THE HALO SURROUNDING NATIVE ENGLISH SPEAKER TEACHERS IN INDONESIA

The Native Speaker Fallacy, a commonly held belief that Native English Speaker Teachers (NESTs) are inherently better than Non-NESTs, has long been questioned by ELT researchers. However, this belief still stands strong in the general public. This research looks to understand how much a teacher’s nativeness affects a student’s attitude towards them, as well as the underlying reasons for their attitudes. Sixty seven respondents in two groups were asked to watch an animated teaching video, after which they completed a questionnaire that used Likert-scales to assess comprehensibility, clarity of explanation, engagement, and preference. The videos for both groups were identical apart from the narrator; one spoke in British English, while the other, Indian English. In addition, they were also visually identified as Caucasian and Asian, respectively. The video was controlled for speed of delivery. The quantitative data were then triangulated using qualitative data collected through open questions in the questionnaire as well as from a semi-structured interview conducted with 10 respondents. The data show that there is a significant implicit preference for NEST teachers in the video, as well as in respondent’s actual classes. However, when asked explicitly, respondents didn’t rank nativeness as a very important quality in English teachers. This discrepancy between implicit and explicit attitudes might be due to a subconscious cognitive bias, namely the Halo Effect, in which humans tend to make unjustified presumptions about a person based on known but irrelevant information.

In this paper we will explore the Halo effect in English Language Teaching (ELT) in Indonesia.Specifically, how this particular cognitive bias influences the attitudes of learners towards Native English Speaker Teachers (NESTs), given their privileged position in ELT.But first we need to look at the context in which this study takes place as well as the previous research done in the field.

English Language Teaching in Indonesia
In Indonesian primary, secondary, and tertiary education, EFL is usually conducted using methods similar to the Grammar Translation (GT) method (Musthafa, 2001).Though Indonesian EFL teachers have been trained in other methodologies, such as the communicative approach, it seems that most teachers revert to GT due to pressures from the curriculum, standardised testing, large class sizes, and the teaching and learning culture that leans heavily towards teacher-centricity (Coleman, 1996;Musthafa, 2001;Pasassung, 2003).
As such, there has been a proliferation of private language schools that to a larger and lesser extent, use the communicative approach in Indonesia (Adi, 2012).These schools are typified by: having smaller class sizes than in formal schools (10-18 students compared to 40 -50); focusing on more communicative activities (information gaps, role plays, etc.); emphasising learning how to communicate, not just learning grammar; and, employing NESTs.
NESTs are often seen in their marketing materials, and it might give the impression that this more dynamic and communicative teaching approach (very unlike the teaching at their formal schools) is inherently the domain of the NEST.
Apart from one (Watson-Todd & Pojanapunya, 2009), these studies looked at explicit attitudes, and none have investigated perceptions of NESTs in doi: dx.doi.org/10.17509/ijal.v5i2.1352relation to the Halo effect.A discussion of the results of the abovementioned studies are in the literature review.
For practical purposes, this report will use the term native speaker as the Indonesian layperson sees it: a Caucasian English speaker from an inner circle country (see Kachru, 1992;Fig. 1.1), who used English from birth, and whose nativeness is biologically inherited.Though this usage is narrow and quite possibly discriminatory, (Rampton, 1990;McKay, 2002;Cook, 2013), Motha (2006, in Ruecker & Ives, 2014) claims that due to historical factors, "English and Whiteness are thornily intertwined " (p. 496).This can be seen in the discourse in how English teacher job advertisements in East and Southeast Asia seem to focus a lot on nativeness and even having the correct Caucasian look (Ruecker & Ives, 2014).

