個人數位助理多媒體注釋之單字學習成效

全文

(1)國立台灣師範大學英語學系碩. 士. 論. 文. Master’s Thesis Graduate Institute of English National Taiwan Normal University. 個人數位助理多媒體注釋之單字學習成效. The Effects of Multimedia Annotations on EFL Learners’ Vocabulary Learning on PDAs. 指導教授：林. 至. 誠. 博士. Advisor: Dr. Chih-cheng Lin 研究生: 符德馨 Te-Hsin FU 中華民國一年八月 August, 2011.

(2) 2.

(3) 中文摘要本研究旨在探討不同語言以及不同視覺輔助之注釋組合 (combined annotations)輔助閱讀對字彙學習的影響，由於受試者必須在個人數位助理 (Personal Digital Assistant)上進行文章的閱讀，故本研究有別於傳統藉由桌上型電腦(desktop)進行注釋輔助閱讀之研究。參與本研究的學生來自中台灣某國立高中一年級，共有一百二十一位。這些受試學生依據不同的注釋組合分成四組，分別為中文圖片注釋組、中文動畫注釋組、英文圖片注釋組以及英文動畫注釋組。這些受試者在研究開始第一週先接受個人數位助理之操作訓練，第二週則實施前測以確定四組受試者對目標單字 (target words)的認識無差異，均為陌生。第三週時受試者在 PDA 上進行四篇指定文章的閱讀，四篇文章均有五個目標單字，總共二十個目標單字且均為動詞，受試者依據自己的組別在不同的組合注釋輔助下閱讀文章，每閱讀完一篇文章及進行完文末的閱讀測驗活動後，立即在 PDA 上接受五個目標單字的立即後測 (immediate posttest) ，四篇文章故總共進行了四次。立即後測包含兩種能力的測驗:一為受試者拼寫出目標單字(word production)；其二為受試者辨認出目標單字 (word recognition)。在實驗結束一個月後，受試者接受與立即後測相同內容但以紙筆形式進行的延宕後測(delayed posttest)，以得知在實驗一個月後受試者目標單字的學習情形。研究結果顯示，在立即後測中，無論在單字拼寫或單字辨認方面，中文及英文注釋對單字所產生之學習效益並無差異，但圖片與動畫輔助則對單字學習產生不同效益，且其中動畫輔助對單字學習所產生的效益顯著大於圖片輔助。而在延宕後測中，無論在單字拼寫或單字辨認方面，中文或英文注釋以及圖片或動畫對單字學習所產生之效益均無差異，但在單字拼寫方面，語言注釋 (verbal annotations)與視覺輔助工具(visual aids)之間的交互作用達到顯著。本研究證實藉由 PDA 做為閱讀媒介而呈現的以母語(Chinese)或以目標語 (English)為注釋對單字的學習效益，與藉由桌上型電腦做為閱讀媒介而呈現的兩種語言對單字的學習效益，結果相當一致。此外，本研究亦證實動畫輔助有助於以動詞為目標字彙的單字學習，並提供了僅具基本語言能力的中學生作為佐證。最後，本文亦論及相關主題的未來研究以及教學建議。. 、多媒體注釋、行動學習、個人數位助理關鍵字:字彙學習關鍵字. i.

(4) ABSTRACT The purpose of the study is to investigate the effects of different combined annotations with first and second languages for verbal annotations and pictures and animation for visual annotations. It was administered on PDAs to further confirm whether the results are in agreement with those implemented on the desktop. According to the assigned annotation type, four classes of 121 senior high school freshmen in central Taiwan read 4 passages, each of which had 5 target verbs, annotated in the following ways: L1 definition-plus-pictures, L1 definition-plus-animation, L2 definition-plus-pictures, and L2 definition-plus-animation. These participants took an immediate posttest on a PDA after reading each and a delayed parallel posttest with paper and pen one month after the experiment. Two-way ANOVA analyses indicated in the immediate posttests no significant differences between the definition languages but significant differences between the visual aids, with animatied pictures more effective, including word production part and word recognition part; however, in the delayed posttest no significant differences between the definition languages and between the visual aids, whether in word production part or in word recognition part. Besides, different from the results in the immediate posttest, there were significant interaction effects between different verbal annotations and different visual aids found in word production of the delayed posttests. The findings of definition languages were consistent with the previous studies, and this study further offered evidence from younger adults and lower proficient English learners. The performance for animated pictures over still pictures supported the argument that dynamic animation is designed to illustrate concepts of changes and processes such as English verbs. Directions for future research and suggestions for language practitioners are also discussed. Keywords: vocabulary learning, multimedia annotations, mobile-learning, PDA. ii.

(5) ACKNOWLEDGEMENTS I dared not dream such a dream—getting my thesis done. I alone couldn’t have accomplished it. I owe so much to those who accompany me and see me through. First, I am deeply indebted to my thesis advisor—Dr. Chih-cheng Lin. Thanks to his data, the foundation was laid for my thesis. Without his inspiration, I couldn’t have found my way out of chaos in my brain. In busy life, he still spared time to read my thesis and offered me his expert and useful guidance. His full devotion to reseach spurred me to hang on until the accomplishment of my thesis. Without him, the birth of this thesis would have been impossible. Also, I would like to express my sincere thanks to my thesis committee members, Dr. Hsueh-ying Yu and Dr. Jack Tseng. Dr. Yu generously provided her insightful advice, which played a significant role in improving my thesis. Dr. Tseng’s detailed comments and explanation built me a clear picture of how to present my study. Moreover, I would like to extend my thanks to the professors in the Department of English who taught me for the past three summers when I studied in summer program. Their instruction broadened my horizons, making my learning and teaching a different story. Nothing is greater than enjoying teaching and learning. Last but not least, I can’t be too thankful to my family, including my generous parents, my kind parents-in-law, my dear husband and my thoughtful children. Their support and tolerance contributes to what I have achieved.. iii.

(6) TABLE OF CONTENTS 中文摘要.........................................................................................................................i ABSTRACT...................................................................................................................ii ACKNOWLEDGEMENTS......................................................................................... iii TABLE OF CONTENTS ..............................................................................................iv LIST OF TABLES ........................................................................................................vi LIST OF FIGURES .....................................................................................................vii CHAPTER ONE INTRODUCTION...........................................................................1 CHAPTER TWO LITERATURE REVIEW .................................................................4 2.1 Vocabulary Glossing ........................................................................................4 2.1.1 Gloss Types and their Effects................................................................5 2.2 Multimedia Annotations...................................................................................8 2.2.1 Pictures..................................................................................................8 2.2.2 Video ...................................................................................................12 2.2.3 Animation............................................................................................15 2.3 MALL for Vocabulary Learning ....................................................................16 2.3.1 Strengths of MALL for Vocabulary Learning via Mobile Devices ....17 2.3.2 Factors which Make MALL Limited ..................................................21 2.4 The Present Study ..........................................................................................23 CHAPTER THREE METHODOLOGY ...................................................................26 3.1 Participants.....................................................................................................26 3.2 Material and Target Words.............................................................................26 3.3 Instrument ......................................................................................................28 3.4 Procedure .......................................................................................................32 3.5 Data Analysis .................................................................................................33 CHAPTER FOUR RESULTS AND DISCUSSIONS ...............................................35 4.1 Results............................................................................................................35 4.1.1 Pretest Results.....................................................................................35 4.1.2 Posttest Results — Word Production ..................................................36 4.1.3 Posttest Results — Word Recognition ................................................40 4.2 Discussions ....................................................................................................43 4.2.1 L1 Glosses versus L2 Glosses.............................................................43 4.2.2 Animated Pictures versus Still Pictures ..............................................46 4.2.3 Interaction Effects between Language and Visual Aids......................48 4.2.4 Decline in participants’ performance in the delayed posttest .............50 CHAPTER FIVE CONCLUSION ............................................................................52 iv.

