Over the past decades, researchers and teachers (Chen, 2004; Huang, 2004; Lin, 2010) in Taiwan have recognized vocabulary size as one of the crucial factors affecting reading comprehension of Taiwanese EFL learners. An insufficient amount of vocabulary may impede learners’ reading comprehension, including the failure to grasp the main idea of texts, misinterpretation of details in articles, and even to the extent of losing confidence in learning and reading in English. On the contrary, learners with abundant vocabulary tend to benefit from reading because they widen their vocabulary when they read more. The phenomenon, which is defined as “the Matthew Effect” (Adams, 1990), reveals the persistent issue of how to learn vocabulary effectively and efficiently.
Using a word list to memorize vocabulary has long been regarded as an effective and efficient way to learn lexical words, especially in Taiwan. However, learning vocabulary by using L2-L1 word pairs (L2 vocabulary and its corresponding L1 translation) has been found by researchers as being one-to-one correspondences that were actually arbitrary with risk of misunderstanding (Schmitt
& Schmitt, 1995). Acquiring vocabulary through reading, compared to word pairs, is believed to benefit language learners in terms of word knowledge, collocations, vocabulary usage, and language use. Besides, based on the universal L1 learning experiences, unknown words can be inferred and acquired through reading. Learning words through reading is also regarded as “pedagogically efficient” (Huckin &
Coady, 1999, p.82) for there are two activities processed at the same time—one is
reading, and the other is lexical acquisition. For these reasons, learning vocabulary through reading seems to be a practical way for students to increase their vocabulary and is encouraged for ESL and EFL learners.
Researchers concerning vocabulary acquisition in both L1 (Nagy, Herman, &
Anderson, 1985; Shu, Anderson, & Zhang, 1995) and L2 field (Dupuy & Krashen, 1993; Hulstijn, 1992; Pitts, White, & Krashen, 1989; Waring & Takaki, 2003) have recognized that learners have the ability to infer the meanings of unknown words according to contextual clues in reading. In the process of inferencing meanings of unknown words, the form and meaning link of vocabulary is believed to be reinforced because of learners’ mental effort (Rott, 2005; Rott & Williams, 2003).
The retention of inferred words was also proved longer than those explained by synonyms (Hulstijn, 1992). Thus, providing opportunities for learners to infer unknown words seems to facilitate vocabulary learning.
However, some researchers (Hulstijn, 1992; Kelly, 1990; Parry, 1993) have also noticed that the reliance on contextual clues to infer meanings could be problematic.
The major reason is that not all reading materials provide sufficient clues for readers to infer meanings of vocabulary. Some learners may even misidentify the meaning of unknown words and result in the risk of retaining incorrect vocabulary meanings (Nassagi, 2003). To solve this problem, many textbooks and reading materials for ESL and EFL learners have adopted the meaning-given method, which is to provide L1 or L2 glosses, with the belief that the glosses will contribute to both reading comprehension and vocabulary learning.
Though providing glosses seems to be a safe way because it solves the problem of wrong guessing, researchers in another camp have argued that learners tend to
shift their focus to the main idea of the reading material as soon as they are satisfied with the word meanings provided (Hulstijn, 1992; Hulstijn & Laufer, 2001, Mondria, 2003). To be more specific, the meaning of the unknown words will not be obtained by learners due to the little mental effort involved. Recognizing this phenomenon, Hulstijn (1992) suggested that the adoption of multiple-choice glosses instead of single-translation glosses in reading might solve the problem. He claimed that multiple-choice glosses generate more mental effort and induce higher involvement load compared to single-translation glosses. Most important of all, it is the mental effort or the involvement load that contributes to vocabulary learning and retention (Hulstijn, 1992; Laufer & Hulstijn, 2001).
Most of the empirical research investigating gloss types have provided evidence that glosses, no matter in what form, are generally more facilitative to lexical learning than other text enhancements (Hulstijn, Hollander, & Greidanus, 1996;
Jacobs, Dufon, Hong, 1994; Ko, 2005; Watanabe, 1997). However, in terms of the superiority of the single-translation glosses or multiple-choice glosses, empirical studies have not shown the results corroborating the researchers’ prediction (Hulstijn, 1992; Nagata, 1999; Rott, 2005; Watanabe, 1997; Wang, 2005)1. In fact, some latent factors existed in the studies and influenced the validity of the results. One of the problems in most of these studies was that they allocated the identical time-on-task (experiment time) to both of the two treatment groups (single-translation gloss vs.
multiple-choice gloss). According to Hulstijn and Laufer (2001), different tasks
1 According to the Involvement Load Hypothesis (Hulstijn & Laufer, 2001), activities inducing higher involvement load (more elaborate mental effort or behaviors) generate better vocabulary learning and retention. Therefore, readers are expected to learn more words from the reading with multiple-choice glosses for higher involvement load when compared to the reading with
single-translation glosses.
might need different processing time. Learning tasks involving more mental effort could be usually more time consuming. In their study, they preferred to take
‘time-on-task’ as an inherent characteristic and it should not be specifically controlled. In the case of the current issue, processing multiple-choice glosses in reading, which requires more elaborate processing and deeper mental effort, might take learners more time as compared to dealing with single-translation glosses in reading. However, as mentioned previously, almost all the empirical studies investigating these two gloss types distributed the same amount of experiment time to both gloss conditions. Therefore, it is reasonable to question that setting identical time-on-task for the two treatments might be disadvantageous to the multiple-choice group. Thus, a reinvestigation of the lexical learning effects from the two gloss types under different time-on-task might need to be conducted.