The Native Speaker's privileged position in ELT
There is a lot of current research and ELT thinking on the issue of nativeness, with many scholars writing about the Native Speaker Fallacy, a widespread assumption held by many practitioners, managers, and learners that Native English Speaker Teachers (NESTs) are better than Non-NESTs (Phillipson, 1992;Medgyes, 1992;Canagarajah, 1999).This is sometimes attributed to the "the Chomskyan notion that the native speaker is the ideal informant in grammatical judgements and is therefore the ultimate authority on language use" (McKay, 2002, p. 42).Many learners believe that the benchmark for pronunciation and grammar is the native speaker, and aim to have pronunciation and grammar usage that is similar to native speakers (Timmis, 2002).This is exemplified by some English language learners in Canada, who believed that only "Canadian English" is real English, and it can only be taught by a male, Anglo Saxon Canadian teacher (Amin, 1997).One study indicates that when played the same recording in English, comprehension drops when the speaker is visually identified as Asian instead of Caucasian (Rubin, 1992).
Another possible reason for this attitude is that historically, Non-NESTs, especially in less economically advantaged places such as non-white neighbourhoods in post-Apartheid South Africa, are quite often not as well trained as their NEST counterparts (Chick, 1996).Some argue that even now, many Indonesian English teachers lack the requisite mastery of language and pedagogic training to teach effectively (Sholihah, 2012;Dewi, 2011).In contrast, Indonesian regulations require foreigners to have both a degree in languages and a practical teaching qualification in order to teach English in Indonesia (Menteri Pendidikan Indonesia, 2009).It is easy to see why many of the students at these schools would welcome a well-trained and fluent NEST.
There is also a disconnect between what the layperson and the ELT scholar defines as a native speaker.The Collins Online Dictionary (Anon, 2015) defines it as a person "who speaks that language as their first language rather than having learned it as a foreign language."However that is only one of the three different approaches that Cook (2013) outlines in defining nativeness: 1.The historical approach, dependent on the language one inherits from one's parents; 2. The components approach, looking at the characteristics implicit in nativeness; such as an intuitive understanding of the rules and an ability to be creative in language usage; and, 3. The social identity approach, ones nativeness is dependent on if one identifies with a particular culture, or group of people.
As you can see, the latter approaches look at nativeness as something fluid and changeable, e.g.given enough training an L2 user can gain a deep understanding of grammar to become a native speaker equivalent (Medgyes, 1992;Kachru, 2005).Alternatively, an immigrant can be so immersed in a culture that she identifies more with her adopted country, picking up its linguistic cues, and then become a native speaker.
However, society as a whole still tends to believe that nativeness is inherited as well as a prerequisite to good language teaching.This is exemplified by: 1.The institutionalisation of nativeness in Asian immigration laws in countries such as Korea and Indonesia, which require NESTs to hold a passport from a largely white, native English speaking country, such as Australia, Ireland, or the United Kingdom (Ruecker & Ives, 2014, Menteri Pendidikan Indonesia, 2009; 2. ELT job listings and hiring practices, where there is an explicit preference towards native speakers (Clark & Paran, 2007;Mahboob & Golden, 2013), especially Caucasian NESTs (Ruecker & Ives, 2014); and, 3.In Japan, NESTs are paid more than Non-NESTs, just for the sake of their nativeness (Butler, 2007).
Ruecker (2011) argues that these types of regulations, assumptions, discriminatory job listings and hiring practices only further perpetuate the native speaker fallacy by "making them invisible and less likely to be challenged" (p.407), feeding a positive feedback loop.

The debate over the term Native Speaker
There has been an effort to break away from the term native speaker (see Kachru, 1992;Paikeday, 1985;Medgyes, 1992).Edge (1988) proposed the terms more or less accomplished English speakers, whereas Rampton (1990) tried to shift focus from nativeness to language expertise and affiliation, saying that nativeness "spuriously emphasizes the biological at the expense of the social….(and mixes up) language as an instrument of communication with language as a symbol of social identity" (p.98).However, Medgyes (ibid.)argues that these terms and definitions often overlap and lack rigour; preferring himself to use the term "native speaker" despite its inherent issues.Kachru's concentric circles have also come under scrutiny, with both Graddol (1997) and McKay (2002) arguing that the centrality of the inner-circle (native speaker, norm making countries) has implied that native speakers are the only source of correct usage and are the best teachers, thus bestowing upon them special status in defining language pedagogy.Graddol (ibid.)proposed an update to the concentric circles, suggesting three overlapping circles (Fig. 1), in effort to change the traditional view of rigid biological and geographical demarcations.
This has also been mirrored by Kachru (2005) himself, who proposed that the inner circle should now be conceived as the whole group of proficient English speakers who have "functional nativeness", regardless of how it is used or learnt (Fig 2).Graddol, 2006, p. 110) This view of looking at nativeness not as a binary but as a continuum is also echoed in the writings of Medgyes (1992), who claimed that there exists a continuum on which language learners constantly move along as long as they are studying the language (Fig. 3).
Figure 3. Medgyes modified version of the interlanguage continuum (Medgyes, 1992, p. 342) Although the term is still used widely by the laity and scholars alike, Canagarajah (2005) argues that the distinction between native and non-native speakers no longer applies; globalisation, the spread of English, and the blending of cultures has created a world in which speakers of different varieties of English will use whichever variety they are comfortable with to communicate, and it will probably not be an inner circle variety.

