\nDicle University, Diyarbakir, Turkey
\n<suleyman.basaran
![\"atmark\"](\"http:\/\/www.tesl-ej.org\/atmark.png\")
dicle.edu.tr><\/p>\nNe\u015fe Cabaro\u011flu<\/strong> The ubiquitous use of Internet-based mobile devices in educational contexts means that mobile learning has become a plausible alternative to or a good complement for conventional classroom-based teaching. However, there is a lack of research that explores and defines the characteristics and effects of mobile language learning (LL) through language learning podcasts. This study aimed at investigating possible effects of the use of podcasts via mobile devices on LL beliefs. Using a pre-test\u2013post-test research design, the Beliefs about Language Learning Inventory (BALLI) was given to a non-random convenience sample of 187 university students before and after the twelve-week podcast-based LL program. Collected data were analyzed by using descriptive statistics and a Wilcoxon signed rank test. For triangulation purposes and to gain deeper understanding of the process, qualitative data were collected with semi-structured interviews with 16 participants selected through quota sampling strategy. Detailed comparative analysis of BALLI and interview data revealed that podcasts had positive effects on certain types of LL beliefs. <\/p>\n Keywords<\/em>: podcast, language-learning beliefs, belief change, BALLI<\/p>\n The Internet is the most important innovation of late 20th and early 21st centuries. The advent of the Internet and developments in communication technologies are cited with and compared to the discovery of electricity and of the light bulb (e.g., Zukowski, 2007). In fact, it is the Internet that has accelerated globalization and thus turned the world into a global village, creating a knowledge economy (Loy, 2000) that has changed many conventions. This new system not only curbed distances in terms of immediate access to the furthest corner of the world, but also created abundant and diverse resources and even parallel worlds such as Second Life.<\/p>\n Educational conventions and practices must certainly change in this constantly evolving world (Collis, 2005). Easy access to fast Internet connections and common use of mobile devices might have implications for second or foreign LL. Delivery of individualized but at the same time comprehensive content in real time via the Internet and digital technologies provides an effective means for creating appropriate learning environments that meet personal needs (Zhang & Zhou, 2003). Hence, a paradigm shift has occurred in academic circles in general and English language teaching in particular. According to Hedberg and Lim (2004), educators have adopted e-learning-related technologies both to extend conventional methods and to develop new skills and tools for learning and instruction. Such technologies have provided instructors with new possibilities and choices to overcome persisting problems stemming from lack of resources and time constraints, enabling students to enjoy new learning experiences. Current technological means can be applied to learning activities such as collaboration, digital storytelling, oral conversations, multimedia messages and podcasting, which can enhance second and foreign LL. <\/p>\n The assertion that novel Internet-based applications may enhance face-to-face communication and even replace it in LL settings is well grounded in previous research (e.g., Cabaro\u011flu, Ba\u015faran & Roberts, 2010; Cabaroglu & Roberts, 2007; Hampel & Hauck, 2009; Miwa & Wang, 2011; Ware & O\u2019Dowd, 2008; Warschauer & Kern, 2000). Yet, continued research is crucial to explore and define characteristics and effects of mobile language learning and more specifically foreign language learning through LL podcasts. It is especially the impact of LL podcasts on language beliefs that needs to be investigated. A deep understanding of such psychological constructs in mobile settings might facilitate development and implementation of novel educational strategies for more efficient foreign LL.<\/p>\n The purpose of this study is to describe and explain the process and the impact of using podcasts as LL objects on learners\u2019 beliefs about learning English as a foreign language (EFL). The study aims to investigate whether there is a positive change in EFL students\u2019 beliefs about EFL. More specifically, it aims to find out whether there is any difference between students\u2019 beliefs about language learning before and after using podcasts as LL objects and aids.