(7) 5.1 Summary of the Findings...............................................................................52 5.2 Pedagogical Implications ...............................................................................53 5.3 Limitations of the study .................................................................................56 5.4 Suggestions for Future Research ...................................................................56 REFERENCES ............................................................................................................59 APPENDICES .............................................................................................................66 Appendix A: Four Reading Passages...................................................................66 Passage A: The Bicycle................................................................................66 Passage B: Gifts of Love..............................................................................67 Passage C: Stan’s Worries ...........................................................................68 Passage D: The Bullfrog and His Shadows .................................................69 Appendix B: Target Words...................................................................................71 Appendix C: Pretest .............................................................................................72 Appendix D: Immediate Posttest for The Bicycle................................................74 Appendix E: Immediate Posttest for the Bicycle .................................................76 Appendix F: Delayed Posttest for the Bicycle.....................................................78 Appendix G: Delayed Posttest for the Bicycle ....................................................79. v.

(8) LIST OF TABLES Table 1. Procedures of the Experiment......................................................................32 Table 2. One-way ANOVA Summary for the Pretest ................................................36 Table 3. Summary of Descriptive Statistics for the Pretest .......................................36 Table 4. Summary of Descriptive Statistics for Word Production in the Posttests....36 Table 5. Two-way ANOVA Summary for Languages and Visual Aids in Word Production of the Immediate Posttests.........................................................38 Table 6. Two-way ANOVA Summary for Languages and Visual Aids in Word Production of the Delayed Posttests ............................................................39 Table 7. One-way ANOVA Summary for the Examination of the Simple Main Effect in Word Production in the Delayed Posttests...............................................40 Table 8. Summary of Descriptive Statistics for Word Recognition in the Posttests..41 Table 9. Two-way ANOVA Summary for Languages and Visual Aids in Word Recognition of the Immediate Posttests.......................................................42 Table 10. Two-way ANOVA Summary for Languages and Visual Aids in Word Recognition Delayed Posttests.....................................................................43. vi.

(9) LIST OF FIGURES Figure 1. Screenshots showing the layout of annotations for the L1 glosses-plus-still pictures group and the L1 glosses-plus-animated pictures group................28 Figure 2. Screenshots showing the layout of annotations for the L2 glosses-plus-still pictures group and the L2 glosses-plus-animated pictures group................28 Figure 3. Screenshots showing the layout of the first part of word production for the L1 glosses-plus-still pictures group and the L1 glosses-plus-animated pictures group ..............................................................................................30 Figure 4. Screenshots showing the layout of the first part of word production for L2 glosses-plus-still pictures group and the L2 glosses-plus-animated pictures group ............................................................................................................30 Figure 5. A screenshot showing the layout of the second part of word production for four groups ...................................................................................................31 Figure 6. Screenshots showing the layout of word recognition for L1 groups (left) and L2 groups (right)..........................................................................................31. vii.

(10) CHAPTER ONE INTRODUCTION It is generally agreed that successful vocabulary acquisition lays solid foundations for the mastery of a language. As a brick is to a house, so words are to language acquisition. As Wilkins (1972) noted, “without vocabulary nothing can be conveyed” (p.111). However, for most second language learners, learning vocabulary is not an easy task, especially when they encounter infrequent vocabulary or large vocabularies. In particular, they tend to have difficulty remembering large vocabularies (Oxford, 1990). To reduce stress from encounters of unknown words, glosses or annotations on a text are designed as a helpful technique and aid, through which language learners naturally give more attention to and have repetitive meetings of the target words to enhance vocabulary learning. In addition, glossed words provide information on the target words for language learners to assist them avoid erroneous guesses (Hulstijn, 1992) and improper inference to promote reading comprehension. Less frustration, more interest. Glossing increases learners’ interest in reading and further encourages autonomy (Myong, 2005). As technology advances, multimedia annotations, the digital genre of glosses, emerge and make it possible to diversify glossing presentations, whether in the modes of still pictures, static animations, videos or films. Various modes of annotations provide opportunities for language learners to process information in more than one channel simultaneously so that language learners can process linguistic information deeply and conceptualize vocabulary concepts particularly. (Paivio, 1971, 1990; Mayer, 1997, 2001). So far, most studies have agreed that the combination of verbal. 1.

(11) and visual annotations works better in facilitating vocabulary learning than the isolation of verbal annotations or visual ones. Thus, combined glosses are employed in the present study to help the participating learners learn English vocabulary. With the advent of mobile devices, researchers in language education have been interested in exploring the potential benefits of learning foreign languages via mobile devices. It seems that mobile language learning can be convenient, personalized, and cheap. Beyond the constraints of time and space, mobile devices render language learning ubiquitous (Levy, 2009). In addition, mobile devices with proposed systems embedded can provide personalized or context-aware vocabulary, according to the needs of and the proficiency of language learners (Chen & Chung, 2008; Chen & Hsu, 2008; Chen & Li, 2010). Lastly, compared with the cost spent on other learning facilities like language laboratory, the mobile device is much cheaper (Browne & Culligan, 2008). In light of the advantages granted by the mobile devices, research on mobile. language. learning. has. opened. up. new. realms. for. investigation. (Kukulska-Hulme, 2009). Coming to proponents’ expectations, studies have lent support to the assertion that mobile devices can be effective tools for vocabulary development. Multimedia annotations via mobile devices are worthy of research due to hardware appropriateness, namely the capacity of mobile devices suitable for the installation of multimedia annotations and the screen of mobile devices suitable for the presentation of multimedia annotations. For future practical and efficient use, studies of multimedia annotations on mobile devices are necessary, but, however, now, still scant. Though there has been a surge in the study of glosses administered on the desktop computers, glosses installed on mobile devices have yet to be thoroughly investigated. The purpose of this study is to understand how glosses supported by mobile devices, i.e. the personal digital assistant (PDA), contribute to vocabulary development. Therefore, the 2.

(12) research questions in this study are addressed as follows:. 1) Do L1 and L2 glosses differ in their effect on Taiwanese senior high school students learning English verbs in terms of word production and word recognition in the posttests? 2) Do still pictures and animated pictures differ in their effect on Taiwanese senior high school students learning English verbs in terms of word production and word recognition in the posttests? 3) Are there any interaction effects between the two variables: language type (L1 or L2) and visual aids ( still pictures or animated pictures)?. 3.

