Cultural Identification and Language Learner
Autonomy in the Western and Non-Western World:
An Exploratory Study
Geert LoerS2333562
MA thesis, Department of Applied Linguistics Faculty of Arts
University of Groningen
Supervisors:
dr. R.G.A. (Rasmus) Steinkrauss dr. H.I. (Hilde) Hacquebord
Date of completion: June 16th, 2018 Word count: 13.726
Acknowledgements
Creating a new research instrument for a master’s thesis project is a rather optimistic undertaking, let alone testing it as well. Therefore, the past period has been both positively thrilling and distressing. After having completed the various steps in the design of the
instrument and the analyses of its results, a feeling of necessity to express gratitude emerges, for this entire project would not have been feasible without the help of a number of people. Here, I would like to give thanks to those who have prominently influenced the entire process described before.
First and foremost, a great thanks has to be given to dr. Rasmus Steinkrauss, who
supervised this thesis project not only by doing what is expected of a supervisor, but also by providing great insights into the aspects of this project which would make it a success. Additionally, his efforts and support in constructing the instrument for this project deserve acknowledgement, as they created structure in an otherwise quite disordered process.
Secondly, the opportunity to receive help from the supervisor of my current internship makes for drs. Jeroen van Engen to deserve great credit. Although my thesis project has not officially been connected to my internship at the Language Center of the University of Groningen, Jeroen has had a significant impact on the developed research tool in the design and distribution steps of this process. In other, even more grateful words, the most notable reason for which a large number of respondents was reached has been the efforts of Jeroen in distributing the instrument on behalf of the Language Center. This means that the strength and success of this study rely in part on his indispensable help.
Additional thanks are owed to the participants of the pilot study in this project. Whereas all subjects of the final version of the developed questionnaire ought to be thanked for their time and effort, the participants in the pilot study additionally took the time to have a face-to-face discussion about the various parts of the instrument, concerning its content and clarity.
Conclusively, the thesis project which is described in this article is considered a joint effort for which students, teachers and other professionals have worked in order to make a valuable contribution to the study of the relationship between cultural identification and autonomous language learning. It is for this great effort that I sincerely hope to have added to the knowledge about these topics in linguistic research, and especially to have created a solid starting point for future research on these matters.
Contents
Introduction... 2
Background... 4
The Definition and Implications of Learner Autonomy... 4
Technical, psychological and political autonomous language learning... 5
The Cultural Heritage of Autonomous Language Learning... 6
Ethnicity and cultural identification... 7
Western and non-western cultures... 8
Language learning autonomy in the western and non-western world... 10
The Dynamics of Language Acquisition and Influences on Learning Strategies... 13
Research Objectives... 14
Method... 16
Subjects... 17
Materials... 19
Procedures... 23
Design and Analyses... 24
Results... 26
Effects of Cultural Identification on Autonomy Perception... 27
Effects of Previous Learning Experiences on Autonomy Perception…... 28
Effects of Gender on Autonomy Perception... 29
Discussion... 31
Cultural Identification and Language Learning Autonomy... 31
Previous Experiences, Gender and Language Learning Autonomy... 33
Conclusions... 36
References... 38
Appendices... 42
Abstract
It is frequently assumed that the way in which autonomy is valued in the process of learning a foreign language may be influenced by the culture with which a learner identifies.
Investigating non-Dutch students at Dutch universities who are learning Dutch, the present exploratory investigation uses a newly developed questionnaire to find empirical evidence for this influence in both western and non-western learners. Previous learning experiences and gender have also been found to possibly influence the perception of autonomy, and are therefore incorporated in the research instrument. A Wilcoxon Signed-Rank test revealed that 60 western learners did not perceive autonomy differently from 22 non-western learners, Z = .058, r = .013, although there was a slight tendency for western learners to perceive autonomy more positively. In contrast, previous experiences turned out to be significantly, and
moderately strongly, positively correlated with the perception of autonomy, Pearson’s r(299) = .39, p < 0.001. Finally, a Wilcoxon Signed-Rank test revealed that 108 male learners, in spite of a weak trend, did not perceive autonomy significantly better than 203 female learners,
Z = .54, r = .12. In general, it appears that language learning autonomy is not significantly
influenced by cultural identification or gender. Previous experiences, however, have been found to influence autonomy perception. The absence of representation of different cultures in the present study and the environment in which the participants learn Dutch may have
implications for the validity of the results.
Keywords: Language learning autonomy, cultural identification, previous learning
Cultural Identification and Language Learner Autonomy in the Western and Non-western world: An Exploratory Study
Since the beginnings of research on learner autonomy, it has been questioned whether the process of independently learning a foreign language is influenced by the cultural
characteristics of the learner (Palfreyman & Smith, 2003). It is established that autonomy originated in Europe and that it is more strongly embedded in language learning practices in the countries of this continent (Schmenk, 2005). The question remains whether this is due to the overlap between the characteristics of learner autonomy and of the different cultures of these European countries, or if there may be other reasons, yet to be discovered, why learner autonomy is not yet as firmly rooted in cultures in other parts of the world. It is stressed that cultural and political influences on the way learner autonomy is valued in particular cultures around the world have been heavily underestimated, and that autonomy itself cannot be seen as a way of learning languages or a support for classroom learning that is beneficial for all foreign language learners as opposed to learners from one particular cultural background (Schmenk, 2005). Research into the possible embeddedness of cultural characteristics on the perception of learner autonomy is therefore assumed to provide new insights into the
prospects and limitations for language learning for learners of varying backgrounds (e.g. Benson, 2007; Palfreyman & Smith, 2003; Riley, 1988; Schmenk, 2005).
Therefore, this study uses a questionnaire to shed light on the differences between western and non-western cultures with regards to the perception of learner autonomy. Testing the three versions of learner autonomy as described by Benson (1997) and explained further in this investigation, it is examined whether cultural characteristics of learners from the western world make for these learners to perceive autonomy in a more positive way than learners with non-western cultural characteristics. As this is an exploratory study, an almost entirely new questionnaire is developed as a measure of the cultural perception of autonomy. For the initial validation stages of the questionnaire, foreign language learners of varying backgrounds are tested in a specific setting. In this instance, learners from western and non-western cultures who are students at Dutch universities in the Netherlands and learning Dutch as a foreign language are tested and given a so-called ‘autonomy score’. Learners who identify with Western-European cultures and learners who identify with East-Asian cultures are then compared with respects to their autonomy score. In other words, whereas cultural
identification is taken into consideration in the analyses of this study, nationality is not; the importance lies in characteristics of particular cultures, and identification with these
autonomy. Cultures which do not fall under the categories of western and non-western are excluded to increase homogeneity between participants of a particular group in the present investigation. In future studies, the implemented questionnaire can be elaborated upon and used to test learners from additional different cultures.