<\/p>\n In parallel with the increase in the number of technological innovations, there has been a surge of research concerning the effect of mobile technology upon LL. Earlier research asserts that mobile technologies can motivate foreign language learners via portable and flexible learning more than localized classroom learning (Norbrook & Scott, 2003; Sandberg, Maris & de Geus, 2011). Kukulska-Hulme (2005), for example, reported the findings of three studies that underlined the potential of mobile technology applications for foreign and second LL. In another study, Kukulska-Hulme and Shield (2007) elaborated on the effectiveness of mobile technologies such as podcasting in foreign LL by asserting that it can provide students with an affective and low-cost tool for taking control of what they learn and thus improving their language proficiency. Learners\u2019 beliefs about LL have been studied extensively (e.g., Baker, 2008; Horwitz, 1988; Horwitz, 1995; Pan & Block, 2011; Schulz, 2001). Teachers\u2019 beliefs and perceptions about using technology have also been investigated (e.g., Ottenbreit-Leftwich, Glazewski, Newby, & Ertmer, 2010). However, earlier research is scant on addressing belief change and specifically on the effect of emerging technologies on belief improvement. A review of studies about mobile learning and podcasting and LL beliefs is given below. <\/p>\n Mobile learning (M-learning) is a maturing field (Pachler, 2007) of research that has grown out of the ubiquitous use of digital technology and affordable portable devices. According to Witherspoon (2005), developments in IT and digital technologies have created a new academic eco-system and have benefits for \u201ctomorrow\u2019s environment for learning\u201d (p. 3). Witherspoon also asserted that \u201c[t]echnology is changing everything from pedagogy to system-wide decision-making\u201d (p. 8) and that the \u201cacademic eco-system is complex and ever-changing\u201d (p. 11). The fact that digital technology has potential for LL does not necessarily mean that it simplifies the process. Rather, it adds new challenges to the already highly complex process of foreign language instruction. New technology demands adoption of new strategies by learners and new additional roles by teachers (Chen & Chang, 2011; Yang & Chen, 2007). Moreover, Fischer (1992) claimed that basic social patterns are not easily changed by new technologies and that they withstand even widespread innovations, adding that effects of new technologies are modest and differ from one technology to another and can be contradictory.<\/p>\n M-learning has been boosted by the recent popularity of portable audio and video players and free delivery of digital content in the form of podcasts, which is a compound term coined from \u201ciPod and \u201cbroadcast\u201d, meaning broadcasting via Internet to be played on iPods (Bankhofer, 2005). As with other types of technologies, these new developments were first taken up by young people, affecting their leisure activities such as listening to music and playing games. Nowadays, however, podcasts are also used to learn foreign languages, and there is a huge amount of free content on the Internet. Rosell-Aguilar (2007) compared the impact of podcasting on LL to the impact of the arrival of the Internet to stress their importance in LL. In a qualitative study of critical theory and popular culture in a secondary classroom, Bausell (2006) found that literacy practices developed in popular culture were adopted at a rapid rate, the use of podcasting as an alternative means for students to express themselves holding significant pedagogical potential. Podcasts can be effective tools for integrating with the target culture and enabling students to gain a sense of belonging to a community and thus overcome social and psychological barriers such as low self-esteem, anxiety and poor motivation. Some learners who systematically listen to podcasts enter a state of \u201cflow\u201d and temporarily forget that they are listening to a foreign language (McQuillan, 2006), which is very important for acquisition. However, various decisions concerning the content and comprehensibility level of podcasts and creating opportunities for comprehensible output make teacher help and guidance indispensable (Timu\u00e7in, 2006).<\/p>\n As the use of podcasts for learning is a relatively recent phenomenon, it is not surprising to see that there is little research on their pedagogical potential and implications in foreign LL (Celik, 2012; Rosell-Aguilar, 2007). Although there are some descriptive and informative studies that mostly focus on positive impact of podcasts in education (e.g., Cebeci & Tekdal, 2006; Dlott, 2007; Troutner, 2007; Zukowski, 2007), the lack of a comprehensive theoretical framework to research is apparent, with the ways and practices in which podcasts can effectively be used still under debate (e.g., Beheler, 2007; Golonka et al. 2012; Stanley, 2006; Zielke, 2007).<\/p>\n Beliefs about LL are accepted as learners\u2019 metacognitive knowledge about themselves, their goals and needs (Bernat & Gwozdenko, 2005). Beliefs affect attitudes and motivation (see Baker, 2008 for the impact of learners\u2019 beliefs and Chen & Goh, 2011 for that of teachers) and students\u2019 efficiency in classroom setting (Horwitz, 1988). In the literature on constructs that affect LL and acquisition, the terms belief and perception are used interchangeably (e.g., Mori, Sato & Shimizu, 2007; Schulz, 2001; Tse, 2000). Therefore, for the purpose of this study the terms belief<\/em> and perception<\/em> will be taken as synonymous. <\/p>\n With regard to student beliefs, Horwitz developed a 34-item Likert-scale data collection instrument in 1985, the Beliefs about Language Learning