(13) CHAPTER TWO LITERATURE REVIEW This chapter mainly presents previous related studies. Section 2.1 discusses the idea of vocabulary glossing without the aid of the multimedia. In section 2.2, there are studies on multimedia annotations with a focus on the effects produced by the different modes of annotation: textual glosses, pictorial glosses, video glosses, and animation glosses. Section 2.3 reveals the review of studies on MALL for vocabulary learning under two subheadings: strengths of MALL for vocabulary learning via mobile devices and factors which make MALL limited. Finally, this chapter ends with what will be done in the present study. 2.1 Vocabulary Glossing A gloss or an annotation functions like a dictionary and provides short definitions or explanations of the unfamiliar words for learners (Yun, 2011; Nation, 1983; Pak, 1986; Lomicka, 1998) to prevent learners from doing the wrong guessing due to improper language ability and insufficient clues in context (Myong, 2005). Besides, the provision of glosses in the text made the interruption of reading process as little as possible (Nation, 2001:175; Myong, 2005). These two assist reading comprehension. Besides, very often the words glossed are marked. These marked glossed words draw learners’ attention to the words and have learners regard the words as items to learn. Glossing leaves learners “noticing” (Nation, 2001: 63) the words, the first psychology condition that the activity uses to help reach the learning goal. This is confirmed by Bowles (2004), whose study indicated the groups with glosses noticed more of the target words than the no-gross group. As Schmidt (1990) asserted, only when learners 4.

(14) “notice” the target word can it be processed further. Then, while glossing is being done, learners see the words in the text, see them again in the gloss, look back at them in the text to see how the meaning in the gloss fits the context (Nation, 2001: 176; Watanabe, 1997): three meetings with the words. Repetition of and involvement in the words are beneficial to vocabulary learning. Also, glosses help learners overcome the text which may be too difficult to read without glosses (Nation, 2001: 175) and encourage learners’ autonomy (Myong, 2005). Based on the above-mentioned, glossing facilitates language learning. In fact, a lot studies confirm the positive effect of glossing on reading comprehension (Davis, 1989; Jacobs, 1991, 1994; Jacobs, Dufon, & Hong, 1994; Hulstijn, Hollander, and Greidanus, 1996; Lomicka, 1998; Bowles, 2004) and on vocabulary learning (Hulstijn, 1992; Jacobs, Dufon, & Hong, 1994; Watanabe, 1992). 2.1.1 Gloss Types and their Effects The words can be glossed in the learners’ first language or in the learners’ second language and the choice of the language has no different effect on language learning. As Nation (2001: 175) said, “the first requirement of a gloss is that it should be understood”. As Jacobs, Dufon and Hong’s (1994) study showed, L1 and L2 glosses had no different effect on the recall of the passage and on the translation of the glossed words and that learners favored L2 glosses as long as they were comprehensible. Similarly, Bell and LeBlanc (2000) also reported no significant difference between the effects of L1 glosses and L2 glosses on the multiple-choice reading comprehension test (Myong, 2005). However, different from the subjects in Jacobs, Dufon and Fong’s (1994) study, the subjects in this study expressed their preference for L1 glosses. Besides, Myong (1995) found L1 contributed to better vocabulary learning but made no difference to the effect on comprehension, compared with L2 glosses. In 2005, Myong still found L1 glosses as well as L2 glosses enabled learners to comprehend more easily,. 5.

(15) but further pointed out that more than half of the subjects in the study preferred L2 glosses for their reading material. Traditionally, glosses may be presented in the text directly after the glossed word, in the margin on the same line as the glossed word, at the bottom of the page containing the glossed word or at the end of the whole text (Nation, 2001: 175) with the intention of supporting learners’ reading comprehension (Yun, 2011; Nation, 1983; Pak, 1986; Lomicka, 1998). Some research has been conducted on the effects of where to put glosses. Frenskner (1990) reminded that gloss location and presentation method might affect reading comprehension positively or negatively (AbuSeileek, 2011). Watanabe (1997) indicated that glossing right after the glossed word didn’t work so well. Besides, Holley and King (1971) examined the effect of glossing on vocabulary learning and comprehension and found no difference between glosses in the margin, at the foot of the page and at the end of the text. Besides, while Jacobs, Dufon and Fong (1994) was investigating learners’ views on the location of glosses, they found glosses most were presented at the bottom of the page or in the right margin and that learners favored marginal glosses considering they were faster to access (AbuSeileek, 2011). There are no inconsistent agreement over the argument that multiple-choice glosses, compared with single ones, contribute to better language learning. When provided with multiple-choice glosses, learners have to make mental effort to process the provided choice items of the target words and thus develop vocabulary learning so Hulstijn (1992) claimed that “choices did make a significant difference” (Nation, 2001: 176). However, Hulstijn also mentioned that choices might lead to very different results in that some learners made wrong choices and thus disturb learning. Later, Watanabe (1997) found no difference between the effects produced by these two types of glosses. But, later, it was found that the provision of multiple-choice glosses did have an immediate positive effect on receptive as well as productive word knowledge but did 6.

(16) not retain any more word knowledge after five weeks than readers in the control condition (Rott, William, & Cameron, 2002). As the researchers explained, this finding might suggest that the mental effort exerted through multiple-choice glosses did not lead to the sufficiently deep level of processing so as to promote long-term retention of form–meaning connections. Besides, Rott (2005) observed that learners with access to multiple-choice glosses and those in single glosses comprehended main ideas equally well. However, the former showed a tendency to comprehend more supporting ideas. Glosses can carry different amounts of information. Hulstijn, Hollander, & Greidanus (1996) pointed out learners provided with word-level marginal glosses had much better retention scores than those provided with dictionaries. Grace (1998) indicated that learners with sentence-level glosses, in which there was a strong relation between the subject, target word, verb, and function, performed significantly greater short-term and long-term retention of correct word meanings than those without due to deep processing and intake of meaning. Later, Grace (2000) also found that participants in the translation group, who could access an English dialogue-level translation, regardless of their gender, had significantly higher scores on both posttest 1 and posttest 2 than those who did not have this option. Then, AbuSeileek (2011) made a comparison between glosses with 1-2 words, those with 3-5 words, and those with 6-7 words. His results revealed that glosses with 3-5 words produced noticeable differences in learners’ achievement. AbuSeileek (2011) explained that glosses with 3-5 words neither provided difficult synonym to understand like glosses with 1-2 words nor causes attention-split like glosses with 6-7 words. As is presented, glosses contain word-level, sentence-level, and dialogue-level textual definitions. What’s better, with glosses administered on the multimedia, they manage to be presented in diverse forms, including still pictures, dynamic videos, animations etc. Versatile glossing information and modes are more likely to satisfy learners of different 7.

