Introduction

The investigation of language learning autonomy, that is, exploring the ways in which students take responsibility for their own learning processes (Benson, 2011), is important since technological advancements of the twenty-first century increasingly allow for students to learn and use English outside the classroom. As the ways students use and learn English through technology might differ between boys and girls, we think it is important to conduct focused studies that investigate gender-based differences. Hence, the aim of this study is to explore one particular aspect of language learning autonomy: students’ autonomous use of technology with special emphasis on gender-related differences. This research endeavor is important in a context like Hungary, a monolingual country, where learners of an additional language must be autonomous in order to find opportunities to use the language they learn and thus improve their proficiency. It is not only important to map gender-based differences in students’ autonomous use of technology but also to study what predicts male and female students’ autonomous use of technology. To do so, we have chosen to investigate different individual difference variables, such as motivation, self-efficacy and emotions. Previous studies have shown that motivation is an important predictor of learner autonomy, an example of which is autonomous use of technology (Kormos & Csizér, 2014); therefore, we included students’ motivated learning behavior, specifically, the effort they intend to invest in language learning, in our analysis. Additionally, we hypothesized that self-efficacy beliefs may also be important when it comes to learner autonomy because, on the one hand, it appears intuitively plausible that learners should be convinced of their own abilities to perform well in order to be autonomous, and on the other, empirical evidence has already been found linking the two constructs in both quantitative and qualitative studies (Csizér et al., 2021; Han & Reinhardt, 2022; Honarzad & Rassaei, 2019; Tilfarlioglu & Ciftci, 2011). Finally, we were interested in the extent to which language learning emotions influence learner autonomy since empirical research surveying the relationship between these two constructs is scarce despite the fact that the investigation of emotions is an increasingly important area in applied linguistics. The crucial role of emotions is highlighted, as it is becoming increasingly clear that cognition and emotions are intricately intertwined and that positive emotions are likely to be important contributors to the cognitive process of language learning (Dewaele et al., 2022a; Fredrickson, 2003, 2008; Li & Wei, 2022; Shao et al., 2020).

Our study is novel in several respects and aims to fill various research gaps. First, examining what students do to learn a language on their own is especially informative in a context where the public education system has often been criticized for not being able to fulfill its role in teaching students a second language. This is the case even though the total number of language lessons in Hungarian public education can be considered high among countries of the EU (Öveges & Csizér, 2018). Second, mapping the interplay of various individual difference variables is important to enhance our theoretical knowledge of language learning processes, and although such studies already exist (Dewaele et al., 2022a; Li & Wei, 2022; Shao et al., 2020), ours is novel since it measures a wider array of emotions than what is usually found in earlier research endeavors. Third, when gender-related differences are identified, these findings might provide assistance in customizing instruction to fit the needs of individual learners better, thereby enhancing differentiated instruction.

Theoretical Background of the Study

Language Learning Autonomy

Learner autonomy has been defined in multiple ways, but Holec’s (1981) classic definition, which includes “the ability to take charge of one’s learning” (p. 3) still provides a solid foundation for the idea that language learning autonomy involves taking responsibility for learning at various stages of this complex process (Benson, 2011), encompassing language use as well (Illés, 2012, 2019). In the current study, we measured one aspect of learner autonomy in second language learning, which was students’ autonomous use of technology that has been operationalized as students’ taking responsibility for learning English through their independent use of technology. Language learning autonomy has been researched from various perspectives in recent years, and its relationship with L2 motivation has been examined in Hungary (Csizér & Kormos, 2014; Kormos & Csizér, 2014), similarly to its tendency to decrease over the years students spend in public education in this country (Albert et al, 2018a, 2018b). However, there have been few studies investigating the relationship between language learning autonomy and gender. In a Turkish context, Varol and Yilmaz (2010) found both similarities and differences in terms of autonomous learning activities inside and outside the classroom, while Şakrak-Ekin and Balçıkanlı (2019) concluded that female university learners demonstrated greater autonomy than male students. Similar results were observed in the Chinese context, where females outperformed males when learner autonomy was measured (Zhao & Chen, 2014).