Inventory (BALLI). Horwitz (1988) suggested that if students have preconceptions and negative beliefs about the way languages are learnt and particularly if their beliefs are different from teachers\u2019 beliefs and practice, this may lead to poor confidence in the teacher, dissatisfaction with the course and poor achievement amongst learners. This idea was later confirmed in a number of studies. For instance, Mantle-Bromley (1995) found that some students may come to FL classes \u201cwith certain attitudes, beliefs, and expectations that may actually prove harmful to their success in the classroom\u201d (p. 383). Teachers need to investigate their students\u2019 beliefs so that they can be supportive, help them overcome their feelings of \u201cisolation and helplessness,\u201d and \u201coffer concrete suggestions for attaining foreign language confidence\u201d (Horwitz, Horwitz & Cope, 1986, p. 132).<\/p>\n Learners\u2019 LL beliefs, which may differ across groups (Horwitz, 1999) need to be investigated before implementing any program or introducing any innovation (Sakui & Gaies, 1999). Horwitz (1995) explained how students\u2019 affective reactions influence LL. She further stressed that it is essential for teachers to give priority to the emotional needs of their students because affective factors influence the learner\u2019s willingness to engage in the activities, a necessity in developing second language proficiency. In a study on the comparison of learners\u2019 and teachers\u2019 beliefs about LL and syllabus design, Bulut and \u00dc\u011f\u00fcten (2003) found that learners\u2019 perceptions did not match those of the teachers. Unlike teachers, students had highly positive perceptions of grammar. The study also revealed that listening and speaking were the most favored domains, while reading was the second and writing the least favored. The findings changed the teachers\u2019 classroom activities; for instance, a future program included more listening comprehension activities. According to Schulz (2001), a mismatch between teachers\u2019 and learners\u2019 beliefs about LL can be harmful; therefore, teachers should investigate students\u2019 beliefs and make sure that they are modified to avoid any conflicts between students\u2019 beliefs and classroom activities. Horwitz, Bresslau, Dryden, McLendon and Lu (1997) provided further evidence for the need to adjust instruction to learners\u2019 needs and expectations and suggest multiple ways of teacher collaboration to do so (see also Fox, 1993; Horwitz, 1988). <\/p>\n One important question tackled by researchers is whether beliefs are flexible. Cabaroglu (1999) as well as Cabaroglu and Roberts (2000) tested the widespread view that beliefs are inflexible and reported to have observed some development. They found that early confrontation of pre-existing beliefs and self-regulated learning opportunities were the most important factors in positive belief change. Additionally, Tse (2000) used an autobiography technique to analyze perceptions of 51 adult FL learners about classroom atmosphere and instruction. The study revealed three categories or themes: classroom interactions, perceived level of success, and attributions of success and failure. Students were reported to believe that instruction focused too little on oral communication, that their proficiency was low, and that their failure was a result of lack of effort. Conjecturing that negative perceptions can lead to anxiety and lack of motivation, Tse (2000) wondered \u201cwhy the changes in FL pedagogy that have been implemented over the past few decades have appeared to do so little to change student perceptions of the classroom\u201d (p. 82). She hypothesized that probably the changes in classroom instruction were not sufficient to overcome the social milieu, the cultural assumptions that promote the view that L2 learning is difficult and relatively rare for adults. <\/p>\n Analyses of changes in 146 pre-service EFL teachers\u2019 beliefs about LL also demonstrated that beliefs are difficult to change and that considerable effort is needed to change detrimental beliefs (Peacock, 2001). Stressing the need for further research on belief change, Bernat (2005) commented that studies of the time did not explain how learner characteristics affect the nature of beliefs and that there was a need for an interdisciplinary approach so as to find out how cognitive and personality psychology provides a foundation for a possible relationship between learner beliefs and personality. As the main aim of the study was to analyze the process and the impact in a specific social context, rather than making broad generalizations about the effects of using podcasts in LL, a convenience sampling strategy was adopted. Convenience sampling is more viable when description rather than generalization is the goal (Dawson, 2002). To collect quantitative data and describe and compare LL beliefs before and after a twelve-week program in the academic year of 2009-2010, 187 first-year students at the education faculty of a state university in southeast Turkey were taken as a convenience sample. The program required participants