Perceptions of NEST and Non-NEST teaching strengths
There have been many probes into the different advantages and disadvantages of NESTs and Non-NESTs (see Arvizu, 2014;Medgyes 1992;Lasagabaster, & Sierra, 2002;Lee & Lew, 2001;Ma, 2012;Mahboob, 2004;Moussu, 2006;Wu & Ke, 2009), but it seems that their respective strengths and weaknesses are quite often complementary.For instance, Ma (2012) reported that EFL learners found understanding NESTs instructions and explanations difficult, whilst finding the same easy with Non-NESTs.It can be seen from Table 1 that NESTs and Non-NESTs have few overlapping strengths, and many complementary ones.It follows that for most contexts, it would be beneficial for schools and learners to have a mix of NESTs and Non-NESTs.

Preferences towards NESTs and Non-NESTs
Many attitudinal studies show that there is a preference in learners and school managers for NESTs.In Indonesia, there seems to be a high preference for NESTs in tertiary education (Dewi, 2011;Murtiana, 2012), with 86% of students in one study believing that NESTs are a prerequisite to successfully learning English, and 91% believing non-NESTs are less effective teachers.In Taiwan, there are similarly strong beliefs (Wu & Ke, 2009), but in Thailand (Watson-Todd & Pojanapunya, 2009), Spain (Lasagabaster & Sierra, 2002) and Mexico (Arvizu, 2014), the preference isn't as marked.
There is also a clear preference for hiring NESTs by school (Clark & Paran, 2007;Mahboob, 2004Mahboob, , 2005;;Moussu, 2006;Ruecker & Ives, 2014).Some have explained this phenomenon by pointing to market forces, with management responding to the student's demand of NESTs; however, Holliday (2008) points out that if students demanded teachers based on gender, management might not be so accommodating.It seems that, at least in East and Southeast Asia, this pandering has led to discriminatory hiring practices in which not only are Non-NESTs paid less than NESTs (Butler, 2007), but it is also implied that Non-NESTs need not apply (Ruecker & Ives, 2014;Watson-Todd & Pojanapunya, 2008).

The Halo effect
The Halo effect is defined as the "widespread human tendency to make unwarranted inferences about a person's unknown characteristics on the basis of known but often irrelevant information" (Forgas, 2011, p. 812).This bias causes people to think, among others, that: a woman's writing ability is higher if she is attractive (Landy & Sigall, 1974); the same man's appearance, mannerisms, and accent is perceived more positively if he is friendly (Nisbett & Wilson, 1977); and, a political candidate is seen as more competent if she is attractive or has a familiarity to the voter (Verhulst, et al. 2010).
Given the privileged position of the native speaker in defining language pedagogy (McKay, 2002); the problematic intertwining of the English language and race (Motha, 2006in Ruecker & Ives, 2014); as well as the fact that intelligence and competence is often based on physical appearance (Moore, et al. 2011;Verhulst, et al., 2010), it is easy to propose that the so-called Native Speaker Fallacy is a type of halo effect surrounding NESTs as defined in this study, at least in Indonesia.Especially when considering the high rates of students who think they are essential to the learning process (Murtiana, 2011).Haselton, Nettle, and Andrews (2005) claim that these biases are hard-wired into the human brain by evolution.As such, awareness might not be enough to mitigate its effects.

Call for research
As the privileged position of the NEST in ELT has been put to question, it is important for researchers to explore not only the 'hows and whys', but also the extent to which it biases the learner.Although there have been probes into the Halo Effect in education (see Shevlin, et al., 2000), there has been no writing on it in ELT apart from a blog post (Michelioudakis, 2014).That said, there has been research in ELT on how non-language factors, such as accent and ethnicity can affect learner attitudes towards teachers in general (Rubin, 1992;Boyd, 2003), including matched-guise research focusing on attitudes towards accents alone (Saravanan & Poedjosoedarmo,1996;Dalton-Puffer, Kaltenböck, & Smit, 1997).
This research asks similar questions, but using teaching videos taught by an Asian and a Caucasian teacher.This is to explore the extent to which the widespread bias towards NESTs colours a student's judgment of a teacher's efficacy in Indonesia The research questions are to explore: (1) Do Indonesian students perceive native speakers to be better language teachers?(2) How do Indonesian students define native speakers?(3) What do Indonesian students believe are the differences and similarities in being taught by NESTs and Non-NESTs?And Why do students believe what they do?