(17) preferences and different performances, and the following section includes studies on glosses with the assistance of the multimedia. 2.2 Multimedia Annotations The advent of the multimedia activates annotation. Annotation nowadays is not fixed to the traditional form, text. As Chun and Plass (1996) proved, aided by the multimedia, the annotation takes its several forms other than text, such as pictures, videos and animations, which make direct association with actual objects and create concrete ideas, thereby, effectively enhancing vocabulary acquisition. To have a global understanding of the effect different annotation modes have on vocabulary learning, researchers have conducted many studies. This section focuses on these studies in terms of different annotation modes, texts, still pictures and dynamic videos and animations. 2.2.1 Pictures The combination of textual and pictorial glosses contributes to learners’ better performances than the isolation of text or pictures. Kost, Foss, and Lenzini (1999) conducted a study on 56 English GFL university learners of beginning level, who were required to read a narrative under one of the three gloss conditions: L1 translation, pictorial gloss, and L1 translation along with pictures in the gloss. The results indicated that subjects under the condition of the combination of textual and pictorial glosses outperformed those under the condition of textual gloss alone or pictorial gloss alone on the recognition of target words in both short-term memory and retention. The combination of textual and pictorial glosses also has a positive effect on learners’ listening comprehension. Jones and Plass (2002) conducted a study to test the effect of textual and pictorial annotations on learners’ listening comprehension and vocabulary acquisition. 171 English FFL university learners were. 8.

(18) randomly assigned to one of the four listening groups: no annotations provided, only written L1 annotations provided, only pictorial annotations provided and both written L1 and pictorial annotations provided while going on with the listening task. On the posttests, it was found that the students remembered word translations and recalled the passage better when they were provided with both written and pictorial annotations while listening than one of these types or no annotations and that pictorial annotations created much larger effect than written annotations, especially for delayed tests. Yanguas (2009) further related multimedia gloss effect to Schmidt’s (1990) noticing hypothesis. Ninety-four college-level Spanish participants read a text under one of the four conditions: no gloss, L1 text gloss, picture gloss and text plus picture gloss. His findings revealed that gloss groups noticed and recognized significantly more of the target words than no gross group and that, in view of comprehension, the combination gloss group significantly outperformed all the other groups. The L2 textual gloss instead of L1 combined with pictorial annotations still leads to learners’ better performances than the isolation of text or pictures. Yoshii, and Flaitz (2002) examined the effect of the annotation type on L2 vocabulary retention. 151 Japanese ESL university learners of beginning or intermediate level read a short story with 14 target words of verbs under one of the three conditions: L2 text-only, picture-only and L2 text plus picture. The results confirmed the overall advantage of the combination of text and picture glosses over text-only and picture-only types on the immediate and delayed tests. As for retention, the participants’ scores on the delayed tests, regardless of the annotation type, declined equally from those of the immediate tests. Different from 14 verbs as the target words Yoshii and Flaitz focused on, Shahrokni (2009) tackled 25 concrete nouns. 90 adult elementary Iranian EFL learners read the texts for comprehension, which were displayed in one of the following types of gloss, L2 text, pictures, or L2 text-pictures. The findings showed that the 9.

(19) textual-pictorial group outperformed on word recognition as well as picture recognition tests, confirming the Dual-coding Theory, according to Shahrokni (2009). To further make sure the effect of textual glosses in the target language and in the native language, Yoshii (2006) included four gloss types in the study: L1 text only, L2 text only, L1 text plus picture, and L2 text plus picture.. The analysis of the results indicated no. significant difference between L1 and L2 glosses for definition-supply and recognition tasks,. significant. differences. between. picture and. no-picture. glosses. for. definition-supply test only and different patterns of vocabulary retention over time for L1 and L2 gloss groups. The above-mentioned studies, regardless of L1 text provided or L2 text, show a similar result—the combination of textual and pictorial gloss has a larger positive effect on language learning, including word recognition, vocabulary retention, listening comprehension, reading comprehension. This result supports Paivio’s (1986, 1990) Dual-coding Theory—information processed in more than one sensory modality, such as verbally and visually together, is more effective for learning than information processed singularly. Among them, according to Jones and Plass (2002), the results of their study further extended Mayer’s Generative Theory of Multimedia Learning (1997, 2001) to listening, which claims learners have to select relevant textual and pictorial information from the target text, organize these two types of information respectively into coherent mental representations and then integrate them with each other into a mental model. However, there are other studies showing different results. Wu and Chu (2007) compared the effects of three modes of vocabulary annotation on vocabulary retention of high school students. 169 high school students read a text under one of the four conditions on a computer program: no annotation, text annotation (L1) only, picture annotation only, and combined text (L1)-picture annotations. The results indicated 10.

(20) that although students in all three annotation groups attained significantly better gains of vocabulary on the immediate and delayed posttests than the control group (no annotation), the three modes of annotation yielded similar effects on vocabulary gains through on-line reading. Besides, Acha (2009) investigated the effect of three different annotation modes on children’s vocabulary learning. Participants of 135 third and fourth graders read a short English story with the target words annotated verbally, visually or both verbally and visually. It was reported that the recall rate of children in the ‘word-only’ group was better than the recall rate of children in the ‘word and picture’ and the ‘picture-only’ groups. In these two studies, young learners, teenagers and children, under the condition of text-plus picture didn’t outperform as the adult subjects did in other subjects. This provides some food for thought. According to Wu and Chu (2007), while text annotation only or picture annotation only might have been adequate for senior high school students to do incidental vocabulary learning in the task, text-plus-picture annotations might produce “redundant information,” which was no aid to vocabulary retention. On the other hand, Acha (2009) attributed the worse performances of ‘picture-only’ group and ‘word and picture’ group directly to “a higher cognitive load (Sweller, 1994),” which a picture involves. Besides examining the effect of different annotation modes, some studies are conducted from the perspective of the learning style. Yeh and Wang (2003) not only investigated the effectiveness of three annotation modes: text annotation (L1+L2) only, text (L1 + L2) plus picture, and text (L1 + L2) plus picture and sound but also aimed to determine whether learners with certain perceptual learning styles benefited more from a particular type of annotation. As was expected, the combination of text and picture was the most effective type of vocabulary annotation among three types. Furthermore, it reported that Chinese ESL/EFL learners preferred visual stimuli to auditory stimuli 11.

(21) and that the effect of the learners’ perceptual preference on the effectiveness of vocabulary annotation types was not evident to vocabulary learning. Chen, Hsieh and Kinshuk (2008) also, by using SMS and MMS messages for the study of English vocabulary, experimented with four types of annotations—words only, words with written annotation, words with pictorial annotation and words with both written and pictorial annotation—and found that the content with pictorial annotation benefited learners with lower verbal and higher visual ability while the content with both written and pictorial annotation benefited learners with higher verbal and visual ability and that, from the perspective of Cognitive Load Theory (Sweller, 1994), the provision of both written and pictorial annotations was not appropriate for learners with lower verbal and visual ability. 2.2.2 Video The text-plus-video annotation can’t rise above the text-plus-picture annotation in facilitating vocabulary learning. Chun and Plass (1996) did a series of studies on GFL university learners, who were asked to read a 762-word narrative by using CyberBuch on the computer. The target words were annotated in three modes: L1/L2 text definition, L1/L2 text plus picture and L1/L2 text plus video. The findings were quite inspiring, for the application of multimedia contributed to a higher rate of incidental learning (25% accuracy on production tests, 77% on recognition tests) than expected, compared with Coady’s (1993) 5% - 15% and Knight’s (1994) 5% - 21% for supply-definition, 23% - 55% for select-definition. Besides, among three modes, scores for words annotated with pictures + text were significantly higher than those for words annotated with video + text or text only, which, according to Chun and Plass (1996), still confirmed the dual-coding theory of Paivio (1986). What’s more, the value of annotation was recognized in the study for the more a learner looked up a. 12.