to listen to podcasts repetitively and do related tasks. Students were given the Beliefs about Language Learning Inventory (BALLI) before and after the program to observe whether there was any change in their beliefs. Qualitative data were also collected in order to gain a better insight into participants\u2019 beliefs and perceptions and possible changes. Qualitative data were necessary also for triangulation purposes, for the pre-test\u2013post-test design of the study did not comprise a control group. As the number of the participants for the main study (N=187 students) was rather high, 16 students were selected based on their background, bio-data and scores at the first administration of BALLI. At this stage, a proportional quota sampling strategy was adopted to ensure that all student groups with different characteristics were represented (Dawson, 2002). <\/p>\n The rationale behind the choice of such a design is justified: to treat all the students equally and fairly, the entire target population was chosen as the sampling unit. In other words, as the use of podcasts in LL was expected to bring about positive results in students\u2019 learning, it was believed to be more ethical to include all members of the target population in the research. The purposive and non-random sampling strategy of the study and the authors\u2019 dual roles as teachers and researchers prevented them from following an experimental research design. Therefore, quantitative data were analyzed descriptively. Experimental research and statistical tests of significance have failed to answer the most essential question, \u201chow,\u201d concerning educational processes. In Van Lier\u2019s terms, \u201cwe know very little about what actually goes on in classrooms\u201d (1990, p. 3). Focusing only on the end-product and the effect of an intervention as is done in experimental studies in social sciences and especially in education would fall short of explaining the highly dynamic factors that are interwoven during the process and what students bring into the classroom (House, 2002). Complemented with descriptive analysis of quantitative data, a qualitative design that entails a case study, interviews and participant observation can serve as a \u201cunique and valuable source of information that complements and informs theory, research and practice\u201d (Marczyk, DeMatteo & Festinger, 2005, p. 148). <\/p>\n The course evolved around the use of podcasts as LL materials and related tasks. The program consisted of 16 weeks of study, with four weeks missing due to religious holidays (two weeks) and midterm exams (two weeks). Each week, students listened to three sections of one podcast before the class hour and did related task-based activities in pairs and\/or in groups during the class. Neither course books nor any other teaching materials were used so as not to interfere with the effects of podcasts. Participants had to listen to the podcasts that were recorded on their mobile phones or mp3 players several times and study the transcripts for each section before classes so that they could participate in the group activities to do the tasks. They were also expected to listen to the sections studied in the class throughout the week that followed. <\/p>\n Interviews with the participants revealed that average listening frequency for each section was four times before the class and once or twice after the class. Although students were encouraged to listen more frequently, some of them expressed during the interviews that they could not listen more because they had to do work for other courses and that listening to the same sections was boring after a few repetitions. The podcasts were specifically designed and produced by experts for the British Council to be used by elementary level young adults and were relevant to the proficiency level of the participants (British Council, n.d.). The podcasts and their support packs with transcripts were downloaded and used with the authorization of the British Council. Each 20-minute podcast comprised eight segments, such as Conversations in English<\/em>, I\u2019d like to meet<\/em>, Quiz<\/em>, Our person in\u2026<\/em>, Your turn<\/em>, Carolina<\/em>, The joke<\/em>, and Tom the teacher<\/em>. Students were given audio files, transcripts of podcasts and tasks for each section to make sure that they all were in possession of all course material. <\/p>\n Tasks usually required students to collaborate with each other and produce something to be presented or discussed in the class or sent to the British Council\u2019s LearnEnglish<\/em> team that produced the podcast. An excerpt from Series 1 Episode 1, Section 2 is given below to give an idea of what tasks that were done by participants look like:<\/p>\n You listened to Zara from Bristol talking about why she would like to meet Angelina Jolie. Is there a famous film actress that you\u2019d like to meet? If you can think of someone, make some notes to answer these questions: <\/p>\n Now put your notes together to write a paragraph about the person and why you\u2019d like to meet her. If you want, you can send your paragraph to learnenglishpodcast@britishcouncil.org.