METHOD Overview
This research was designed to elicit respondents' implicit attitudes towards NEST and Non-NEST teachers.Firstly, a quasi-experiment in which participants from a private English language teaching school in Bandung, Indonesia, watched different teaching videos (Fig. 4 and 5) and answered a questionnaire was done.Secondly, a further interview was conducted to shed further light on the matter.
Both research tools had a large focus on gathering qualitative data, which can help to better interpret how variables are related as "the telling anecdote may be much more revealing and influential than almost any amount of figures" (Blaxter, et al., 2010, p. 205).Data from both research tools can be accessed through the links in the Appendix.
One thing to note is that this research utilised convenience sampling (see Dornyei, 2007, p. 98), the only criteria to selection of age and a willingness to volunteer.As such, findings might not be generalizable.The Quasi-Experiment Participants were put in two groups (G1 and G2) and were asked to watch one of two videos that taught the usage of the prepositions of place, and then completed a questionnaire (see Appendix).The videos were identical except for one key difference; G1's video was narrated in a British accent and had a picture of a Caucasian teacher with a British sounding name (ostensibly a NEST); G2's video, was narrated with an Indian accent and a picture of an Indian teacher with an Indian sounding name (ostensibly a non-NEST).
The Questionnaire Participants then completed a questionnaire asking them to rate the video and the teacher (see Appendix).Questions were designed to elicit respondents' implicit attitudes towards the ostensible NEST and Non-NEST in the video, as well as explicit attitudes and preferences towards NESTs and Non-NESTs in general.Data was then analysed to see if there was a statistical significance in the difference between teacher/video ratings in G1 and G2 using the Chisquared test.The hypotheses for the tests are as follows: There is no observed difference in rating between both groups.
There is a higher rating for the teacher in G1 than the teacher in G2.
There is a higher rating for the teacher in G2 than the teacher in G1.
Qualitative data was gathered through a post questionnaire interview, and the questionnaire itself, using open questions that required both short answers (Fig. 6) and extended answers (Fig 7).

The Interview
Semi-structured interviews were conducted with 5 participants from each group (10 in total).The interviews were done in a mix of Indonesian and English, depending on the preference of the interviewees.As the researcher is a multilingual speaker of English and Indonesian, interviews were transcribed and analysed in their original language.This qualitative data were then analysed using the moves outlined in Miles and Huberman (1994, pp. 9-12).

Limitations of the study
The biggest limitation to the study is that the videos were narrated by different people.This increases the number of variables, impacting validity (Blaxter, et al., 2010).Perhaps participants were responding to teacher's accent, or pace of delivery, not their perceived nativeness.For this reason, the questionnaire also elicited participants' attitudes towards the teacher's accent, personality, pace of delivery, and choice of words (see Fig. 6).Of course, out of those variables, choice of words was exactly the same, and pace of delivery, though slightly different, was very similar (725 words in ~5 minutes for both videos).This research was designed so the only variables that influence participant ratings were: the Caucasian and Indian pictures, and the British and Indian accents and names (see Fig. 4 and 5). Figure 6.Sample questions from questionnaire (A).Short answer questions and Likert scale question.
Figure 7. Sample questions from questionnaire (B).Extended answer questions Another limitation is the relative unfamiliarity of the Indian accent compared to the British accent to Indonesians.The rationale for using the Indian variety of English is that even though it isn't regarded as norm-making (Kachru, 1992), it is still a native variety, and as such might be more useful in exploring the attitudes towards perceived nativeness and the 'thorny intertwining' of English and whiteness described by Motha (2006, in Ruecker & Ives, 2014, p. 496).
Finally, the convenience sampling along with the low sample size is also another limitation.It is quite likely that this sample is not representative of Indonesians in general, so we should be careful not to generalise the results that widely.

