The predominant role textbooks play in classrooms is clear from the research evidence. Lowry and Moser (1995) once indicate that textbooks are responsible for seventy to ninety percent of students' learning content. Christenbury and Kelly (1994) also claim that ninety-one percent of teacher respondents use textbooks regularly. Wade and Moje (2000) further point out that many teachers rely on textbooks to structure content, organize lessons, and provide suggestions and materials for teaching and assessment. Similarly, in Taiwan, studies show textbooks dominate the content of teachers' instruction and are the essential sources of students' learning (Chan, 2004; Chen, 2003; Huang, 2003; Li, 2003; Liao, 1999; Wu, 2002). Therefore, although some educators consider textbooks at best provide a base or core of materials as jumping-off points for teachers and classes, teachers still tend to follow the sequence and pacing of the textbook regularly (Skierso, 1991). Garinger (2002) as well notes that even with the development of new technologies that allow teachers to generate materials of high quality, the demands for textbooks continue to grow. Such heavy reliance on textbooks reflects the preliminary importance of the evaluation and selection of textbooks. Just as Garinger (2001) affirms, "choosing a course textbook is a prospect that must be respected as it has significant impact on the ability of students to meet their language objectives, and affects both the process of how they learn and the outcomes" (Introduction section). Van Els et al. (1977) also claims, "the dominant role that a textbook plays in FLT makes the selection of the textbook an important decision" (p. 298). Harmer (2000) further indicates that even if teachers might omit, replace, add or adapt the teaching materials when they decide the textbooks are not appropriate, too many of the alternatives listed above would make students begin to wonder why they are using the book in the first place, especially if they have bought it themselves. As a result, it is believed that with additional consideration and attention on the textbook evaluation and selection, the outcome of students' learning would be enhanced.
However, the term "textbook evaluation and selection" seems like a new conception to most English teachers in junior high school in Taiwan, and has caused chaos among them (Ou, 1996). Over the last few decades, these teachers have never been asked to be responsible for such a task because they do not make decisions about the English language curriculum or the textbooks used in classrooms. Teachers simply implement the unified series of textbooks edited by the National Institute for Compilation and Translation. It was not until 2002 that the government liberalized the authorized version of English textbooks, left the market open to the commercial publishing companies, and empowered English teachers to take an active role in evaluating and selecting the textbooks. Nevertheless, according to Chan (2004), this educational innovation has resulted in teachers' anxiety about "the conflict of change of roles, increased workload, pressure from parents and the lack of self-confidence"(p. 6). Also, in both Chan's study (2004) and Lin's study (2004), most junior high school English teachers consider that the textbooks produced by the National Institute for Compilation and Translation are the best even though they feel satisfied with the commercially published textbooks. Therefore, in the present study, one of the attentions would be given to explore how junior high school English teachers react to the elimination of the centralized English textbooks as well as the new tasks of evaluating and selecting textbooks. Basically, after the announcement of liberalizing the authorized textbooks, textbooks have been the subjects of research in Taiwan. Since 1993, researchers have been devoted themselves to establishing the systematic evaluation criteria for teachers (Hu, 1998; Huang, 1997; Ou, 1993, 1994, 1997; Tzeng, 1997; Wang, 2001). With the flourishing open market of commercial publication, more and more studies about English textbooks issues are then conducted. Part of the studies focuses on the content analysis of the textbooks, hoping to offer suggestions for textbook editing and selection (Chang, 2002; Chen, 2001; He, 2004; Iang, 2004; Ma, 2003; Pan, 2004; Su, 2004; Yeh, 2003). Part of the researchers devotes themselves to setting the evaluation criteria for English textbook (Fan, 2000; Hu, 1998; Wang, 2001). Part of the studies aims to investigate the considerations or influential factors of textbook compilation (Cheng, 2003; Chung 2003), the process of textbook selection in schools (Chen, 2002; Huang, 2004; Ie, 2003; Kang, 2003) or the retrospective evaluation of the textbooks (Chan, 2004; Chen, 2000; Chen, 2003; Hsieh, 1998; Huang, 2003; Kuo, 2003; Cho, 2002; Leu, 2004; Li, 2003; Liao, 1999; Lin, 2002; Lin, 2004; Wang, 2004; Wu, 2002).
Nevertheless, most of the studies are conducted in quantitative methods, namely, survey research, to gain a general result about textbook issues, and thus are unable to get a closer look at respondents' perceptions or to capture the nuances in data. Besides, few of them are conducted on the issues of teachers' perceptions or attitudes towards the decentralization of the textbooks or the tasks of evaluating and selecting textbooks. Only in Huang's study (2004), Tzeng's study (2002), and Ie's study (2003), are elementary school teachers' perceptions about the decentralization of the authorized textbooks, or their own professional knowledge and capabilities for textbook selection examined. And only in Li's study (2003), the problems that teachers encountered between the predictive and retrospective evaluation are investigated.