Another associated issue which interests the writer of the current study is that some of the studies in Taiwan have investigated how adults and senior high students manipulate multiple-choice glosses in reading (Lin, 2010; Huang, 2007; Ke, 2003;
Wang, 2005), but few have tackled the learners at junior high level. Similar to senior high school learners, cultivating the skills of inferencing word meanings from context is regarded as one of the main objectives according to the Nine-Year Curriculum Guidelines (see website of the Ministry of Education). Test items regarding inferencing unknowns words constantly occur in the Basic Competence Test every year. This implies that junior high learners are assumed to have the basic inferencing ability before entering senior high school. However, to date, no research has really reported the effects of using multiple-choice glosses in junior high level, let alone compared it to the effects of using single-translation glosses in reading for
the beginning learners. Considering the junior high participants’ ability, in the current study, the increase in word recognition2 is regarded as the lexical learning effect from the glosses. Thus, the current study is going to compare the effect of gloss types on meaning recognition. Furthermore, in contrast to senior high school learners which are more skillful learners in terms of reading and inferencing, the perceptions towards the multiple-choice glosses of the junior high learners, who are less skillful learners, also deserve some attention.
1.2 Purpose of the Study
There are three aims in the current study. First, the current study is going to explore whether multiple-choice glosses can generate a better amount of meaning recognition than single-translation glosses. Two treatment groups are set up—one is STG (single-translation glosses) and the other is MCG (multiple-choice glosses).
How many vocabulary they could recognize in the immediate and delayed posttests from the two treatment groups are of concern. Second, the current study also attempts to verify whether the participants in the multiple-choice glosses group need more time to finish the task, as Laufer et al. (2001) have suggested. Thus, a MCG-E group (multiple-choice glosses with time-extended) is set up to be compared with the MCG (multiple-choice glosses). The potential influences of the attributed time-on-task on word recognition are investigated. Last, since multiple-choice glossing is supposed to be new to most of the junior high learners, their perceptions towards multiple-choice glosses might be worthy of investigation. Their major
2 Meaning recognition is one category of vocabulary knowledge, referring to the learners’ ability to select definitions or L1 translations to target words (Schmitt, 2010; Laufer & Goldstein, 2004). When it comes to vocabulary acquisition, meaning recognition is the first and the most essential part that needs to be acquired. It is also highly correlated with the classroom performance (Laufer et, 2004).
reasons for the liking of multiple-choice glosses are further explored through focused-group interviews.
1.3 Significance of the Study
Junior high school students in Taiwan, though with at least 5 years of English learning experience3, learning vocabulary through reading is one of the main skills they must cultivate during this period. Among ways to draw students’ attention to vocabulary, both single-translation glosses (STG) and multiple-choice glosses (MCG) have been proven to have positive effects. Most of the textbooks or reading materials for senior high learners and adults have widely adopted STG, and some begin to use MCG because they have recognized the great value of MCG in terms of lexical learning. However, at present, an investigation of the effects of different gloss types on students at junior high level is a missing piece even though beginning learners deserve more attention (Coady, 1997). Understanding learners’ perceptions towards MCG might provide an avenue for instructors to know learners’ difficulties and to manipulate MCG in a more effective way. Besides, by testifying the influence of the variable “time-on-task”, the effectiveness of multiple-choice glosses might be further understood. The current research aims to provide some pedagogical implications for instructors or textbook writers for junior high students in Taiwan.
As for the theoretical implications, the learners’ perspectives from the focused interviews might provide some insights for future researchers who are interested in the effects of glossing.
3 The ninth graders in the current study have received their English instructions since they were at fifth grade. However, most of them had their first English lesson even earlier in cram schools as most parents believe it is beneficial for their learning.
1.4 Research Questions
The current research addresses the following questions:
1. Does multiple-choice glosses (MCG) generate better effects on word recognition than single-translation glosses (STG)?
2. Does the factor ‘time-on-task’ have any effects on multiple-choice glossed groups in terms of word recognition?
3. What are the perceptions towards multiple-choice glosses of the learners from the MCG and MCG-E groups?
1.5 Definition of Terms
Single-translation Gloss (STG): Single-translation Gloss refers to the provision of L1 corresponding translations to the target words.
Multiple-choice Gloss (MCG): In the current study, there are two L2 items for each of the target words. One is the corresponding L2 translation of the target word and the other is the distractor. The participants will need to choose the correct one by evaluating the clues in the context or utilizing their background knowledge.
Word Recognition: It refers to the ability to match the given L2 words to their corresponding L1 meanings correctly.
1.6 Organization of the Thesis
The present section describes the outline of the subsequent chapters in this
thesis. In Chapter Two, the benefits of reading and the important role the inferencing skill plays in lexical growth will be firstly reviewed. The limitations and possible solutions will then be discussed. After that, the notion of using multiple-choice glosses, its theoretical accounts, and some empirical studies will be reviewed.
Limitations of these studies will be discussed. Chapter Three outlines the methodology, including participants, materials, and instruments used for data collection. The procedure will then be outlined. Chapter Four presents the results and the major findings. The theoretical and pedagogical implications, limitations and suggestions for future research will be focused in Chapter Five.
CHAPTER TWO LITERATURE
To consolidate the claim of the current study, this chapter refers to the associated theories and empirical studies in literature. First, reasons for learning words through reading are provided. Then, the importance of lexical inferencing skills and its facilitative effects on lexical learning will be presented. Section three provides the theoretical accounts and empirical studies concerning the lexical learning effect from adopting glosses (both single-translation and multiple-choice glosses) in reading. The limitations of these empirical studies are discussed in section four. A summary of the overall literature review is provided in the last section.