Second Language Learning Motivation

The investigation of the role of motivation in the second language learning process has a long history. It has been convincingly argued that motivation is one of the most important individual difference variables (e.g., Dörnyei & Ushioda, 2021), as it encompasses both the effort and persistence students invest in learning the L2. We have labelled the construct that measures effort and persistence as students’ motivated learning behavior, which is in line with previous investigations in the Hungarian context (Csizér & Kormos, 2012; Kormos & Csizér, 2008). The extensive field of second language learning motivation research is evolving in different directions including theoretical advancements, filling contextual gaps, and discovering new ways of exploring situation-specific and individual profiles of motivated learners (Lamb et al., 2019). Still when it comes to investigating gender-based differences, the number of studies is relatively low, and the results are mixed. The majority of studies, including those conducted in the Hungarian context, have found clear differences with girls reporting higher levels of motivation (Dörnyei et al., 2006; Iwaniec, 2019; Kissau et al., 2010; Öztürk & Gürbüz, 2013; Ryan, 2009; Williams et al., 2002). However, researchers in Sweden were unable to confirm gender-based differences (Henry & Cliffordson, 2013; Sylvén & Thompson, 2015). As Iwaniec (2019) explains, these studies often do not investigate the possible reasons behind gender-based differences, but it is assumed that they “mirror the popular beliefs, which frame language learning as a feminine activity” (p. 139).

Emotions

In the language learning context, Shao et al. (2019) defined emotions as “affective experiences that are directly tied to language learning activities and resulting learning outcomes, a dynamic process which is determined by appraisals of socio-culturally shaped L2 learning tasks” (p. 2). This definition states that emotions serve as a link between the individual’s inner and outer world and emphasizes the interactive dynamic relationships between the two, clearly highlighting the significance of emotions in all human activities. Within the field of second language acquisition, anxiety was the first emotion that received increased attention and targeted research efforts (Horwitz et al., 1986; MacIntyre & Gardner, 1991); however, currently studying language learning enjoyment has become increasingly popular (Dewaele & MacIntyre, 2014; Dewaele et al., 2022b; Zeng, 2021). This trend is most likely connected to the growing influence of positive psychology in our field in general (Derakhshan, 2022; MacIntyre et al., 2016; Mercer & MacIntyre, 2014; Wang et al., 2021) and more specifically to the predictions of the broaden and build theory which claim that the evolutionary significance of positive emotions lies in their ability to encourage exploration of the environment and the simultaneous building of resources (Fredrickson, 2003, 2008). Besides these two prominent emotions, a third one that has been explored extensively in connection with language learning is boredom (Derakhshan et al., 2022; Pawlak et al., 2020, 2021, 2022a, 2022b). However, as regards gender differences, the results are typically mixed. Although female learners have reported higher levels of language learning anxiety as well as enjoyment than males in a number of questionnaire studies (Dewaele & MacIntyre, 2014; Dewaele et al., 2016), such differences could not be confirmed in others (Mierzwa-Kamińska, 2021; Piniel & Zólyomi, 2022). Also, when it comes to boredom, male learners were found to experience this emotion more intensely in the Hungarian context (Albert et al., 2018a, 2018b), while in a Turkish study no gender differences were found (Coşkun & Yüksel, 2022).

Self-efficacy

Bandura (1986, p. 391) defined self-efficacy as “people’s judgments of their capabilities to organize and execute courses of action required to attain designated types of performances”. Although this definition lends itself well to the investigation of language learning processes, self-efficacy, despite its potential importance, has received relatively little emphasis in applied linguistics/language pedagogy research, where the focus was mainly on its relationship with language learning motivation and strategy use (Kim et al., 2015; Piniel & Csizér, 2013). As regards gender differences, Iwaniec’s (2019) study focused on these constructs and found that there was no conclusive evidence, as some studies reported higher levels of self-efficacy for women (e.g., Mills et al., 2007), while other studies found no gender-related differences (Kissau et al., 2010). In her study in Poland, Iwaniec (2019) did not find any meaningful differences between boys and girls concerning their beliefs of being able to learn English. One explanation might be that self-reported self-efficacy data might produce a bias against female participants, as “females tend to underestimate their abilities, whereas males tend to overestimate them” (p. 138). Interestingly, her correlational results were significantly stronger for females between self-efficacy and self-regulation than for males, which might indicate the greater importance of self-efficacy beliefs for female language learners when it comes to regulating their own learning independently.