The participants
The 67 participants were divided into two groups, G1 and G2, which had 35 and 32 respondents respectively.A large majority of the participants were young adults (less than 8% were over 25, none under 30) and speak Indonesian as a first language, though it must be noted that G1's mean age was higher.Around a third of participants in both groups were in high school; 10% of G1 work, while none of G2 do; and the rest study at university.G2's English ability was more uniform; more than half are self-identified pre-intermediate English users with the rest being intermediate and upper intermediate.This is contrasted to the more diverse G1, with 40% elementary users and 20%, 25%, and 15% of pre-intermediate, intermediate, and upperintermediate users.Both groups had similar experiences in formal and non-formal English learning.However, there was much more agreement in their purposes for learning English (Table 2).
Even though there are differences, both groups are learning English for the similar purposes.A similar proportion of respondents need English for: Very much YES! ---1 ---2 ---3 ---4 ---5 ---6 ---Very much NO! 2.3."Mastery of the language" is thought by some to be one of the characteristics of effective language teachers.Do the teachers need to be native speakers to have "mastery of the language?"Why? "Keahlian bahasa" sering dianggap sebagai karakteristik penting untuk menjadi guru bahasa yang efektif.Apakah karakter ini hanya dimiliki oleh "native speaker?"It must be conceded that this difference in demographics might be partially responsible for the divergence in the responses in the questionnaire.That said, this school markets itself to a certain socio-economic strata, and as such the students students who enrol there are have many common traits, they are mostly relatively wealthy, welleducated, and comparatively sophisticated.This is supported by their having enough disposable income to study English in a relatively pricey non-formal educational institution, as well as the similarities in their purposes for learning English, which include the relatively expensive further education abroad in ESEs.That said, we must still keep in mind the fact that these groups are not homogeneous, and that this diversity might be the cause of their divergent attitudes.
G1 or G2 is the group (respectively NEST, and Non-NEST), 'Q' is for Questionnaire, 'I' for Interview, and the number at the end is the participant number.E.g.G2Q.31 is Questionnaire respondent 31 in the Non-NEST group, and G1I.03 is Interviewee 3 in the NEST group.

Implicit preferences towards video and teacher
Four questions were asked to determine respondents' implicit attitudes towards their teacher and video, namely: Q1. Was this video engaging?Q2.How clear was the explanation?Q3.Was the teacher easy to understand?Q4.Would you like to be taught by her in reality?
The use of a Chi-Squared test uncovered a statistically significant difference in three out of the four questions (Figures 8 to 11 This data are supplemented by the qualitative data gleaned from the open questions regarding accent, pace of delivery, choice of words and personality.Each response was given a numerical value; '+1' for a positive comment, '0' for a neutral one, and '-1' for a negative comment.Using this quantification (Table 3, Q columns), there is a definitely higher rating for G1 in two categories, namely accent, and pace of delivery.However, the difference wasn't as marked for choice of words.For personality, respondents showed no preference.We can also see recurring themes in each category (Table 3).The equally positive response to the choice of words and personality suggest that the only variables the participants responded to were the teachers' names, pictures, accents, and pace of delivery; all of which are indicative of their perceived nativeness.The big difference in attitudes towards pace of delivery is slightly odd as pace was only different by 1% (1.4 wpm).It does seem that the reasons for the divergent results were because of these factors: the foreignness of the G2 narrator's accent (Table 3, items 'strong', 'difficult', and 'not native', divergent ratings for comprehensibility [Q3]); these particular respondents want to be taught by the native speaker accent (see Table 4); and, the visual identification of the G2 teacher as Asian might have decreased comprehension compared to the Caucasian G1 teacher (Rubin, 1992).
With the limitations in mind, this data does seem to be indicative of there being an implicit preference towards NESTs in Indonesian private language schools, most else being equal.Specifically for this study, the factors that might have influenced results were the desirability of the 'native accent' (Timmis, 2002.), the intertwining of nativeness and race (Motha, 2006), and the Native Speaker Fallacy (Phillipson, 1992).
The results imply that race and accent are important to Indonesian language learners, and that these factors (desirable accent and physical appearance) could influence a student's perceptions of competence in a teacher.This is true in other parts of the world as well (Ruecker and Ives, 2014), e.g. in Thailand people are more to accept white Non-NESTs than an Asian NEST, suggesting that whiteness is desirable, prestigious even (Watson-Todd and Pojanapunya, 2008).This implicit preference is also in accord with research into explicit preferences in Indonesia (Dewi, 2011;Murtiana, 2012) as well as the explicit attitudes reported by respondents in this study (see General Teaching Preferences).
What is interesting is that this preference exists in this school, which places a heavy emphasis on professional development.Both NESTs and Non-NESTs are required to, among others: 1) have a CELTA (Cambridge ESOL practical teaching course entitled 'Certificate in Teaching English to Speakers of Other Languages') to be hired full time, 2) attend at least 6 practical workshops a year, 3) be observed at least twice a year, 4) create and follow through with development plans based on aforementioned workshops and observations.It is also not uncommon for teachers from other schools to be trained there, either in short courses introducing language teaching, or the CELTA as it is also the only approved CELTA centre in Indonesia.That said, many other minority and non-Caucasian teachers, including myself, have had to 'invest a great deal of energy in establishing themselves as authentic teachers in the eyes of both their students and their colleagues' (Amin, 1997:581).It's telling that a statistically significant preference exists even at a school which places such a high emphasis on training their teachers to the same standard.
However, I'd like to put forward the very uncontroversial idea that this preference isn't a failure on the part of the Non-NESTs teachers.I believe the issue is partly to do with the context in Indonesia which is already somewhat biased towards NESTs (see next section), combined with the biases in human psychology that are difficult if not impossible to dispel (Haselton, Nettle, & Andrews, 2005).Just as physical attractiveness can implicitly influence how intelligent or competent a person is perceived (Landy & Sigall, 1974;Moore, et al., 2011), a teacher's race and accent can also implicitly influence a learner's attitude.I'll finish this section with this comment from an interviewee, talking about essential teaching characteristics: "If they look at you and think that you're not native, they're gonna think you're not as eloquent as natives, er… no matter how good you are, they're gonna put you down because of how you look."G1I.05