Although previous research has not found an answer to the question whether language learning autonomy is influenced by cultural identification, the western-based nature of the phenomenon might indicate that learners of Dutch from western cultures are more likely to value an autonomous learning process than learners from non-western cultures. It is therefore expected that those learners who identify with western cultures will receive a higher
autonomy score than those who identify with non-western cultures. However, it needs to be stressed that the development of a new questionnaire and the lack of empirical evidence for cultural bases of autonomy perception make for this investigation to be an exploratory study. Future studies may then refine the methodology and analyses used in this investigation and apply them to increase the knowledge on influences of cultural characteristics on learner autonomy.
Background The Definition and Implications of Learner Autonomy
Learner autonomy has been originally characterised as a learner being “capable of taking charge of his own learning” (Holec, 1980, p. 4). This definition has been drawn up as part of the initial studies on autonomous language learning, but has been added to by later studies in the sense that autonomy is not merely a capability to learn languages independently, yet rather includes the learner to give direction to their personal developmental process and to possess a skillset that allows the learner to obtain a higher proficiency level outside of the classroom setting (Benson & Voller, 2014). Although autonomy is now accepted to include critical self-reflection on the personal language learning process, there is no specific definition that is shared by all investigators of this phenomenon, especially when the perspective of the language teacher as opposed to the learners themselves is to be taken into consideration (Palfreyman & Smith, 2003). The authors state that independence may be an objective for language learners in general, but that autonomy can also occur when a learner is guided towards independence by a teacher, meaning that the process of becoming more autonomous is the learner’s personal responsibility but does not have to occur entirely individually (Palfreyman & Smith, 2003).
The extensive amount of literature agreeing with the independent, self-reflective nature of autonomy, where learners can still be guided in the process of becoming more autonomous, leads the current investigation to implement the definition as drawn up by Holec (1980) and added to by Benson and Voller (2014). In this study, therefore, learner autonomy is
characterised for all language learners as a capability to learn a foreign language
independently and to give direction to the personal developmental process, which includes a critical self-reflection on the learner’s own language development (see Benson & Voller, 2014; Holec, 1980). Benson (1997) additionally states that there has not been one definition of autonomy that is generally supported by all researchers on this phenomenon. According to the author, this may be due to the fact that autonomy is a concept with a three-way
subdivision into technical, psychological and political autonomous language learning (Benson, 1997). This division will be used in the present study to cover all aspects of autonomy; if it is indeed the case that cultural characteristics make learners perceive
autonomy differently, an all-embracing view on the nature of autonomy is needed to cover the possible different interpretations of its meaning.
In understanding the construct of language learner autonomy, several additional characteristics that may be valued differently by learners from varying cultures need to be
distinguished. Firstly, autonomy is not necessarily a capability that a learner either possesses or lacks; rather, gradually taking charge over the personal learning process is an important aspect of autonomy and the ability to learn autonomously, therefore, needs to develop as the learner discovers how to direct the learning process and take control over the used materials (Little, 2007; Reinders, 2010). Secondly, it is essential to incorporate the language teacher in the understanding of learner autonomy; in taking charge over the learning process, autonomy especially occurs to its full potential when a learner is given more responsibility by the language teacher and other foreign language learners (Little, 2007). Thirdly, although there is a significant influence of the teacher and peers on autonomy, Oxford and Nyikos (1989) add that there are multiple factors which determine the choice of a particular learning strategy, and that opting for autonomous language learning as a way of improving foreign language
proficiency can thus be affected by multiple factors. As is mentioned before, cultural identification is the investigated factor in this study that is assumed to influence the
perception of the characteristics of learner autonomy explained here, and in turn might lead the learners to choose autonomy as a learning strategy to a lesser or higher extent.
Conclusively, the characterisation of autonomy as a capability to learn independently is well supported, but it is to be reconsidered which factors influence the extent to which autonomy plays a role in the language learning process and how successful autonomous learning can be.
Technical, psychological and political autonomous language learning.
It is clear that language learner autonomy is a complicated construct with many aspects to be studied. Therefore, in order to investigate autonomy in learners with different backgrounds, all different aspects of autonomy need to be taken into consideration. The three versions of autonomy as described by Benson (1997) are used in this investigation to test independent learning in western and non-western students as thoroughly as possible. These versions are technical, psychological and political autonomy, and are generally regarded as sub-definitions of the notion of taking charge of one’s own learning process as described by Holec (1980). Therefore, the three versions will result in an overall view on autonomous language learning for the participants in this investigation. In other words, these three versions are not
investigated separately, but result in an overall view when combined into a single indication of the perception of autonomy for all participants.
Firstly, Benson (1997) describes technical autonomy as language learning in a non-classroom setting and in which the learner does not experience influences from a foreign language instructor. Secondly, psychological autonomy represents the learner’s ability to regard independent learning as beneficial for a higher proficiency level, and indicates a
responsibility for the personal learning process in a classroom or a non-classroom setting. Lastly, political autonomy represents taking control over this process, and entails that the learner manages the content of their learning in either a classroom or a non-classroom context. In general, it is stated that these different versions are, to a certain degree, expressed
differently in each individual language learner; for example, psychological autonomy can be expressed in the confidence of a learner to independently increase their proficiency level, and other versions are expressed in other ways of learning independently. Nevertheless, it is clearly stated that there is significant overlap between the three versions of autonomous learning, and that these versions are rather different views on the concept of autonomy as opposed to three separate segments which, taken together, comprise this concept (Benson, 1997). Regarding the present study, it would be infeasible to try to distinguish between participants of varying cultural background and investigate either of the three versions of autonomy separately; technical, psychological and political language learning autonomy will be analysed all together. In this way, the three versions represent all possible aspects of autonomy in a foreign language learner, but add up to a single ‘profile’ of autonomy for the subjects in this study.