Methods

Design of the Study and Research Questions

In order to investigate the impact of some individual difference (ID) variables on students’ autonomous use of technology, we conducted a quantitative questionnaire study to be able to infer statistically from the sample to the population. The research questions were as follows:

  1. 1.

    What characterizes the differences between girls and boys concerning second language learning motivation, emotions, and autonomous use of technology?

  2. 2.

    To what extent do second language learning motivation, self-efficacy and emotions impact the autonomous use of technology of boys and girls in Hungarian secondary schools?

Context, Population, Sampling and Sample

As we would like to generalize our findings to the population, it is important to point out here that our population included secondary school students in Hungary. Generally speaking, in the Hungarian three-tiered education system primary school education starts at the age of 6, and it can last for four, six or eight years; therefore, students transfer to secondary schools at the age of 10, 12 or 14. As a result, secondary school education can last for eight, six or four years, with an additional year dedicated to intensive language learning between primary and secondary education in certain cases. Our population consisted of secondary school students studying in the 8 + 4 system, as this is the most frequently occurring setup, and we selected students from years 9 to 12. Concerning the sampling technique, we decided to use quota sampling (Dörnyei, 2007), and first selected secondary schools having the four-year program from three different regions in the country: the capital city, Western Hungary, and Eastern Hungary. The final sample consisted of 1,152 secondary school students (467 boys, 682 girls, 3 with missing data for gender) from three schools from the capital and eight schools from other six towns. The age of the students ranged between 14 and 20 (M = 16, SD = 1.22, 3 missing), and they were all L1 speakers of Hungarian and were learning English at school at the time of the data collection. Their self-reported level of English was between B2 and A1 with an average starting age of 9.2 years (SD = 3.1) of learning the language.

The Instrument

The instrument used in the study aimed to tap into 12 constructs related to students’ autonomy (Csizér & Kormos, 2012), motivation (Csizér & Kormos, 2012; Dörnyei, 2005, 2009; Kormos & Csizér, 2008), emotions (Albert et al., 2021; Pekrun, 2014; Pekrun et al., 2011) and self-efficacy (Albert et al., 2018a, 2018b; Piniel & Csizér, 2013), which were all measured with the help of scales containing 5-point Likert-scale items. Despite the fact that there are a number of reliable scales available for measuring similar constructs (Dewaele & MacIntyre, 2014; Mahmoodzadeh & Khajavy, 2019; Pawlak et al., 2020, 2022a, 2022bs), we opted for developing our own scales for a range of emotions specifically for the Hungarian context. As we measured nine distinct emotions, the use of short scales was necessary, and the reliability of those scales was established via extensive piloting (Albert et al., 2021; Csizér et al., 2021).

  1. 1.

    Motivated learning behavior (5 items, α = 0.82): signals the extent to which learners are ready to invest energy in learning English (example: I can honestly say that I do everything I can to master the English language).

  2. 2.

    Autonomous use of technology (5 items, α = 0.83): reflects learners’ abilities to utilize internet- and computer-based opportunities to improve their English knowledge (example: I often use the Internet to improve my English).

  3. 3.

    Enjoyment (6 items, α = 0.78): refers to learners’ feelings of enjoyment while taking part in the activities and topics during English lessons (example: I enjoy the topics that we discuss in English lessons).

  4. 4.

    Hope (6 items, α = 0.78): measures how hopeful learners feel about achieving success in learning English at school (example: I feel hopeful about overcoming challenges in the process of learning English).

  5. 5.

    Pride (5 items, α = 0.88): taps into the extent to which learners feel proud of their achievements in English learning (example: I am proud of my achievements in language learning).

  6. 6.