Respondent definition of 'native speaker'
Before we continue, it would be beneficial to see how the interviewees defined 'native speakers' (sample size = 10).Someone who; -Learns and/or uses the language from a young age (n=8); -Understands the culture of a 'native country' (n=5); and, -Uses it actively in daily life (n=4).
This quote is indicative of the general opinion of respondents in this matter: "(a native speaker is) someone who had been growing up with the language and the culture, affected by the language and culture..." G2I.01 It seems that most respondents seem to view nativeness only in the historical approach (Cook, 2013), seeing it as being inherited from the culture and the environment one is during childhood.Race was also mentioned in the interviews (n=3); "(to be a native speaker) you don't have to, you know, … happen to be, er… white… " (laughs) G1I.01 This seems to show that while you don't have to be white, being white is a presumption of being a native speaker.This is aligned with regulations governing which nationalities are allowed to teach English in Indonesia, namely, those with passports from a mainly white, English-speaking country (Menteri Pendidikan Indonesia, 2009).
Half the interviewees said that nativeness is inherited, with one saying that: "however linguistically competent we become, in English for example, we will never be as perfect as a native speaker."G2I.02 (emphasis mine) This is supported by questionnaire data, with responses such as 'NESTs speak better' (n=2) and 'NEST pronunciation is better' (n=9).I believe that the Indonesian regulation mentioned above is partly responsible for what learners believe.This institutionalised discrimination invisibly reinforces the notion that nativeness is based on hereditary and accent (Ruecker & Ives, 2014).When asked about the 'correct' English pronunciation, one respondent replied: "In English?Yes… This.. (laughs) … I don't … like speaking this, I think as long as it's understandable… (interviewer: understandable to whom?) … To everyone.Like in a class."G1I.01 Scholars also point to the fact that English doesn't belong exclusively to the inner circle countries anymore, saying that 74% of communication done in English are between nonnative speakers, and that non-native English speakers outnumber native speakers by at least twoto-one (Graddol, 2006).Even though many are still of the opinion that English belongs to the native speaker, one respondent mentioned the role that English plays in the global village; "Because in this world we are speaking English to everyone in international relationship, not only native speaker."G2Q.01 Another finding is that half of the interviewees (n=5) have similar beliefs as Medgyes (1992), in that nativeness isn't binary, and native-like competence is achievable for non-natives.Questionnaire data implies that, although difficult, you can learn to become a native speaker (n=7).Knowledge of culture is seen as a badge of nativeness (interview data, n=5), with two interviewees identifying knowledge of idioms to be another (G2I.4/G2I.1).It's interesting to note that when asked in the questionnaire whether nativeness is a prerequisite for 'mastery of the language', 80% said 'no' (n=54).So it seems that the Chomskyan notion of natives being the only source of 'correct' English (McKay, 2002) is being challenged.