The Cultural Heritage of Autonomous Language Learning
Although language learner autonomy as a benefit to the enhancement of foreign language proficiency was not widely researched before its instantiation by Holec (1980), it can be argued that the principles of taking charge of one’s own learning, together with a positive attitude towards this way of learning, trace back to the Enlightenment period in Europe: The values of autonomous language learning are considered to be a product of the
individualisation which started to increase in this period, and these values indicate a capability to take decisions concerning personal growth (Benson, 2012; Schmenk, 2005). It is for this cultural heritage that autonomous learning is regarded a western principle (Schmenk, 2005). Surprisingly, researchers who have considered or stated that language learner autonomy is clearly a western-based principle also occasionally attempt to make a case for its homogenic and universal character, arguing that autonomy is becoming a universally more wide-spread strategy for language learning (Little, 2007; Schmenk, 2005). For example, Schmenk (2005) explains that an increasing amount of research focuses on the prospects for successful language learning autonomy in other parts of the world, such as in Asian contexts, and that autonomy could be perceived positively by any language learner across the globe (Schmenk, 2005). This would indicate that there should not be any discernible cultural influence on
learner autonomy, in other words, that a learner’s cultural background is an arbitrary factor of autonomous learning. However, researchers such as Oxford, Holloway and Horton-Murillo (1992) provide examples of empirical evidence supporting the influence of culture on learning strategies in general, and also state that many learners who identify with a particular culture often opt for a similar way of learning a foreign language (Oxford et al., 1992). Additionally, empirical studies on the connection between cultural background and learner autonomy as a specific way of learning a language have shown the need for further research on the
differences between autonomous learners from different backgrounds (see Benson, 2007). Benson also demonstrates that previous research has mainly focused on learners with similar backgrounds, such as students who identify with principles from East-Asian cultures (2007). Contrary to the statements from the research explained above, the notion that autonomy is becoming a universally homogeneous capability in language learners from both western and non-western cultures is admitted by Benson (2007) to be rather lacking in empirical support, mainly due to the fact that non-western contexts have not been frequently investigated in research on autonomy.
Ethnicity and cultural identification.
In addition to the statements which support or attempt to disprove the relationship between culture and language learning, most previous research fails to incorporate a clear distinction between the so-called ‘western’ and ‘non-western’ parts of the world. Before a distinction which is relevant for this research on autonomy is made, a description of the concept of culture needs to be provided. Brooks (1968) provides a general but clear view on the concept, by stating that culture represents the social order to which a language learner is attached, and that entails the way in which a learner is expected to think and behave. An extremely important aspect of this is interaction, which influences and leads the learner to adopt these thoughts and behavioral patterns (Brooks, 1968). For this investigation, therefore, cultural identification is more important than the country a particular participant is from; it is assumed that a learner from a particular country does not necessarily identify with the cultural principles of this country, but possibly identifies with the principles of another culture. As is mentioned by Altugan (2015), cultural identity is at least partly dependent on experience and developed competences, which make a learner identify with the principles of a particular culture. For the research instrument which is created for this investigation, cultural
identification is incorporated in questions about participants’ background and ethnic group. Furthermore, in-group ties are tested with various statements to which participants of the research instrument can agree or disagree on a six-point Likert scale. These in-group ties
identify to which extent a participant feels like they belong to an ethnic group (Cameron, 2004). In other words, agreement to these statements indicates a feeling of belonging to a particular ethnic group with similar cultural characteristics, and disagreement indicates that this feeling is not very strong or clearly present, which might mean that a language learner does not feel strongly connected to the principles of this group (see Cameron, 2004).
To avoid confusion about the meaning of ethnicity among subjects, a definition as extracted from the elaborate consideration of the concept by Nagel (1994) is used. This definition of ethnicity entails the set of cultural characteristics and values which are shared by a particular group of people within a country and that this group identifies with (Nagel, 1994). The author incorporates many aspects in her definition of ethnicity, specifically focusing on identification with a collective cultural identity. Due to this focus on cultural identification, the definition as drawn up by Nagel (1994) lends itself very well to this investigation:
Although ethnicity is considered to be a set of values of a particular group of people within a country, it is not excluded that someone identifies with the values of a group that falls outside the national boundary of this individual’s native or home country. For this study, it is possible that a person identifies with either those values of the majority of people of a particular culture, albeit a group of people in their own country or a different one, or that they identify with the values of a certain minority: An example would be that a native-Dutch language learner can identify more with the ‘Frisian’ minority culture than with the more national or supranational ‘Dutch’ ethnicity or culture.
Western and non-western cultures.
Although Kumaravadivelu (2003) strongly states that research on linguistic behavior, and supposedly for that matter also behavior towards autonomy, cannot rely on differences
between the cultures of language learners, the author fails to make the distinction between the culture of the country that a particular learner is from and the culture with which a learner identifies. In other words, the culture of a learner’s native or home country can indeed not be automatically assumed to be of clear influence on behavior as a learner can identify with other, perhaps contrasting cultures. When looking at cultural identification as opposed to cultural background, the possible existence of differences between learners who identify with contrasting cultures cannot be so easily overlooked as is done by Kumaravadivelu (2003). As most researchers of linguistics, and of learner autonomy in particular, seem to look at cultural background more than cultural identification, research thus far has often homogenised
language learners according to their geographical environment (Schmenk, 2008). It is therefore important to keep in mind that the current investigation is not interested in the
cultural background (i.e., culture of the country of origin or residence) of language learners, but much more in cultural identification, in this case identification with either western or non-western cultural principles.
In order to provide a solid reason for separating western and non-western learners,
therefore, several findings on cultural identification in Western Europe and East-Asia need to be discussed. Firstly, the historically western-based nature of autonomy leads to the choice to distinguish between western and non-western cultures. Additionally, according to research on culture and perception, differences in the cultures with which either European, excluding Eastern Europeans, or East-Asian individuals identify, represent varying general patterns in the way these individuals think and perceive. These patterns are generally expressed by a greater independence in language learning among those who identify with western cultures (Varnum, Grossmann, Kitayama & Nisbett, 2010). Their study, which includes the empirical findings of many other investigations on cultural identification and the way individuals think, is the foundation of the division of cultures which is chosen for the present study: Those individuals who identify with Western-European cultures are categorised as western learners, and those who identify with East-Asian cultures are categorised as non-western learners (Varnum et al., 2010). Moreover, Sakai, Takagi and Chu (2010) state that cultural norms of East-Asian language learners often prevent these learners from wanting to be independent in their learning process, whereas Western-European learners are more frequently encouraged to take an individual approach in this process (Sakai, Takagi & Chu, 2010). Despite the fact that this statement already supports the division adopted in this investigation between Western-European and East-Asian cultures, more information on language learning autonomy in the western and non-western world is elaborated upon in the next section of this paper.