    Curiosity (6 items, α = 0.83): measures how curious learners feel about learning English, and the topics and activities they encounter during the English lessons (example: In English lessons, we deal with topics that arouse my curiosity).

  7. 7.

    Anxiety (5 items, α = 0.69): taps into learners’ feelings of inhibition experienced in connection with English language activities in school lessons (example: I get frustrated if I can’t understand an English-language text).

  8. 8.

    Boredom (5 items, α = 0.79): measures the extent to which learners feel bored during the activities and topics in the English language lessons (example: I get bored by the activities in English lessons).

  9. 9.

    Apathy (4 items, α = 0.77): refers to learners’ feeling of hopelessness related to success in English language learning in school (example: I feel hopeless about ever mastering English in the school).

  10. 10.

    Confusion (5 items, α = 0.78): measures the extent to which learners feel confused about language learning in class (example: Sometimes I feel confused because I don’t understand what is happening in the English lessons).

  11. 11.

    Shame (5 items, α = 0.80): taps into learners’ feelings of shame about their achievement and actions during English lessons (example: I feel ashamed if I can’t answer a question during our English lessons).

  12. 12.

    Self-efficacy beliefs (6 items, α = 0.93): reflects learners’ beliefs about their abilities to successfully learn English (example: I believe that I can do the speaking tasks we are given during English lessons).

Data Collection and Data Analysis

The project started in 2018 with the careful development of the questionnaire that included two rounds of empirical validation (Albert et al., 2021; Csizér et al., 2020, 2021). Data collection for the main study, reported here, started in the autumn term 2019 and finished in the first months of 2021. Due to the COVID-19 pandemic, data collection was switched to online form in 2020, but the paper-and-pencil and online versions of the questionnaire were identical.

Ethical issues were considered carefully. Permission to include students in our study was sought from the schools, and data were collected in an anonymous manner, without any personal data based on which students could be identified. Students’ participation was voluntary, and they could opt out of answering any questions both in the paper-and-pencil and online versions of the questionnaires.

The steps of data analysis in the current article were dictated by our research questions: Based on the distributional characteristics of the scales (Csizér et al., in preparation), we decided to use parametric statistical tests. First, the reliability of the scales was calculated with the help of Cronbach’s alpha values; these results can be seen next to the description of the scales in the methods section. To answer our first research question, we used independent samples t tests along with the Cohen’s d measure of effect size (for the analytical cut-off points we used the guidelines by Cohen et al., 2018). For our second research question, we ran regression analyses separately for boys and girls to arrive at comparative results.

Results

What Characterizes the Differences Between Girls and Boys Concerning Second Language Learning Motivation, Emotions and Autonomous Use of Technology?

Our first research question can be answered based on Table 1, containing descriptive statistics of the scales as well as independent samples t tests to compare the mean scores of boys and girls. Out of our 12 scales, nine show significant gender-based differences. It seems that while girls showed an overall higher level of motivation to learn English, boys scored significantly higher on the autonomous use of technology scale (t = 3.579, p = 0.001, Cohen’s d = 0.22) indicating that they are more independent learners than girls, although the effects sizes are small in both cases. In addition, boys had a significantly higher level of self-efficacy (t = 5.662, p = 0.001, Cohen’s d = 0.34) meaning that they are more confident that they can successfully learn English; nevertheless, this effect size is still small. As for positive emotions, girls seemed to be slightly more curious about the learning process, which was coupled with a minimally stronger feeling of enjoyment (as indicated by the low effect sizes), while there was no gender-based difference when hope and pride were measured. Stronger effects could be witnessed in the case of negative emotions. Specifically, girls were more anxious than boys (t = −7.995, p = 0.001, Cohen’s d = 0.48) and girls experienced more confusion (t = −4.793, p = 0.001, Cohen’s d = 0.28) and shame (t = −6.388, p = 0.001, Cohen’s d = 0.39) in connection with language learning, but they did not seem to be as bored by the process as boys (t = 3.868, p = 0.001, Cohen’s d = 0.23).

Table 1 Gender-based Differences Concerning the Scales of the Analysis
Full size table

To What Extent do Second Language Learning Motivation, Self-efficacy and Emotions Impact the Autonomous Use of Technology of Boys and Girls in Hungarian Secondary Schools?