(22) certain annotation type, the more he used this type as the retrieval cue for remembering words. However, they made no mention of why scores for words annotated with video + text were worse than those annotated with pictures + text. Different from Chun and Plass’s (1996) argument, Al-Seghayer (2001) held the argument for the text-plus-video annotation. 30 various ESL university learners of intermediate level read a narrative text on a hypermedia-learning program. In the narrative text, the 10 target words were annotated in the following three forms: printed text definition alone, printed text definition associated with still pictures and printed text definition associated with video clips. The study yielded the conclusion that the combination of text and video was more effective than that of text and picture in teaching unknown vocabulary. Al-Seghayer (2001) claimed that that’s because “video better built a mental image, better created curiosity leading to increased concentration, and embodied an advantageous combination of modalities (vivid or dynamic image, sound, and printed text).” While some studies take sides for the picture-plus-text annotation or the video-plus-text annotation, others assert either of them along with textual annotation benefits vocabulary learning. For example, Akbulut (2007) investigated immediate and delayed effects of different multimedia annotations on incidental vocabulary learning and reading comprehension of advanced foreign language learners. 69 Turkish EFL university learners assigned to three types of annotations—L2 textual definitions, L2 definitions coupled with pictures and L2 definitions coupled with short videos—read an annotated text for comprehension. The analysis of the data through a vocabulary pretest, posttest, delayed test, a reading comprehension test and a questionnaire displayed that the combination of definitions with aided visuals, regardless of the type of visual annotation, was more effective in facilitating both incidental vocabulary learning and retention than the provision of only definitions, in line with the Generative 13.

(23) Theory of Multimedia Learning (Mayer, 2001), according to Akbulut (2007). Akbulut (2007) thought Mayer’s (2001) Generative Theory of Multimedia Learning based the success of a multimedia presentation in facilitating learning on transfer and retention, that is, the use of multiple vocabulary tasks and the use of a delayed posttest respectively in the study. Furthermore, some studies pointed out when choosing the annotation mode, learners’ preference or ability should be taken into consideration to facilitate language learning, including reading comprehension and vocabulary acquisition. Plass, Chun, Mayer, and Leutner (1998) used the same narrative as in 1996 and conducted their study on 103 English GFL university learners, with twelve of the twenty-four target words annotated with L1 text and L1 text plus picture, and with the other 12 target words annotated with L1 text and L1 text plus video. They observed that the students, besides their better performances in vocabulary acquisition of the posttests when they selected both visual and verbal modes of glosses than only one or no gloss, comprehended the story better when they had the opportunity to receive their preferred mode of annotation, namely visualizers’ choice of visual activities and verbalizers’ choice of verbal activities. In another study, Plass, Chun, Mayer, and Leutner (2003) examined when multiple representations of information in second-language learning help and hinder learning. 152 English GFL university learners of intermediate proficiency read the same 762-word German story on a multimedia computer program while receiving no annotations, verbal annotations, visual annotations or both for 35 key words in the story. The results revealed that the rank order of effects from different modes of annotation on learners’ vocabulary acquisition replicated that found in the previous study by Plass et al. (1998). Besides, what’s more important, it reported that vocabulary acquisition was worse for low-verbal and low spatial ability students than for high-verbal and high-spatial 14.

(24) ability students when they received visual annotations for vocabulary words, but not when they received verbal annotations and that text comprehension was worst for all learners when they received visual annotations. The findings in these two studies provide evidence for the generative theory of multimedia learning (Mayer, 1997, 2001). When a learner actively selects, organizes, and integrates relevant verbal and visual information when presented, meaningful learning is enhanced. Besides, the findings in Plass et al. (2003) also support the cognitive load theory (Sweller, 1994, 1999), which assumes that multimedia learning processes require cognitive resources but are executed under the constraints of limited working memory. To have a clear understanding of the effects when captioned video are used, Sydorenko (2010) conducted a study on 26 second-semester learners of Russian. 8 of them saw video with audio and captions (VAC); another 9 of them saw video with audio (VA) and the other 9 saw video with captions (VC). The findings showed that groups with captions (VAC and VC) scored higher on written than on aural recognition of word forms, while the reverse applied to the VA group. That is, it is suggested that the VAC group learned more word meanings than the VA group and that captioned video tends to aid recognition of written word forms and the learning of word meaning, while non-captioned video tends to improve listening comprehension as it facilitates recognition of aural word forms. Besides, according to the data through the questionnaire, learners paid most attention to captions, followed by video and audio, and acquired most words by associating them with visual images. 2.2.3 Animation In some studies, animations, instead of videos, are applied to annotation. Lin (2009) examined whether animation annotations facilitate the learning of motion verbs. 15.

(25) better than pictorial annotations. Seventy 8th graders were assessed by production and recognition tests using a pretest and two posttests after they read two passages, in which 20 target word of verbs were annotated in three forms: L1 text-plus-animation group, L1 text-plus-graphics group and L1 text-only group. Although the two visual groups outperformed the text-only group, differences between the dynamic animation group and the static graphics group were not detected. As Lin (2009) explained, static graphics with L1 textual definitions could provide sufficient information for EFL teenage beginners to acquire basic action verbs in this study while the dynamic property of animation somewhat distracted learners’ attention from learning, though it also helped them to construct mental images. Kuo and Chiang (2006) also aimed to investigate the effects of multimedia annotations on both vocabulary recognition and reading comprehension. Four intact classes from an elementary school in central Taiwan read a narrative story with 20 content word annotated, which was used as the instrument to implement different versions of annotations: no annotation, textual annotation, animation annotation, and text-plus-animation annotation. Post-tests results showed on both vocabulary recognition and reading comprehension task, the text-plus-animation annotations group was found to perform the best among the four groups. The results supported the dual-coding theory (Paivio, 1986), according to Kuo and Chiang (2006). 2.3 MALL for Vocabulary Learning Researchers are giving more attention to mobile assisted language learning, for it develops with the advantages of e-learning and, it, what’s better, breaks through spatial, temporal boundaries in learning, two of the factors limiting learning process (Chen & Chung, 2007). Chan et al.(2006) even use the term “seamless learning” to specify the continuity of the mobile learning experience. Toward this new phase in the evolution of 16.

(26) technology-enhanced learning, most of the subjects in previous studies hold positive attitude (Kiernan & Aizawa, 2004; Levy & Kennedy, 2005; Thorton & Houser, 2005; Chen & Hsu, 2008; Chen & Chung, 2008; Clough, Jones, McAndrew, & Scanlon, 2008; Huo, 2008; Kennedy & Levy, 2008; Lu, 2008; Song & Fox, 2008; Abdous, Camarena, & Facer, 2009; Cavus & Ibrahim, 2009; Basoglu & Akdemir, 2010; Wong & Looi, 2010; Sandberg, Maris, & de Geus, 2011), whether MALL is administered on mobile phones, Personal Digital Assistants (PDAs), tablet PCs, or other mobile tools (Chen & Chung, 2007; Basoglu & Akdemir, 2010). The following is what has been done with MALL. 2.3.1 Strengths of MALL for Vocabulary Learning via Mobile Devices First, as Kiernan & Aizawa (2004) claimed, “mobile phones proved to be popular learning tools with the learners involved” in the task-based learning project, so the mobile device manages to arouse learners’ interest in learning English vocabulary via it. In Thornton and Houser’s project, Vidioms (2005), after the college sophomores looked through the L1 explanations of English idioms, L2 scripts and animations on the Vidiom website via mobile phones and PDAs, they gave significantly positive evaluations about these materials. Rich multimedia via mobile devices could hold learners’ interest. Another study on informal learning also supports the effect of the PDA and smart phone on the learning interest. It is found that participants were ready to adjust the existing features of their mobile devices, PDAs or smartphones, to match their learning needs, and might develop informal learning interests because of owning the device, and their variety of effective informal learning activities was the result of their interest in informal learning (Clough, Jones, McAndrew & Scanlon, 2008). Besides, when a system was implemented on the PDA, more than half of the learners (73%, 73%, 61% respectively) agreed that using PDA to learn English or English. 17.

