggested by Kline (2011). Furthermore, according to Henseler et al. (2015), to confirm discriminant validity, we must ensure that the HTMT inference results do not demonstrate confidence interval values of 1 for any of the constructs. All the measurement constructs in the current study have the discriminant validity below the threshold value of HTMT .85 (see Table 2). Hence, the discriminant validity of these four measurement constructs has been established.

Validation of Formative Measurement Model
As stated earlier, International Posture is an HCM that includes both reflective and formative measurement models. As the validity and reliability of the lower order constructs have been achieved in the previous sub-sections, the validity and reliability of the higher order constructs are presented in this sub-section. The higher order constructs are formative in nature and thus represented in the formative measurement model discussed below.
In the formative measurement model, three assessment steps were taken, namely (1) assessing convergent validity, (2) addressing collinearity issues, and (3) assessing the significance and relevance of the formative indicators. These assessments were conducted to validate the formative measurement model for the higher order model of the affective variable, International Posture. Details of the measurement properties of the formative construct of International Posture are as shown in Table 3 below. Table 3. Measurement properties of the formative construct of International Posture Based on Table 3, the path coefficient of the formative construct of International Posture is 0.836. According to Klassen and Whybark (1999), the value that is more than 0.7 suggests that the construct has a sufficient degree of convergent validity. Regarding multi-collinearity, all the indicators of the formative construct satisfied the variance inflation factor (VIF) values and are consistently below the threshold value of 5 as stated by (Hair et al., 2017) and 3.3 as stated by (Diamantopoulos & Siguaw, 2006). As such, it can be concluded that collinearity is not an issue for the estimation of the PLS path model in the current study.
Besides, IAT, IFO, IFA, and IIV were proven to be significant indicators of the International Posture construct with t-values of 4.501, 4.887, 2.598, and 3.430, respectively. As such, we can conclude that the reflective-formative measurement model for International Posture in the current study is valid and reliable.
Overall, the result of CFA in this study confirmed the validity of the four measurement constructs (IFO, IAT, IIV, IFA). These four measurement constructs were included in the study of Yashima (2002), and three of these measurement constructs (exclude IFO) in the study of Yashima et al. (2004). Yashima et al. (2004) omitted the construct IFO as Yashima's participants were language learners that were in contact with the native speakers of English. However, the construct of IFO was included in the current study as our participants were Malaysian undergraduates studying in a public university in Malaysia. Other than that, the constructs of IFO and IAT were also included by Yashima (2009) in the second order model of Attitudinal and Behavioural Propensity. However, the constructs of IIV and IFA were omitted as the study by Yashima (2009)  International Posture utilised in Yashima (2002Yashima ( , 2009) and Yashima et al.'s (2004) second order model of Attitudinal and Behavioural Propensity.

Research Model
Based on the assessments that have been presented in the above sections, it is clear that the construct of International Posture is valid and reliable for Malaysian English language learners. After all the analyses for testing the validity of International Posture were completed, the final model of the scale developed to measure this variable is presented in Figure 5. Based on the results presented above, to answer research question 1, the researcher has identified the level of each measurement construct under the effective variable of International Posture. Based on the results gathered, the four measurement constructs are at the mean level of above 3.5 from the scale of 1 of completely disagree and 5 of completely agree. IFO is the highest construct with a mean of 4.3. Among the indicator item for this construct are "learning English will allow me to know various cultures and peoples" and "Learning English will allow me to make friends with foreigners". The undergraduates' interest to communicate with the international

International Posture
Interest in International Vocation/Activities Intergroup Approach-Avoidance Tendency

Interest in Foreign Affairs
Intercultural Friendship Orientation community using the English language are shown through this measurement construct.
In the measurement construct of IAT with the mean of 3.9, an example of indicator items stated is "I would talk to an international student if there is one in class". This construct represents the undergraduate's readiness to engage in a communication event using the English language voluntarily. Another two measurement constructs are IFA and IIV with the same mean score of 3.8. IFA measurement construct concerns the undergraduates' interest in foreign affairs. An example of the indicator item for this measurement construct is "It is important for me to know the current international issues". It is believed that the undergraduates' interest motivates them to learn the language that eases the process of them accessing the international information. Lastly, IIV measurement construct is measured through indicator items such as "I want to work in an international organization such as United Nation". The undergraduate's interest in travelling abroad getting involved with international activities requires them to be able to communicate using the English language.
In research question 2, the researcher validated the capability of these four constructs in measuring the affective variable of International Posture among Malaysian undergraduates. The process of validating these four constructs involved higher order and lower order model's CFA through a reflective measurement model and formative measurement model. In the reflective measurement model, internal consistency, convergent validity, and discriminant validity tests were conducted to ensure the reliability and the validity of the lower order measurement model. Through the analysis of internal consistency, seven indicator items (refer to Table 1) were deleted due to low loadings. The deletion of these indicator items improved the composite reliability (CR) to above the threshold value of .70 in all four measurement constructs, proving the reliability of the indicator items measuring the four measurement constructs. The convergent validity above .50 in all the measurement constructs signifies that the convergent validity has been established. It indicates that each indicator included in the scale reflects the convergence construct as compared to indicators measuring other constructs (Urbach & Ahlemann, 2010). The validity of the reflective measurement construct was measured through HTMT discriminant validity test. No further deletion of indicator item was made as the results satisfied the threshold value of the convergent and discriminant validity tests. In the formative measurement model, the assessments conducted were convergent validity test, collinearity test, and the significance and relevancy of the formative indicators were valuated. All the tests conducted in this phase satisfied the threshold value required (refer to sub-section 4.3.1). Therefore, no deletion of formative measurement construct was made at this phase, and all the measurement constructs (IFO, IAT, IIV, IFA) were proven to be reliable and valid in measuring the affective variable of International Posture among the Malaysian undergraduates

CONCLUSION
To conclude, this study has revealed that the International Posture among the Malaysian undergraduates is high as shown in the measurement construct levels. On top of that, this construct has been proven as valid to be operationalised in the context of Malaysian English language learning. The scale of International Posture that has been adapted in this study specifically for Malaysian English language learners is valid and reliable, and all set to be used to measure this affective variable in future studies.
Other than the operationalisation of International Posture in the Malaysian context, the study also has contributed towards the underlying design of the variable International Posture. The underlying design of this variable covered integrative and instrumental constructs of motivation in English language learning through the four dimensions (IFO, IAT, IIV, IFA) as well as its respective indicators. Therefore, this variable has been carefully studied through its lower order and higher order models by employing PLS-SEM as the statistical analysis tool. Hence, the theoretical background embedded in the model could be clearly observed and, when needed, an amendment can be made to suit future re-examination and confirmation of the affective variable, International Posture.
Undeniably, studies on International Posture are still limited, especially in the Malaysian English language learning context. This variable, derived from the socioeducational model of second language acquisition by Gardner (1985), has the potential to predict English language learners' WTC and encourage better L2 proficiency. With a scale that has been tested and validated specifically in this context, hopefully it will be used in more studies that focus on this area in the future. However, in the current study, this validated scale is only tested among undergraduates with ages range from 19-24 years old. Hence, the applicability of this scale might differ in case of other language learners with different age groups.