The division for this investigation is further explained by discerning specific western and non-western cultures with which learners can identify, and which comprise the groups of western and non-western cultures for this investigation when taken together. The complete categorisation can be found in Table 1. This categorisation, consisting mainly of cultures which appear to be geographically located in western Europe, is taken from Paldam (2002), who identifies a group of western cultures on the basis of the overlap between the behavioral patterns of individuals who identify with these cultures. This categorisation is supported by statements made by Huntington (1996), who focuses on the cohesion between individuals who identify with these cultures. On the opposing side, Chinese, Japanese, South Korean, Taiwanese, Singaporean and the culture associated with Hong Kong are used for those language learners who identify with East-Asian (i.e., non-western) cultures, as the
characteristics of these cultures belong to a so-called general “East-Asian identity” (Huat, 2004, p. 200). The author states that this East-Asian culture has emerged from fading boundaries among nationalities, and that the cultures of the countries pertaining to this East-Asian identity have become very coherent and similar (Huat, 2004).
Table 1. Categorisation of Cultures Across Western and Non-western Principles
Cultural group Cultures
Western – Western Europe Austrian, Belgian, Danish, Finnish, French, German,
Icelandic, Irish, Italian, Luxembourgish, Norwegian, Swedish, Swiss, British
Non-western – East Asia Chinese, Hong Kongese, Japanese, Singaporean, South Korean, Taiwanese
Regarding the western category, North-American cultures are excluded as countries in this part of the world do not share tendencies toward learner autonomy with Western Europe (e.g., Benson, 2007; Oxford, 1996). In the questionnaire for this study, learners who identify as Australian, New Zealander, Portuguese or Spanish are excluded for the possible identification of these learners with other cultural groups that do not share characteristics with the cultures used in this investigation (Paldam, 2002). For example, Paldam states that the Portuguese and Spanish cultures have many similarities with Latin-American cultures, which makes them less identical to the cultures shown in Table 1 (2002). Finally, The Netherlands are excluded as this investigation looks into autonomy perception among language learners who are not Dutch and who are students at Dutch universities learning the Dutch language; it is expected that these learners cannot yet be members of the Dutch cultural group due to the fact that a limited proficiency level often indicates the absence of acculturation into a cultural group, or at least a lack of experience communicating with members of this group and therefore insufficient time to have acculturated (Barker & Galasinski, 2001). In general, any culture that has not been categorised as shown in Table 1 will not be incorporated in the investigation, as there is no sufficient evidence for cultural similarity with other cultures of the same group. Looking at the cultures that have been incorporated in Table 1, it is important to keep in mind that
identification with these cultures is investigated in this study, as opposed to country of origin. Language learning autonomy in the western and non-western world.
Apart from the western origins of autonomy, language learning autonomy in particular is also greatly promoted in western cultures: The Council of Europe of the European Union
promotes autonomy in language learning by, for example, attempting to implement the so-called European Language Portfolio (ELP) in foreign language teaching practices throughout the countries in the European Union, and by raising awareness on the benefits of autonomy (Little, 2004; Ushioda, 2006). According to Little (2002), the Council of Europe has had a clear vision on language learning and teaching since the last decades, of which learner
autonomy is an essential part. By promoting learners’ critical awareness, autoregulation of the learning process and learning goals, and auto-evaluation of the learning process, the Council of Europe continuously seeks to increase the degree to which language learners in the western world learn languages autonomously (Little, 2002). Additionally, nine of the fourteen cultures mentioned in Table 1 of Western-European cultures have been an active part of the promotion of the ELP (Little, 2002). Representatives of these nine and the other five cultures have played an active part in the project of the European Union revolving around the ELP (Little, 2011). It seems as though the many efforts of the European Union to increase language learner autonomy have increased the embeddedness of autonomy in the transgovernmental and transnational culture of Western Europe. In addition to promoting autonomy in western cultures, it has already been explained that the origins of taking responsibility for one’s own learning process traces back to the Enlightenment period (Schmenk, 2008). The concept of learner sovereignty and making decisions which benefit the personal learning process originated in Europe and seem to remain embedded in Western-European cultures, where thinking critically is still valued as an aspect of the aforementioned responsibility for one’s own learning process (Schmenk, 2008).
Looking at the East-Asian cultures with which the non-western learners in this study identify themselves, autonomy does appear to be a part of language learning practices in countries such as Korea, Singapore and China: Supporting a more individual responsibility for personal growth and decision making has been proven to benefit learners in East-Asian
contexts in areas such as satisfaction, interest in what is learned and becoming more
competent learners (e.g., Jang, Reeve, Ryan & Kim, 2009; Lim & Wang, 2009; Zhou, Ma & Deci, 2009). It is, however, evident that autonomy is not promoted or embedded in East-Asian cultures to the same extent as it is in western cultures, and the small amount of research which has been conducted on autonomy in East Asia shows how much less common it is in cultures in this part of the world (Hardré et al., 2006). Additionally, as opposed to the ELP and
language learning autonomy in western cultures, studies such as the one on Singaporean learners by Lim and Wang (2009) have focused on learner autonomy in general instead of autonomy in the language learning process, which has been the subject of even less research
in East-Asian cultures and therefore shows that autonomy in language learning is not largely promoted in these cultures. In general, comments made by Sakai, Takagi and Chu (2010) seem to summarise a great part of the differences between autonomy in Western-European and East-Asian cultures: Non-western learners have a desire to increase their autonomy in the learning process and to develop a more individual learning style, but for East-Asian learners “... educational and behavioral norms [...] simply discourage their autonomy” (Sakai, Takagi & Chu, 2010, p. 12). Although the authors, from Japanese and Chinese descent themselves, appear to follow the line of multiple East-Asian researchers who make a personal case for more autonomy in East-Asian cultures, this discouragement of autonomy due to cultural norms does appear to be supported by empirical findings (Chan, 2001). Additionally, it is stated that these cultural norms may not only slow down the development of autonomous language learning in East Asia simply due to particular characteristics of these cultural
beliefs, such as a high degree of respect for the knowledge of the teacher, but that a reluctance to adhere to aspects of critical reflection on the personal learning process, which have been explained to be valued in western cultures, could be a reason to distrust autonomy as well (Littlewood, 1999). Since this indicates differences in the perception of autonomy between learners from western and learners from non-western cultures, the next section contains an explanation of the way in which the research instrument of this investigation takes possibly different perceptions of autonomy into consideration when looking at the description of the concept.
Another example of differences between western and non-western language learners is that language learning autonomy may be modelled distinctively in the minds of learners from different cultures, which makes the formulation of the concept a delicate task (Schmenk, 2008). For example, Littlewood (1999) states that traces from the East-Asian cultural beliefs of collectivism, acceptance of power differences and the perception of success can be of influence on the approaches to various learning strategies of students in East Asia. For this reason, a general explanation of autonomy which is applicable to learners from all contexts, in this case Western European and East-Asian, is required. This explanation cannot be directed towards students from Western Europe only and needs to take various educational norms into consideration (Littlewood, 1999). For this reason, the instrument which is developed for this investigation only uses descriptions of the concept of autonomy and as it has been elaborately described by Benson (1997) and previously in this study, as opposed to the term ‘autonomy’ itself. By using these descriptions, all aspects of autonomy are made clear and possible
misinterpretations of the concept by learners from either of the two investigated cultural groups are avoided.