(27) vocabulary is a very interesting mode (Chen & Chung, 2008; Chen & Hsu, 2008; Chen & Li, 2010). The researchers used the proposed system administered on PDA to promote learners’ vocabulary performances or reading ability of English news. Their interest increased the time learners’ spent on the proposed system via PDA. As for another mobile device, the Tablet PC, students claimed that the tablet computer motivated them to learn English and benefited their learning because of its convenience, its satisfying functions, and the creative practice of the target words via the Tablet PCs (Huo, 2008). For example, they were very interested in writing English words and sentences with the colorful stylus, which left them a deep impression of the English. The results of the multiple-choice questions indeed showed that the experimental group learned more English words via the Tablet PC during the project than the control group via paper and pen. Second, mobile devices break through the barriers the classroom set on learning. Learners who used tablet PCs noticed its learning convenience without the limitation to a certain place for they could get on the Web even out of the classroom (Huo, 2008). Most also agreed that learning English by PDA is very convenient because they can perform English learning at any time and any place (Chen & Chung, 2008; Chen & Hsu, 2008; Chen & Li, 2010). For example, Chen and Hsu (2008) managed to administer the Personalized Intelligent Mobile Learning System on the PDA, which could recommend appropriate English news articles for ubiquitous learning based on learners’ different abilities and at the end of the experiment certainly enhanced the reading performance of most learners except one of mental factors. At the same year, the personalized mobile English vocabulary learning system was successfully conducted on the PDA by recommending suitable English vocabulary to learners of different abilities (Chen & Chung, 2008). Identical with the former, the latter also allowed learners to perform ubiquitous English vocabulary learning without constraints of time and place, thereby 18.

(28) contributing to learners’ significant promotion of vocabulary learning performance. Two years later, Chen and Li (2010) further incorporated the situational learning approach with ubiquitous learning via the PDA. At the posttest, learners who used personalized English vocabulary learning systems with context awareness performed better than those who used personalized English vocabulary learning systems without context awareness. Mobile devices facilitates learning, enhancing learning due to the lack of limitations to a certain time or a certain place. With mobile devices, out-of-classroom can be integrated with in-classroom, namely informal learning was a complement to formal learning in the classroom. The pupils from the 3rd condition who learned by working with the Early Bird material in the classroom and the MEL application via the smartphone out of the classroom were the only condition that scored significantly higher on both posttest measures than the other conditions (Sandberg, Maris and de Geus, 2011). There is one more study to show Mobile phones are equipped with the potential “to synergize formal (in-class) and informal (out-of-class) language learning spaces” (Wong & Looi, 2010). In Wong and Looi’s studies, subjects were required to make use of sophisticated mobile phones to take photos in real-life contexts and then made sentences with the newly acquired prepositions or idioms to describe the photos. With the assistance of the mobile phone, subjects went through the learning process on the basis of inductive vocabulary learning, language learning meaning-making through their own proposition-context or idiom context associations. Cavus & Ibrahim (2009) also mentioned repetitive delivery of SMSs caused the learning process to take place out of the classroom environment while the students were involved in their daily activities. Mobile learning, which crosses “spatial and temporal boundaries,” and is very likely to be “interwoven with the learner’s everyday life” (Kukulska-Hulme et al., 2009), facilitating learning. As Sandberg et al. said, “formal school learning can be augmented by learning in an 19.

(29) informal context away from school” (Sandberg, Maris, & de Geus, 2011). Third, the mobile device works as a platform. Take the PDA for example. In a one-year multiple-case study, besides referring to the downloaded dictionary, students made integrated use of various tools on the PDA with the computer for data collection, situated, constructive, reflective, explorative, and conversing uses for their incidental vocabulary learning (Song & Fox, 2008). Besides being a platform for the downloaded dictionary, the PDA can function as a platform for different personalized learning systems (Chen & Chung, 2008; Chen & Hsu, 2008; Chen & Li, 2010). These systems were administered on the PDA, successfully promoting learners’ reading ability of English news and vocabulary learning performances. As for the mobile phone, it incorporated “game elements” in the mobile and motivated 5th graders to use the MEL application, in the form of serious game in their leisure time plus classroom English lessons and this benefited their learning (Sandberg, Maris and de Geus, 2011). Besides, Basoglu and Akdemir (2010) administered the vocabulary program, ECTACO Flash Cards, on mobile phones (Basoglu & Akdemir, 2010), which was expected to be used by the students in the experimental group in their extracurricular times while the students in the control group were expected to study the new words on paper for six weeks, using vocabulary flashcards. The results of the study indicated that although both the use of vocabulary learning program on the mobile phone and the use of flashcards on papers improved students’ vocabulary learning, the former improved students’ achievement almost two times more than the latter. The mobile phone is a successful platform for the vocabulary learning program. Fourth, mobile text messages are the catalyst for vocabulary learning. The messages provided learners with a trigger for extra practice (Levy & Kennedy, 2005; Kennedy & Levy, 2008). Extra practice creates key conditions that may lead to a word being remembered, identified by Nation (2001): noticing, retrieval, creative or 20.

(30) generative use. It is little wonder that Lu’s study in 2008 showed that that “the higher the reading frequency, the higher the vocabulary gains, and vice versa” and that students learning target words by being delivered SMS messaging had greater vocabulary gains than their counterparts learning identical words via print materials, corresponding to the results in Thornton & Houser’s study (2005), in which students provided with SMS-based materials remembered over twice more than students given web-based or printed materials remembered. It is true with another study. Cavus & Ibrahim (2009) divided a total of 48 messages into three group, with each group having 16 messages. The same 16 messages were sent three times during the 9-day experiment, considering spaced repetition fosters learning (Wozniak, 1990). They found that such spaced repetition of SMS text messaging gave birth to students’ success in learning new technical English words. 2.3.2 Factors which Make MALL Limited Technological problems hinder students’ participation (Lu, 2008). First, users’ technological proficiency affects learning. Teachers expressed they lacked confidence in their proficiency in handling a smart phone (Sandberg, Maris, & de Geus, 2011). If a teacher can’t handle a teaching aid well, the learning of his or her students may be influenced. Also, about 20% of the students felt they encountered many difficulties while using the tablet PC, because they were unfamiliar with computers, which led to their longer time spent on solving the problems of a Tablet PC (Huo, 2008). Besides the mobile phone and the Tablet PC, students’ technical capabilities had effect on their use of the podcasts (Abdous, Camarena, & Facer, 2009). Second, the mobile device can’t get on the Internet smoothly. Some students were frustrated by the technical problems while posting photos and sentences to the web via their phones (Wong & Looi, 2010). In Thorton & Houser’s study (2005), the complaint about the wait time for video. 21.