<\/p><\/blockquote>\n In order to investigate learners\u2019 beliefs, a Turkish version of the Beliefs about Language Learning Inventory (BALLI) (Horwitz, 1985) was distributed to 187 Turkish university students (see Appendix A<\/a>). The instrument was given before and after the twelve-week podcast-based course. The data were analyzed by using descriptive statistics and a Wilcoxon signed rank test. Qualitative data were collected from 16 participants through semi-structured interviews conducted in Turkish before and after the program (see Appendix B<\/a> for an English version of the interview items) and analyzed with content analysis techniques to gain a deeper understanding of the process, observe whether or not possible change was due the use of podcasts and related tasks as resources and thus address reliability and validity issues.<\/p>\n Students were enrolled in the programs at the faculty of education of the Turkish university, coming from underprivileged or lower-middle class families that lived in villages or small towns in southeast Turkey. Their age range was between 18 and 22, and 117 of them were male, 70 female. Almost all had been state high-school graduates with little or almost no command of English when they entered the faculty programs. With a few exceptions, their technology-related skills were limited. As part of the study, students were asked about their computer skills and whether they had e-mail accounts and mp3 players during the orientation phase of the program (one week before the program started). It turned out that few of them had e-mail accounts and chat-experience on arrival, and interestingly, almost all students that did were male. Most of them, however, had mobile phones and some had mp3 players. Three different versions of BALLI were designed by Horwitz: one for foreign language teachers (1985) with 27 items, one for ESL students (1987) with 27 items, and another one for U.S. students learning a foreign language (1988) with 34 items. In this study, this third version was adapted for the students to identify their beliefs about learning EFL. The inventory was translated into Turkish, making minor changes concerning language and nationality to adapt it for the students (Appendix A<\/a>). The BALLI assesses five belief areas: the difficulty of LL, foreign language aptitude, the nature of LL, learning and communication strategies, and motivation and expectations. <\/p>\n Most BALLI items are scored on a five-point Likert scale: 1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, 5 = strongly agree. However, two items (4 and 14) have different response scales. Item 4 relates to the perceived difficulty of English: \u201cEnglish is 1) a very difficult language, 2) a difficult language, 3) a language of medium difficulty, 4) an easy language, 5) a very easy language.\u201d Item 14 is about the amount of time needed to learn a language: \u201cIf someone spent one hour day learning a language, how long would it take them to speak the language very well? 1) less than a year, 2) 1-2 years, 3) 3-5 years, 4) 5-10 years, 5) you can\u2019t learn a language in 1 hour a day.\u201d The BALLI does not give an overall score because it identifies individual learners\u2019 beliefs about foreign LL. Instead, students\u2019 responses to each item are treated separately. <\/p>\n The BALLI has been used in a number of studies. Moderate reliability scores for it have been reported in previous research. For instance, Cronbach\u2019s alpha coefficient results were 0.61 in Park (1995), 0.71 in Kunt (1997), and 0.59 in Kim-Yoon (2000). In the present study, Cronbach\u2019s alpha coefficient was computed as 0.59 for pre-test results and 0.71 for post-test results. These moderate reliability scores are due to the fact that items were designed to be interpreted individually based on the five-step scales and also that participants\u2019 responses to individual items showed wide variability because of the nature of the tool (Kim-Yoon, 2000; Nikitina & Furuoka, 2006; Park, 1995). Nikitina and Furuoka (2006) reassure us that \u201cdespite criticisms and doubts regarding the reliability of BALLI, Horwitz\u2019s instrument can be considered to be a suitable tool for conducting research on language learning beliefs in different socio-linguistic settings\u201d (p. 217). As for the validity of the instrument, the similarities among the factors of the BALLI found across different groups indicate that it has high potential construct validity (Nikitina & Furuoka, 2006; Yang, 1992).