In order to explore the impact of the ID variables included in the study on students’ autonomous use of technology, we ran linear regression for boys and girls separately. Table 2 shows that the variance explained was greater in the case of girls (R2 = 0.403, F = 92.660, p < 0.001) than for boys (R2 = 0.370, F = 46.687, p < 0.001), and there were both similarities and differences between the two genders as regards predictors of autonomous use of technology. Similarities included the strong positive influence of self-efficacy (for boys: β = 0.352, for girls β = 0.428), a more limited contribution from motivated learning behavior (for boys: β = 0.171, for girls β = 0.239), and even less influence from boredom (for boys: β = 0.164, for girls β = 0.084) and confusion, whose effect was negative (for boys: β = −0.102, for girls β = −0.166). In terms of the differences, autonomous use of technology was affected by shame (in a negative way β = −0.089, and pride (positively β = 0.125) in the case of boys, while for girls the positive impact of anxiety (β = 0.103) was significant.

Table 2 Multiple Linear Regression with Autonomous Use of Technology as a Dependent Scale
Full size table

Discussion

Concerning our first research question (What characterizes the differences between girls and boys concerning second language learning motivation, emotions, and autonomous use of technology?), some of the differences that have been identified between male and female learners are in line with previous findings reported in the literature, but some contradict them. For example, whenever gender differences are reported in connection with second language learning motivation, similarly to our study, it is usually females who appear to be more motivated than males (Dörnyei et al., 2006; Iwaniec, 2019; Kissau et al., 2010; Öztürk & Gürbüz, 2013; Ryan, 2009; Williams et al., 2002). The differences observable regarding enjoyment seem to suggest that, similarly to other studies (Dewaele & MacIntyre, 2014; Dewaele et al., 2016) female students enjoyed their English classes significantly more than male students, which is important since positive emotions, enjoyment in particular, are hypothesized to contribute to motivation (MacIntyre & Vincze, 2017). In this sense, the fact that females also reported higher levels of anxiety and shame might be somewhat counter-intuitive, but higher levels of anxiety have been consistently found among females (Dewaele & MacIntyre, 2014; Dewaele et al., 2016;), hinting at the complexity of emotional experiences where simultaneous appearance of positive and negative emotions is not uncommon. Although we found no information on gender differences in connection with language learning related shame in the reviewed literature, it appears to be an emotion that is closely associated with anxiety, so it is plausible that females might be more prone to experiencing it as well, as our results suggest. Regarding other emotions, higher levels of curiosity and confusion among girls appear to be novel findings, but they seem to fit the general pattern indicating higher emotionality of females in general, signaled by their well-known tendency to report both higher enjoyment and anxiety (Dewaele & MacIntyre, 2014; Dewaele et al., 2016). One of the findings that do not appear to fit this pattern is that there was no differences between the two genders in hope and pride, which might be a phenomenon specific to the Hungarian context; however, further research is needed to clarify its causes. Moreover, a higher level of boredom was found in males than in females. Nevertheless, higher levels of boredom in the case of males have already been found in Hungary (Albert et al., 2018a, 2018b a, 2018b); moreover, boredom-proneness has also been consistently reported to be higher in males (Vodanovich et al., 2011), which indicates that males might also be more susceptible to this emotion while learning a foreign language.

Despite the fact that females have been consistently found to be more autonomous in several contexts (Şakrak-Ekin & Balçıkanlı, 2019; Zhao & Chen, 2014), when autonomy was measured in our study specifically as autonomous use of technology, male students scored higher than females. This might be linked to the fact that there are many different facets of autonomy (Benson, 2013), and previous investigations possibly measured learners’ autonomy in relation to their classroom learning and related activities (Şakrak-Ekin & Balçıkanlı, 2019; Zhao & Chen, 2014), while the scale employed in our study was more concerned with using technology for out-of-class language learning. Out-of-class or extra-mural language learning covers a range of activities, and out of these, there is at least one that has been consistently found to be more typical of males, and that is using computer games for language learning (Jensen, 2017; Muñoz, 2020). We hypothesize that male learners’ familiarity with computer games might have extended to other technology-related activities, making them more autonomous in this respect. Male students’ higher self-efficacy beliefs are also not supported by findings of empirical studies in the field of second language learning because previous studies either found no gender differences in this regard (Kissau et al., 2010) or reported higher self-efficacy for females (Mills et al., 2007). However, the observation that male learners tend to overestimate their abilities while female students tend to underestimate them (Iwaniec, 2019) might explain our findings to some extent.