The Dynamics of Language Acquisition and Influences on Learning Strategies It would not be safe to assume that culture is the sole factor that influences a learner’s choice of learning strategy, in particular the choice to learn a language autonomously. Rather, additional factors such as age and gender can also change learners’ experiences and choice of strategy: According to the Dynamic Systems Theory (DST) approach to linguistics, an individual’s language use is dependent on multiple factors which continuously change this individual’s linguistic system (De Bot, Lowie & Verspoor, 2005). This theory is strongly connected to Dynamic Usage-based (DUB) principles, which entail that not only culture, but especially a person’s experiences need to be taken into consideration when looking at
language acquisition, and that other factors such as gender, age and environment can also influence the success of language acquisition in general (Verspoor, Schmid & Xu, 2012). Looking at the choice for a specific language learning strategy, Oxford (1989) states that previous experiences with language acquisition are the biggest influence found so far on a learner’s current learning strategies (Oxford, 1989). Additionally, behavioral patterns in language learning are generally also influenced by these factors (Kumaravadivelu, 2003). In other words, language acquisition in general, and specifically the success of language acquisition and behavioral patterns in learning a language, are evidently affected by these factors (De Bot et al., 2005).
Although an influence of different factors on language acquisition in general has clearly been found in previous research, studies on the influence of age, gender and other factors have provided mixed results when looking at learning strategies in particular, and even studies on previous experiences have not been entirely unilateral (see Oxford, 1989). For example, no clear presumptions have been made in previous research on the possible influence of age or previous experiences on autonomous learning in particular. Concerning the effects of gender on learner autonomy, one study by Varol and Yilmaz (2010) resulted in similar results for male and female students, but merely looked at elementary school learners. The investigation did implement a similar instrument as the current study in the form of a Likert scale, but cannot definitively exclude the possibility of varying results between male and female participants in the present investigation as it did not look at a similar population sample. Similar to the expectation in the present study, results from the study by Varol and Yilmaz do indicate that autonomy may be similarly perceived by both genders (2010). Other studies have
hinted towards females being more autonomous learners, but these studies mainly point out that there are some differences in strategy use, whilst refraining from statements that say females are actually more autonomous language learners (see Hong-Nam & Leavell, 2006; Oxford & Nyikos, 1989). Other factors which are mentioned by Oxford (1989) to lead to varying results in studies on choice of learning strategy, and which are tested with different methods that surpass the limits of a research questionnaire and therefore go beyond the scope of the current investigation, are proficiency level and personality traits. These factors could be included and tested reliably in future research on the perception of language learning
autonomy.
Altogether, the principles of DST and DUB theories which predict the influence of
multiple factors on language acquisition and choice of learning strategy cannot be overlooked, and for that reason it is expected that the factors described in this section are a possible
influence on learner autonomy. The focus of this study remains on the influence of cultural identification on autonomy, as culture is specifically stressed as a possibly important factor of autonomous language learning in much previous research, as opposed to the other factors described here. Looking at previous experiences, the statements on the influence of
experience on learning strategies made by Oxford (1989), specifically concerning the fact that many language learners choose different strategies as they become more experienced learners, make for this factor to be incorporated in the present investigation. Additionally, gender will also be investigated as previous results on the differences between male and female learners, such as those in the study by Varol and Yilmaz (2010), are found to be rather undecisive. Regarding the age of participants, the combination of a lack of previous scientific results on age and autonomy (see Oxford, 1989), together with the similar ages of the target audience in the present investigation indicate that this factor can be omitted in the analyses. Future
research might incorporate participants of various age groups, proficiency levels and other characteristics in autonomy analyses; for this investigation, the perception of autonomy and possible influences on this perception from cultural identification, language learning
experiences and possibly gender are considered sufficient to reach the desired conclusions on the influences of cultural identification on autonomy.
Research Objectives
It has been shown that various researchers make different statements on language learning autonomy. Autonomy is historically a western-based concept, but no proof has been found for a possible difference between western and non-western learners in the way autonomy is
currently perceived in foreign language learning. There are arguments, such as the promotion of autonomy as a learning strategy in non-western cultures and a possible implementation of autonomy in language learning practices across the globe, for a universally homogenous nature of autonomy which would negate the influence of cultural background on autonomous learning, but the only empirical evidence that has been found has mainly been focused on learners with similar backgrounds as opposed to different backgrounds. The possible western-based nature of language learner autonomy, together with the frequently mentioned statement that an investigation into the possible ethnocentric basis of autonomy, make for this
investigation to look into this ethnocentric basis with the expectation that students who identify with western cultures perceive autonomous language learning more positively than those who identify with non-western cultures.
Additionally, it is still unclear to which extent other possible influences can have an effect on learning strategies and autonomous language learning in particular. Although the focus of this study is cultural identification, factors such as experiences with language learning and gender will not be overlooked in this investigation, as these factors possibly influence the perception of learner autonomy as well. For previous experiences, the differences between experienced and less-experienced learners regarding choice of learning strategies lead to the hypothesis that experience with autonomy is a positive influence on the current perception of autonomy. For gender, no significant differences are expected as previous investigations have not led to clear differences in choice of learning strategy between male and female learners.
In order to test language learning autonomy in students from various cultural
backgrounds, this exploratory study aims to create a tool to assess autonomy perception for the three levels on which it occurs taken together: Technical, political and psychological. Additionally, the tool created to test autonomy is supposed to provide a foundation for future research on learner autonomy, and could therefore be used in later investigations on the possible influence of cultural identification on learner autonomy. The present investigation is therefore meant as an initiatory study to find this answer. The alpha level for the analyses of the results of this research instrument is set at .05, as is usual for investigations in social sciences (Lowie & Seton, 2013).
Method
This section provides a description of the participants, materials, procedures and design of the investigation. As is mentioned before, this investigation aims to clarify whether there is a relationship between cultural identification and the perception of language learning autonomy, for which a new research instrument in the form of a questionnaire is created. This
questionnaire was drawn up for students at Dutch universities who identify with western and non-western cultures and who learn Dutch as a foreign language. This questionnaire was administered in a pilot version and a final version, which are both explained in this section concerning the administration of these versions. Only those aspects of the pilot version which different in content from the final version are added to this section. The final version of the questionnaire, which was visually adapted from its online version to increase readability, can be found in Appendix A. An important note is that this study initially intended to investigate autonomous language learning among non-Dutch students at a Dutch university in the city of Groningen, but that this was changed shortly after the research instrument was posted online in order to reach more participants. This means that non-Dutch students at any Dutch
university were able to fill out the questionnaire, as long as they were students and learning Dutch.