<\/p>\n Our research design required a comparative analysis of pre-test and post-test results, since it was about whether there was any difference between students\u2019 beliefs about LL before and after using podcasts. The most commonly used statistics test in such cases is the t-test. The t-test assesses whether or not the means of two groups or pre-test\u2013post-test scores for a single group are statistically different from each other. However, one assumption of t-test is that the data must be sampled from a normally distributed population. A Kolmogorov-Smirnov normality test results showed that the BALLI data did not reflect a normal distribution. Therefore, a Wilcoxon signed rank test was used to compare pre-test\u2013post-test BALLI results. The Wilcoxon signed rank test is a non-parametric statistical test that compares data from two related samples or repeated measurements on a single sample. It is used as an alternative to t-test when the population is not normally distributed. It involves comparisons of differences between measurements and requires that the data be at an interval level of measurement. With these characteristics, it is similar to t-test, but unlike t-test, it has no assumptions about the distribution pattern of measurements. Therefore, it is used whenever distributional assumptions of t-test are not satisfied.<\/p>\n As BALLI does not give an overall score, the Wilcoxon test was run for each item separately. Therefore, instead of determining whether or not participants\u2019 beliefs about foreign LL changed as a whole, the study focuses on any possible differences in their specific views regarding each item in the BALLI. <\/p>\n Concerning participants\u2019 beliefs about foreign language aptitude, the Wilcoxon test results show that there is no significant difference between pre-test and post-test BALLI results for Item 1, Item 6, Item 10, Item 11, Item 16, Item 19, Item 30, and Item 33 (p > 0.05; Table 1). This means that participants\u2019 beliefs about most items regarding aptitude did not change after a twelve-week program of learning English with podcasts and podcast-based tasks. The only positive change occurred in terms of the assertion that some people have a special ability for learning foreign languages (Item 2; M: 2.51-2.85; p < 0.05). It can be assumed that this apparent change resulted from the change in participants\u2019 understanding of the item. We believe that in the first administration of BALLI, participants took the word \u201cpeople\u201d in general, which is the correct and expected understanding. However, in the post-test, their experiences in the classroom made them think of their classmates, whom they had observed to have different levels of ability and performance, and thus understand the term \u201cpeople\u201d to mean \u201cpeople in the class.\u201d Also, it is possible to believe that some people have a special ability for learning foreign languages and at the same time that ability or talent is not so important because anyone can learn a foreign language if they work hard enough. (See Table 1. The number of respondents was 187 for all items).\n\nTable 1. Beliefs about Foreign Language Aptitude<\/em><\/p>\n
\n\u00c7ukurova University, Adana, Turkey
\n<ncabarcu.edu.tr><\/p>\nAbstract<\/h3>\n
Introduction<\/h3>\n
Literature review<\/h3>\n
Mobile learning and podcasting<\/h4>\n
Language learning beliefs<\/h3>\n
\nOne of encouraging studies that provided evidence for positive changes in LL beliefs was carried out by Sim (2007), who showed that beliefs could be changed in a positive way by focusing on active learning and goal setting. Such a focus, Sim believed, seemed to have encouraged more active, responsible and autonomous learning behaviors evidenced in participants\u2019 belief change. Sim (2007), too, encouraged further research on the role of factors dominant in the process of belief change in novel endeavors. Bakker (2008), who provided plausible statistical evidence that instruction of principles of SLA did not have a significant effect on beliefs, claimed that beliefs are not easily changed, after observing that only one belief became significantly stronger by time: \u201cThe instructor should teach the class in German.\u201d She also asserted that gender and LL experience had significant effects on beliefs. Similarly, Hamilton (2009) suggested: \u201c[W]e need to take a step back from our exploration of the pedagogical value of technological functionality, and turn our attentions towards what the teachers and learners bring to the experience in terms of their priorities for and beliefs about language learning\u201d (p. 173). Building on these studies, the present article aims at analyzing possible effects of LL podcasts on LL beliefs of first-year Turkish university students.<\/p>\nMethods and procedures<\/h3>\n
\n
Participants and context<\/h4>\n
\nPersonal observations and informal interviews with students showed that when they had enrolled, they were not ready for student-centered and constructivist approaches to learning in general, and LL in particular. Coursebook-based and teacher-centered instruction at both the state high schools they had graduated from and the programs they enrolled in was the major source of their beliefs. They had many courses with fixed schedules, which seemed to demotivate them. English was a three-hour compulsory course given in two semesters of the first year, and it was only one of the other too many courses for students. Not having been able to master English despite years of formal instruction at state secondary and high schools, they did not believe that they could learn English on a three-hour-a-week course.<\/p>\nBeliefs about Language Learning Inventory and analysis procedures<\/h4>\n
Findings<\/h3>\n