When answering the second research question of the study (To what extent do second language learning motivation, self-efficacy and emotions impact the autonomous use of technology of boys and girls in Hungarian secondary schools?), there seems to be a basic pattern which is typical for both male and female learners. The strongest determinant of autonomous use of technology was learners’ self-efficacy, which has been consistently found to be correlated with and contribute to autonomy in earlier studies (Csizér et al., 2021; Han & Reinhardt, 2022; Honarzad & Rassaei, 2019; Tilfarlioglu & Ciftci, 2011). The relationship between motivation and autonomy is also well-known; although first it was proposed that autonomy might lead to higher motivation (Deci & Ryan, 1985; Dickinson, 1995), it has also been proven since then that motivation can contribute to making learners autonomous (Kormos & Csizér, 2014), which could be observed in our case as well. The negative contribution of confusion is also not surprising since exercising autonomy is probably hindered if the learner feels confused and does not know what to do. Despite this intuitively appealing explanation, further research is needed to substantiate the existence of a negative relationship between confusion and autonomy. The positive contribution of boredom might be more difficult to account for, but it can be hypothesized that autonomous actions, especially ones related to the use of technology and out-of-class learning, might partly be motivated by a certain extent of boredom that the learner experiences in connection with their language classes, as this might urge them to take action and seek out different, possibly more challenging opportunities for language learning. Pawlak et al. (2022b) used similar reasoning about the potential activating effect of intense boredom (Pekrun et al., 2010) when trying to account for the motivating effect of boredom in their study.

As regards gender-related differences, the two additional emotions influencing autonomous use of technology were pride (with a positive effect) and shame (with a negative effect) in the case of boys. Both of these emotions can be hypothesized to affect the learner’s general self-confidence, possibly exerting their influence through this latent variable not measured in the current study. Learners who are proud of their achievements might be braver to act on their own, while feeling ashamed would probably prevent them from executing language learning activities on their own. The positive contribution of anxiety in the case of girls appears to be more difficult to explain, although Pawlak et al. (2022b) also found evidence for its motivating effect. What might happen, besides potential facilitating effects of anxiety, is that female learners who experience feelings of anxiety about their English classes might prefer learning on their own with the help of various technology-based means and might become more autonomous this way. Of course, the contribution of different emotions in the case of the two genders should be confirmed in future studies devoted to examining the hypothesized gender differences.

Conclusion

Our study highlights the existence of differences in the role of ID variables when it comes to learning or using English in the case of the two genders. Although the significance of contextual factors is increasingly acknowledged when studying ID variables (Dörnyei & Ushioda, 2021; Ushioda, 2015), gender differences are not very frequently addressed despite the fact that there appear to be obvious differences between the genders in their emotional experiences, motivation, self-related cognitions, and even their autonomy. If our aim is to deliver optimal learning experiences to our learners, these gender differences should be acknowledged and taken into consideration by teachers in their classroom teaching practice. Based on our findings, teachers should aim to boost their learners’ self-efficacy and motivation and reduce their confusion regardless of their gender, while keeping in mind that boys and girls might react to a number of emotions they experience in their English classes in different ways. Since some of our findings, especially those referring to learners’ emotional experiences, are novel, further research should be devoted to their investigation. This could happen through similar questionnaire studies targeting the same variables; however, collecting qualitative data with the help of interviews might prove even more useful in shedding light not only on the lived experiences of learners but some underlying factors as well. The lack of qualitative data can in fact be considered the biggest limitation of the present study. Qualitative data would also allow for investigating gender-related questions in longitudinal, context-bound designs.