A pilot study was conducted to test the validity and reliability of the developed
instrument. Ten non-Dutch participants who are students at a university in the Dutch city of Groningen and who are learning Dutch were asked to fill out a pilot questionnaire and provide comments on the content of the questions, the clarity of the formulation of these questions, the clarity of the instructions, the length of the questionnaire, and other aspects which could be improved according to these participants. Additionally, a focus discussion with the subjects of the pilot questionnaire was held to discuss any other parts of the questionnaire which would benefit from further clarification. Comments from the participants of the pilot study indicated that a small number of questions needed clearer formulation, and that the concept of ethnicity, which represents cultural identification in the current investigation, was not clear enough. For these reasons, grammatical and lexical changes were made in the questions about previous learning experiences and one question about cultural identification: For previous experiences, two items on the Likert scale were changed from ‘does not apply to me so much’ to ‘does not apply to me that much’ and ‘applies to me quite much’ to ‘applies to me quite a lot’, and for cultural identification the word ‘fabricated’ was changed to ‘artificial’ in the question about the Dutch language being learned too artificially in an instructed setting. Regarding the concept of ethnicity, an explanation of the meaning of ethnicity for this investigation was
added to avoid confusion in the final questionnaire. This definition is extracted from Nagel (1994) and explained previously in this study. General information from the results of this pilot study is added to the explanation on the content of the final questionnaire version in order to justify the choice of particular questions in the research instrument. Additionally, reliability analyses were conducted on the results of the pilot study from all statements with answers on a Likert scale (see Cronbach, 1951; see R Core Team, 2015). These analyses were run on all questions about cultural identification, previous experiences with autonomy and autonomy perception which are incorporated in the questionnaire and explained further in this section. The results of these analyses also accompany the description of the final
questionnaire version in this section. All analyses in the pilot study are done using R (version 3.5.0) (R Core Team, 2013).
Subjects
In total, 539 people filled out the questionnaire through the online platform Qualtrics (see Qualtrics, 2018). Of this total number, 223 answers had to be removed: Most of these
removed answers came from people who started the questionnaire but were either not students at a Dutch university or who were not learning Dutch, and some did not answer enough questions to be able to be incorporated in this investigation. An example of this would be someone who is enrolled at a Dutch university and who is learning Dutch, but who only answered a small number of background questions and did not complete any questions about cultural identification, previous learning experiences and autonomy perception. It is unknown why these people did not complete the questionnaire. Additionally, 4 of these 223 removed answers came from people who did not grant permission for their answers to be used in research.
Deleting these 223 excessively incomplete answers leaves a total of 316 non-Dutch
students who are enrolled in a Dutch university and who are learning Dutch (mean age = 25.3,
SD = 6.1), and can therefore be used in this investigation. Firstly, looking at cultural
identification, the difference between the cultures of participants’ native or home countries and the cultures with which participants identified is apparent in the questionnaire answers: Both western and non-western participants occasionally appeared to identify with other cultures than those of their native countries. For example, a participant from the United Kingdom stated to identify as Eritrean and a Chinese participant identified as Mongolian. Most participants who identified with other cultures, however, opted to state that they felt part of either a bigger culture than that of their native country, or a smaller one. Examples are the
five Chinese participants who identified as Asian or as citizens of the world, five participants from several European countries who identify as Europeans, or the eight participants from Germany who identified with the culture of a small part of Germany only, such as the North. All of these participants are naturally excluded from the analyses on cultural identification of this investigation, as there is no clear support for these examples of cultural identification to belong to either of the cultural groups used in this study. For example, it is unsure whether a learner who identifies as northern German would share cultural characteristics with a learner who identifies as German in general; as opposed to the German culture in general, previous research has not indicated cultures such as the northern German one to share these western cultural characteristics (see Paldam, 2002).
Other participants who were excluded from this investigation are a Chinese person who did not believe to have an ethnicity, four British participants who identified as white people and one Chinese person who identified as a yellow person, one Italian participant who identified as Mediterranean, one German participant who identified as Caucasian and one British person who identified as Afro-Caribbean. These participants were excluded as it could not be safely assumed that they identified with one of the western or non-western cultures as shown in Table 1 of this study. These answers for cultural identification raise questions for future research, for example regarding the German participants who apparently, to some degree, tend to feel part of a rather small cultural group as opposed to the German culture in its entirety, which is used in the analyses here.
After the exclusion of the participants mentioned above, and all other participants who did not identify with one of the western or non-western cultures from Table 1, a total of 64
participants remains who identify with one of the investigated western cultures. 60 participants of this group of western learners averagely agreed with the statements on in-group ties from Cameron (2004) (mean age = 24.0, SD = 5.9, 11 males), and therefore met the requirements for scores on cultural identification as represented in these statements. In other words, it can be safely assumed that these 60 participants feel as if they belong to the group of people who identify with western cultures, whereas this cannot be safely assumed for those 4 learners who averagely disagreed with the statements on in-group ties. Additionally, 22 participants identify with a non-western culture (mean age = 25.1, SD = 5.0, 11 males). All of these participants averagely agreed with the statements on in-group ties and therefore met the requirements for cultural identification scores. Looking at the specific cultures with which these 82 western and non-western participants identify, it is noticeable that all cultures as shown in Table 1 are represented by the subjects of this study, with the exception of Austrian,
Belgian, Icelandic, Luxembourgish and Swedish as western cultures and Singaporean as a non-western culture.
Regarding the other factors under investigation, 3 participants who identify with western cultures and 2 participants who identify with non-western cultures have not learned any other languages than their native tongue and Dutch, and can therefore not be used in the analyses concerning the effects of previous language learning experiences on the perception of autonomy. Additionally, 3 learners who identify with western cultures did not want to share information on their gender, and will be left out of the analyses on the effects of gender on autonomy perception. Overall, the numbers described before are too small for analyses on previous experiences and gender within both groups of cultural identification (i.e., western and non-western), although an interaction analysis of, for example, gender and cultural identification for western and non-western learners would be desired. In this investigation, however, this led to the incorporation of learners from all cultures in the analyses on previous experiences and gender, which means that those learners of Dutch who are not Dutch and who not only identify with a western or non-western culture, but also any other culture, will be incorporated in the analyses on these two factors. This leads to a total number of 299 participants, after having excluded 17 participants who have not learned another language than Dutch and their native tongue, who can be used in the analyses on previous experiences. Additionally, 311 participants can be used in the analyses on gender, after having excluded 5 participants who did not want to state their gender or who do not identify as male or female. 108 of these 311 participants are male. Finally, it is apparent that the average age of all participants is fairly equally distributed and does not leave room for analyses on different age groups for autonomy perception.