Explicit attitudes Attitudes towards NEST and Non-NESTs
When asked directly about the ideal combination of teachers, all but 3 questionnaire respondents said that they would prefer to be taught by a mix of NESTs and Non-NESTs (Table 4).This might be because some respondents would like to 'compare NESTs and Non-NESTs' (n=3).Even though respondents would like a mix of both, like in Spain, it seems that NESTs are more desirable (Lasagabaster, & Sierra, 2002).When elaborating, students cited the reasons in Table 5.
In interviews participants mentioned similar themes to the questionnaire, preferring NESTs because: good pronunciation (n=4); an opportunity to learn about 'native' culture (n=2); and, the 'native' educational culture is more relaxed (n=1).While preference for Non-NESTs is because: they empathise and relate better with learners (n=8); talking to NESTs is intimidating (n=2); and, shared language and culture helps Non-NEST explain better (n=1).
However there was disagreement about who explained better with some preferring NESTs (n=2) and others preferring Non-NESTs (n=3).
"(Non-NESTs) make me more comfortable to learn, (they're) easier to socialise (with) and easier to understand what they said because they can explain it in my language" G2I.2 "Easier to (understand) the subject matter (with NESTs)... they have mastery of the language…so it's easier to understand them."G1I.2 -There is quite a split in opinion with regards to: A.) the ability to teach written communicative skills (G1 = 45.3%;G2 = 62.5%),B.) mastery of the language (G1=54.3%;G2=37.5%), and C.) an indepth knowledge of the culture of the language (G1=20%; G2=50%).
-A final thing to note is that out of the top eight characteristics that both groups agreed to as the most important, five (63.5%) are about linguistic knowledge and pedagogic ability (C1, C2, C4, C6, C7).The interviews yielded another interesting common theme, most (n=8) believed that personality traits were more important than pedagogic ability.Respondents cited the ability to 'read the mood' of a class and to react flexibly to it as one of the reasons why it is so important (n=4).Some (n=4) changed their mind while speaking, first stating pedagogic ability is equal to personality and then suddenly back-peddling.
"(I think it's) equally important, because if it's not balanced, like… if he can teach well but has a lousy personality the kids will get lazy, right?And if he's got a good personality but can't teach very well, er… that can still be tolerated..So I guess I think personality is more important."G2I.01This is supported by questionnaire data, with some respondents explicitly saying it's not about native speakerness, it's about personality (n=3).
A final note to ponder comes up when comparing the top 6 in Table 6 with what the literature has shown to be the strengths and weaknesses of NESTs and Non-NESTs around the world (Table 1, Perceptions of NESTs and Non-NEST teaching strengths synthesised from: Arvizu, 2014;Medgyes, 1992;Lasagabaster and Sierra, 2002;Lee and Lew, 2001;Ma, 2012;Mahboob, 2004;Moussu, 2006;Wu and Ke, 2009).It seems that four out of six are thought to be Non-NEST strengths (C1, C3, C5, and C6).When coupled with the fact that respondents overwhelmingly think that mastery of language (C7) isn't within the exclusive domain of the native speaker, it seems to further hint at the extent of the Halo Effect's power to bias preferences.

CONCLUSIONS
I believe that this study has partly answered the research questions guiding it:

Q1
Do Indonesian students perceive native speakers to be better language teachers?Although there are limitations in the study, it does seem that Indonesians do see NESTs as better teachers.Three out of four preference metrics were statistically significant in favour of the NEST.That said, even though learners had 1.) a statistically significant preference towards the NEST in the experiment; and 2.) preferred to be taught with more NESTs than Non-NESTs; it does seem that many are cognizant that nativeness isn't really necessary for teaching."It's like this, you see?Even though we're Indonesian, it doesn't mean we can teach others to speak Indonesian."G2.I.5

Q2
How do Indonesian students define native speakers?Many seem to take the historic approach (Cook, 2013), believing nativeness is inherited from one's parents and environment.There is an implicit presumption that race is part of the equation, along with accent, and country of origin.That said, this does seem to be changing, with some respondents believing that nativeness is fluid.The belief that

"
I think you can acquire nativeness as you get older...I don't think you can say it's like an imprinted DNA thing."G1I.05

Table 2 .
Purpose of studying english

Would you like to be taught by her in reality? Circle one number.
Apakah Anda ingin diajari oleh guru tersebut?Lingkari satu angka.

Table 4 .
Ideal NEST and Non-NEST preferences

Table 5 .
Reason for students' preferences

Table 6 .
Respondents' perception on the characteristics of effective language teachers