(31) download was made, too. Also, students who had intended to use the mobile, however, later claimed to use it little because the mobile phone took longer to load the pages than the PC (Stockwell, 2008). Numerous students in a PDA-implemented project claimed “access difficulties” made them fail to update their learning portfolios or to use the Internet, which couldn’t meet their expectation that access to the Internet via a mobile device was ubiquitous (Chen & Li, 2010). This is the case with the Tablet PC. Learners mentioned that sometimes they had to wait long to get on line and even encountered disconnections (Huo, 2008). Third, the screen size of the mobile phone and its keypad were two concerns (Stockwell, 2008), while materials presented on the PDA were mostly reported to be clear (Chen & Li, 2010), PDA users reported none of the problems, the tiny screen, the tiny controls (Thorton & Houser, 2005) and more than half of the learners were satisfied with the screen size of the tablet PCs (Huo, 2008). The small screen and pad are difficult to use, which may cause learners to spend more time when the task is completed on a mobile phone (Stockwell, 2010). Stockwell (2007) indicated that the former is “consistently claimed to be a problem for language learning”, contrasting with Thorton and Houser’s findings (2005) that few complaint were made about the sound quality, the tiny screen, and tiny controls. Fourth, due to inconvenient input style, the quantity of language used is limited (Kiernan & Aizawa, 2004). It follows that the content is affected. Besides, over a quarter of the subjects felt that the insufficient content of the SMS lessons given by the teacher, who took the small screen size into account (Lu, 2008). Fifth, cost is also an obstacle (Kiernan & Aizawa, 2004; Stockwell, 2007; Stockwell, 2008). Psychological factors limit the use of mobile device for language learning (Wang & Higgins, 2006; Stockwell, 2008). There is “a disparity between learners’ choice of the mobile phone for their personal uses and for their study” (Stockwell, 2007). Stockwell (2007) observed learners obviously preferred performing the assigned 22.

(32) vocabulary tasks on the computer. In his experiments, learners chose to access the vocabulary activity system, VocabTutor, via the desktop computer rather than via the mobile phone. In fact, more than a half of them even made no attempt to use the mobile phone as a means of learning vocabulary (Stockwell, 2007; Stockwell, 2008; Stockwell, 2010), for they felt the “mobile phone is not for a tool for studying.” (Stockwell, 2008). It was also reported that “many students treated the smartphone as a toy rather than as a learning tool” in the after-school informal setting, though they were actively engaged in the in-class activity (Wong & Looi, 2010). Besides the mobile phone, when learners using the iPod or MP3 players felt they needed extra time to learn how to download podcasts, they would feel that spending the time to use them was not worthwhile, further leading to their refusal to use the iPod or MP3 players (Abdous, Camarena, & Facer, 2009). Environmental factors distract learners, who find it difficult to concentrate on their activities on the mobile (Stockwell, 2010). Some students pointed out that they wanted to study in a quiet environment and this was more achievable with a PC, and others felt they couldn’t concentrate when they started to use the mobile (Stockwell, 2008). Some learners also pointed out the learning environment stopped them from getting access to the vocabulary activities via the mobile (Stockwell, 2007). Thorton & Houser (2005) also mentioned that the environmental factor made many students put off reading their LOTM messages until they could concentrate on them. 2.4 The Present Study This study is going to investigate the effects of different combined annotations administered on the PDA and further to confirm whether the results are in agreement with those implemented on the desktop. On one hand, the textual definition coupled with aided visuals, regardless of pictures, videos, or animations, has been found 23.

(33) effective in facilitating vocabulary acquisition on the desktop (Chun and Plass, 1996; Kost, Foss, and & Lenzini, 1999; Al-Seghayer, 2001; Jones and Plass, 2002; Yoshii, and Flaitz, 2002; Yoshii, 2006; Kuo and Chiang, 2006; Akbulut, 2007; Shahrokni, 2009; Yanguas, 2009; Lin, 2009). However, no similar studies are found to be implemented on the PDA. The PDA has advantage over the desktop in rendering mobile learning without constraint of time and place. Also, the PDA has advantage over the cell phone due to its bigger screen (Thorton & Houser, 2005; Chen & Li, 2010) to present clear materials and its bigger keypad easy to control (Thorton & Houser, 2005). So far, learners’ interest in learning vocabulary has been reported to be successfully aroused by the use of the PDA (Chen & Chung, 2008; Chen & Hsu, 2008; Chen & Li, 2010). Since learners hold positive attitude toward vocabulary learning with the aid of the PDA, if different combined annotations can be successfully implemented on the PDA, the PDA has stark advantage in transforming into a ready learning assistant, which provides personalized vocabulary according to learners’ performance, need and preference for learners to proceed with ubiquitous learning. The main concern of this study is whether animation can serve as a suitable visual aid to depict a series of changes or demonstrate a concept of process. Previous researchers included in their studies diverse parts of speech, nouns, adjectives, verbs, etc. to test the effects of videos as a visual aid on vocabulary learning. Chun and Plass (1996) used different nouns, verbs, and adjectives for each type of annotations, including text, text and pictures, and text and animation (p. 198). So did Al-Seghayer (2001) and Akbulut (2007). In his study, Al-Seghayer included equal amount of nouns, verbs, and adjectives for each type of annotations (p. 213). As for Akbulut’s study, he annotated target words without informing readers of their syntactic categories (p. 505). In Yoshii and Flaitz’s study (2002: 39) and in Yoshii’s study (2006: 89), on the other hand, although they took verbs as their target words, they tested the effects of still 24.

(34) pictures on learning these target verbs. The above combined, there is one issue that needs to attend to is whether or not animated pictures instead of still ones are employed to annotate concepts of changes and processes (Lin, 2009).. 25.

(35) CHAPTER THREE METHODOLOGY This chapter consists of five parts: background information of the participants and their grouping, the choice of reading materials and target words, the introduction of instruments designed for the study, the description of the study procedure and the way to analyze the data of this study. 3.1 Participants A total of 121 freshmen in four separate classes of a senior high school in central Taiwan were invited to participate in the study. These students just passed the high school entrance examination and entered the same high school according to their scores. This meant that they performed almost equally in academic achievement and were supposed to be equipped with 1,000 basic English words suggested by the Ministry of Education after having received at least five years of formal English education since the fifth grade. In other words, these participants were between beginners and intermediates in English proficiency. They were assigned to four treatment groups: L1 glosses combined with still pictures, L1 glosses combined with animated pictures, L2 glosses combined with still pictures, and L2 glosses combined with animated pictures. While the number of the group with access to L2 glosses-plus-animated pictures was 31, the number of the other three groups is 30. 3.2 Material and Target Words Four stories (See Appendix A) were chosen in this study: The bicycle (489 words), Gifts of Love (484 words), Stan’s Worries (456 words) and The Bullfrog and his Shadows (479 words). Assessed by Flesch Reading Ease and by Flesch-Kincaid Grade Level, the four stories were detected with readabilities as follows: 86.8 and 3.2 26.