Materials
The research instrument consists of an introduction, a part on background and cultural identification, a part on previous experiences with language learning, two parts on the perception of autonomous language learning, and final questions about personal
characteristics. In this subsection, more information is provided on the content of the newly developed instrument and its foundation in previous research on identity and learner
autonomy. Firstly, the introduction of the questionnaire states that this investigation looks into experiences of students in the Dutch city of Groningen with autonomous language learning. Additionally, it contains information on the expected time it will take to fill out the
background, age and gender, does not need to be provided, and that the questionnaire is only meant for non-Dutch students who learn Dutch in any way possible.
The introduction is immediately followed by the part on background and cultural
identification, which starts with the questions whether a participant is a student at a university in Groningen and whether they are currently learning Dutch. If either of these questions was answered negatively, a participant would automatically be directed to a page which stated that the questionnaire could not be filled out any further. Wherever the research instrument would be promoted outside of the city of Groningen, a comment was included that participants could answer positively to the question whether they are a student at a university in Groningen as long as they were enrolled in any university in The Netherlands. It is additionally mentioned in the questionnaire that the ways in which a participant learns Dutch, such as through a course, does not matter. This part of the questionnaire then asks participants about their home country. These questions are incorporated to ensure that the analyses of this study are
conducted on a homogeneous group of participants, namely non-Dutch students who are learning Dutch. Then, participants were asked about their ethnicity. It is explained in the research instrument that ethnicity is considered to be “the set of cultural characteristics and values of a particular group within a country” (see Appendix A), and this explanation is accompanied by the same example as that in the background section of this paper. The answer to this question represents the culture a participant identifies with. Finally, four statements on in-group ties as part of cultural identification, as extracted and adapted from Cameron (2004), were incorporated to determine whether a participant actually identified with the ethnicity they had said to have in the previous question. These statements asked participants whether they had a lot in common with other members of their ethnic group, whether they feel strong ties with these other members, whether they find it easy to forms bonds with these members and whether they feel connected to them. Answers to these statements were possible in the form of a six-point Likert scale of agreement, where three points indicate degrees of disagreement and three points indicate degrees of agreement. Participants in the present investigation are expected to identify with a particular culture if they averagely answered in agreement with these statements, and therefore feel as if they are a “part of” a particular group which identifies with that culture (Cameron, 2004, p. 243).
Regarding the results on this part of the questionnaire from the pilot study, reliability analyses for cultural identification (4 items; α = 0.6), showed that three of four items were sufficiently correlated (see Cronbach, 1951). This fourth item, asking about connectedness with other members of the ethnic group (r = 0.13), was not excluded from the final
questionnaire, however, as it is explained before that the concept of ethnicity was not clear to the participants in the pilot study. It was expected that all analyses on cultural identification would be normal after the inclusion of an explanation of ethnicity in the final questionnaire. Additionally, the statement on connectedness was taken from the existing questionnaire on in-group ties by Cameron (2004) that was used for the research instrument of this investigation. Overall results of analyses on homogeneity of the answers given were normal (see Cronbach, 1951).
In the second part of the questionnaire, participants were asked about their general history concerning language learning and autonomy in the language learning process. In this part, questions were asked about participants’ attitudes and behavior towards language learning autonomy in their language learning history. These questions asked participants whether they are used to thinking critically in the language learning process, whether they are used to setting personal learning goals and whether they are used to choosing materials for language learning themselves. These questions were formulated according to information on autonomy as provided by Jiménez Raya, Martos Ramos and Tassinari (2017a; 2017b), and by Arnold and Fonseca-Mora (2017). In addition to this, more specific statements about the application of learning strategies in participants’ language learning history were incorporated. These statements were based on an existing questionnaire on learning strategies by learners of Dutch, administered annually by an independent language instruction organisation related to the university in the city of Groningen (Language Centre University of Groningen, n.d.). These questions concerned the use of technological tools, talking to native speakers in the target language, using materials which are not provided by a language instructor, generally increasing the personal proficiency level without formal instruction, and using visual and audial materials in their language learning history. All eight questions, in the form of statements, were accompanied by a six-point Likert scale which represented the degree to which a statement applied to the participant. This part of the questionnaire represents the influencing factor of previous experiences on a learner’s choice of learning strategy and on autonomy in particular.
In the pilot study, previous experiences with autonomous language learning were investigated using the same analyses as those for cultural identification. Two of eight questions turned out to be inconsistently worded and insufficiently correlated. These
questions, about thinking critically (r = 0.10) and using materials which are not provided by a teacher (r = 0.06), were removed and therefore not incorporated in the final instrument. Feedback suggested that these questions confused participants in the sense that thinking
critically turned out to be an unclear concept and that it was unclear when materials are provided by an instructor or not. The six other items were not excluded (α = 0.68)
Parts three and four of the instrument asked participants about their attitudes towards autonomy, in particular concerning Dutch foreign language learning. Questions in this part were formulated according to the three versions of autonomy as described by Benson (1997) and elaborated upon by Benson and Voller (2014). In total, all questions on technical, psychological and political autonomy were incorporated in the form of statements to which answers on a six-point Likert scale of agreement were asked (see Appendix A). Agreement with a statement would indicate a positive attitude towards language learning autonomy. Seven statements on attitudes towards autonomy were reversed in the sense that a strong agreement would not indicate autonomy, but rather a lack thereof. This was done to have a greater variety of statements and to distract participants from encountering a particular pattern in the statements. As is mentioned by Benson and Voller (2014), there is significant overlap between the aspects of the threefold subdivision into political, psychological and technical autonomy, and looking at them separately would be difficult and create confusion. For this reason, questions on the three versions were taken together and randomised. These questions were distributed among two parts in the questionnaire in order to minimise the size of one single part in the instrument. In total, these questions represent the perception of autonomy of the participants in this study. Examples of political, psychological and technical autonomy can be found in (1), (2) and (3), respectively. It was not mentioned in the questionnaire whether a statement mostly represented political, psychological or technical autonomy.
(1) Learning Dutch in an instructed setting means learning a type of Dutch that is too artificial.
(2) It is my responsibility to improve my Dutch.
(3) I possess the necessary skills to improve my Dutch without instruction.
Regarding the pilot study, reliability analyses were conducted on the statements about autonomy perception. As it has been explained, all twenty-four statements on technical, psychological and political autonomy were taken together as overlap between these versions of autonomy excludes the possibility for separate analyses. An initial problem in these analyses was missing data, which occurred once in three different statements. This problem was overcome by using single imputation in the form of mean substitution (see Acock, 2005). This is, the mean answer score of that question by all other participants as opposed to the
mean answer score of the other answers of that particular participant was used. The mean score which substituted an absent answer is not expected to deviate from the answer that would have been provided, as there are no clear indications for the sample of this pilot study to be extraordinarily diverse and thus to lead to very different answers given. Additionally, as each of the three statements was only left blank by one participant, the risk of a lack in variance within the sample tested is minimal (Acock, 2005). For the twenty-four statements on autonomy perception, both positively worded statements (i.e., a high score indicates a positive perception of autonomy) and reversed statements (i.e., a high score indicates a lack of positive perception of autonomy) were analysed. For positively worded statements, four questions about the availability of tools to learn autonomously (r = 0.19), the direction of the personal learning process (r = 0.00), the desire to learn independently (r = 0.22) and the benefits of independent learning for personal gains (r = 0.16) were deleted to increase item consistency and correlation. For reversed statements, one item concerning a lack of structure in language learning without instruction (r = 0.27) was removed. After deletion, nineteen items for the perception of autonomy were kept (α = 0.85). Political, psychological and technical autonomy were initially represented by eight statements for each of these versions, and after deleting unreliable statements by six, seven and six statements, respectively.
Finally, a part on additional information of the participants was incorporated in the questionnaire. Questions about the native language or languages, other previously learned languages apart from the native languages and Dutch, age and gender were incorporated to create a clearer overview of the participants’ background. Participants were asked about their native language and other previously learned language in order to gather information for future research on the influence of these languages on the process of learning Dutch as a foreign language. Regarding the current investigation, it has been previously explained that age is not taken into consideration as an influencing factor on autonomy perception. Gender is incorporated in the analyses of this investigation, although differences between male and female learners of Dutch are not expected. No analyses on background information such as gender were conducted in the pilot investigation; these analyses were run on the results of the final questionnaire and are discussed later in this study. Finally, consent to use all answers for research purposes was asked.
Procedures
In this section, the entire process for the administration of the instrument to measure cultural identification and autonomy perception is explained. The final version of the research
instrument is a result from background information gathered from previous research on these topics. The final questionnaire consists of questions on background information, four
statements on cultural identification, six statements about previous experiences with language learning, and nineteen statements about autonomy perception. A visually adapted version, which is kept intact content-wise, can be found in Appendix A. The platform used for the final version of the instrument was Qualtrics (see Qualtrics, 2018), which creates a webpage on which participants can answer all questions of the questionnaire. The platform also creates an online overview on which all participants’ answers are automatically gathered and shown to the administrator of the questionnaire.
As is mentioned before, the final version still contained the question whether a participant was enrolled in a university in Groningen, and participants from other parts of the country were informed that they could answer this question positively as long as they were enrolled in any university in The Netherlands. Participants were also informed that they could answer the question whether they are learning Dutch positively if they were learning Dutch in any way, such as through a course, by practicing for themselves, by using platforms on, for example, social media or in other ways. This was done to make sure that not only those learners who learn Dutch through a course would fill out the questionnaire. The comments described here, together with the research instrument itself, were published and promoted through social media pages, such as those specifically designed for language learners, and by sending e-mails to language learners through the networks of Dutch universities. The questionnaire was also promoted to learners of Dutch in courses at Dutch universities. Examples of those universities at which the questionnaire was administered are various universities in the cities of Groningen, Delft, Nijmegen and Rotterdam in The Netherlands.
Design and Analyses
In order to test the research hypotheses of this investigation, various statistical analyses were conducted and are elaborated upon in the next section of this paper. Firstly, the main hypothesis of an expected influence of cultural identification on language learning autonomy is tested by conducting a Wilcoxon Signed-Rank test to test whether learners who identify with western cultures had higher autonomy perception scores than learners who identify with non-western cultures (see R Core Team, 2016a). This test is accompanied by an explanation of the general tendency of participants’ answers concerning their cultural identification and perception of autonomy. Secondly, it is tested whether previous experiences with autonomous language learning indicate a higher autonomy perception score (i.e., a generally more positive
perception of the concept of autonomous language learning). This is tested with a Pearson r correlation analysis (see R Core Team, 2016b). Finally, the effects of gender on autonomy perception are tested with a Wilcoxon Signed-Rank test to compare autonomy scores for male and female participants (see R Core Team, 2016a). All statistical analyses to test the research hypotheses of this investigation are done using R (version 3.5.0) (R Core Team, 2013).
Results
In this section, the possible influences of cultural identification, previous experiences with language learning and gender on the perception of language learning autonomy are
investigated. Statistical analyses are conducted to investigate whether learners who identify with western cultures experience language learning autonomy in a more positive way. It is also tested whether experience with autonomous language learning indicates a more positive perception of this concept. Finally, it is tested whether males or females perceive autonomy similarly.
Similar to the pilot study, imputation in the form of mean substitution was used in the final version to fill in missing answers for the statements on cultural identification, previous experiences and the perception of autonomy in learning learning (see Acock, 2005). The statements belonging to these three topics were all filled in by the total number of 316
participants. For cultural identification, which counts four statements, one answer from a total of 1264 was subject to imputation. This answer is not used in the analyses of this
investigation, however, as it came from a participant who does not identify with one of the western or non-western cultures. For previous experiences, which counts six statements, one answer from a total of 1896 was subject to imputation. Finally, for the perception of
autonomy, which counts nineteen statements, 27 answers from a total of 6004 were subject to imputation, and these answers were divided among eighteen different participants and thirteen different statements.
In order to test whether the statements used in the questionnaire are effectively suited for the analyses of this study, Cronbach’s alpha reliability analyses such as those conducted in the pilot study were run again on all of these statements (see Cronbach, 1951; see R Core Team, 2015). These analyses were run on the answers given by all 316 participants for statements on in-group ties (4 items; α = 0.88), by all 299 participants who had learned another language than Dutch and their native tongue and therefore have previous language learning experiences (6 items; α = 0.66), and on all 316 participants for the statements on the perception of
autonomy (19 items; α = 0.81). No statements on in-group ties as a representation of cultural
identification, on previous experiences with language learning or on the perception of
autonomy in language learning were eliminated, as internal consistency and correlation scores for the statements of each of these three factors would not have increased and as no