(36) for The Bicycle, 88.5 and 3.4 for Gifts of Love, 56.1 and 9.3 for Stan’s Worries and 86.5 and 3.1 for The Bullfrog and his Shadows. This means that these reading passages were generally comprehensible to the participating students, roughly corresponding to the textbook that the students were reading at that time in terms of readability. Five target words were located in each reading passage so there were 20 target words in total (See Appendix B). These words were selected because their difficulty was roughly beyond participants’ proficiency. In other words, most of the target words were unknown to the participants in this study. Besides, these words were all verbs. Verbs were annotated and chosen as the target words to investigate the effects of animated pictures along with verbal annotations on learners’ learning verbs, hoping animated pictures can exert their distinctive effects to the full. Each target word was highlighted red in the text screen. When a participant touched the highlighted target word on the screen with a stylus, its annotation would be presented in the mode, consistent with the treatment group of the participant. That is, the participants in the L1 glosses-plus-still pictures group were provided with the Chinese translation and the still pictures of the target words while those in the L1 glosses-plus-animated pictures group had access to the Chinese translation and the animated pictures of the target words (See Figure 1). On the other hand, while the participants in the L2 glosses-plus-still pictures group were presented with the English definition and the still pictures of the target words, those in the L2 glosses-plus-animated pictures group were provided with the English definition and the animated pictures of the target words (See Figure 2).. 27.

(37) Figure 1. Screenshots showing the layout of annotations for the L1 glosses-plus-still pictures group and the L1 glosses-plus-animated pictures group. Figure 2. Screenshots showing the layout of annotations for the L2 glosses-plus-still pictures group and the L2 glosses-plus-animated pictures group. 3.3 Instrument The instruments in the study used to collect data included a pretest, an immediate posttest and a delayed posttest. Pretest. The pretest in the form of a checklist was designed to assess the participants’ vocabulary knowledge before the experiment. It included 30 items, namely 20 target words and 10 distracters. The participants were asked to give a check 28.

(38) mark and provide Chinese equivalents as well if the words were known to them (See Appendix C). Immediate posttests. The immediate posttest was administered on the PDAs to assess the participants’ vocabulary learning immediately after they read a story. Four parallel immediate posttests were created for the four reading passages. Each of the four immediate posttests included comprehension part and vocabulary part. The former was meant to test the participants’ reading comprehension and its results were excluded in this study. The latter was associated with vocabulary test with focus on word production (See Appendix D) and word recognition (See Appendix E). Word production involved two parts—definition questions and sentence cloze questions. In terms of definition questions, the two groups of participants with access to L1 translation were supposed to answer the target words according to their Chinese equivalents (See Figure 3) while the other two groups of participants provided with L2 definition were supposed to answer the target words according to the given English definitions (See Figure 4). As for sentence cloze questions, four groups of participants were all asked to fill in each blank with the right target word to complete the sentence (See Figure 5). On the other hand, match questions were intended to test participants’ word recognition. The two groups of participants with access to L1 translation were asked to match the target words with their Chinese equivalents, whereas the other two groups of participants given L2 definition were asked to match the target words with the given English definitions (See Figure 6). In order to avoid the interfering effects the test items of word recognition part created on participnats’ word production, partcipants answered the test items in word production prior to those in word recognition.. 29.

(39) Figure 3. Screenshots showing the layout of the first part of word production for the L1 glosses-plus-still pictures group and the L1 glosses-plus-animated pictures group. Figure 4. Screenshots showing the layout of the first part of word production for L2 glosses-plus-still pictures group and the L2 glosses-plus-animated pictures group. 30.

(40) Figure 5. A screenshot showing the layout of the second part of word production for four groups. Figure 6. Screenshots showing the layout of word recognition for L1 groups (left) and L2 groups (right). Delayed posttest. The delayed posttest was conducted one month after the experiment with the intention of assessing participants’ vocabulary learning (See Appendix F). The test items for the delayed posttest were identical with those for the immediate posttest, but the delayed posttest, different from the immediate posttest administered on the PDA, was taken by participants with paper and pen.. 31.

(41) 3.4 Procedure All the procedures designed for this study were implemented on the PDA (personal digital assistant). First, the participants were asked to take the pen-and-paper pretest two weeks before the experiment without prior notice. One week before the experiment, all the materials were installed in the PDA and the participants were instructed how to operate the PDA. In the third week, the students were required to read four reading passages displayed on the PDA in four separate class hours. While reading on the PDA, the participants could move on to the next page or move back to the previous page at their own speed. Besides, by touching the highlighted word, they could read the annotations as previously specified for the four treatment groups. After finishing reading the story, each participant moved on to take the immediate posttest and was not allowed to go back to the reading passage. The participants were required to read four reading passages so this procedure was repeated four times. One month later, all participants were asked to take the pen-and-paper delayed posttest. The procedures of the experiment were as follows (See Table 1): Table 1. Procedures of the Experiment Week. Class activities. Test. Week 1 Learning operations on the PDA Week 2. pen-and-paper pretest st. Week 3 Reading 1 reading passage. 1st immediate posttest. Week 3 Reading 2nd reading passage. 2nd immediate posttest. Week 3 Reading 3rd reading passage. 3rd immediate posttest. Week 3 Reading 4th reading passage. 4th immediate posttest. Week 7. pen-and-paper delayed posttest. 32.

(42) 3.5 Data Analysis The data was collected and analyzed as follows. The pretest has thirty test items and the highest point is 30 with each correct answer counted one point. As for the immediate and delayed posttests, the maximum possible score for word production in these two posttests was 40 points. In each of the four reading passages, there were two types of test items—definition questions and sentence cloze questions—in word production part, each of which had five test items. One point was given for one correct answer to a test item. So, if participants managed to answer all of the definition questions and all of the sentence cloze questions, they would be given 10 points for each reading passage and 40 points in total for four reading passages. As far as word recognition was concerned, the maximum possible score was 20 points. In each of the four reading passage, there were five match questions in word recognitioon part. If participants managed to match all of the target words with their Chinese or English equivalents, they would be given five points for each reading material and 20 points in total for four reading materials. After the results obtained from the three tests were tallied and recorded, the one-way ANOVA was applied to the pretest to examine whether there was a significant difference among the four groups in terms of their vocabulary knowledge prior to the experiment. Then the two-way ANOVA was applied to analyze the results of the word production part and the word recognition part in the immediate posttest and in the delayed posttests respectively to examine whether there was a significant different effect between dual-language annotation groups and between dual visual aids on participnats’ word production and on their word recognition in the immediate posttest and in the delayed posttest and, at the same time, to verify whether there were significant interaction effects between language types and visual aids. The null hypothesis was that the results of the four 33.

(43) groups would be the same. The expected outcome was that the two groups given L1 glosses and L2 glosses would have different effects on learning action verbs and that still pictures and animated pictures would have different effects on learning action verbs in terms of word production and word recognition in the two parallel posttests.. 34.

(44) CHAPTER FOUR RESULTS AND DISCUSSIONS This chapter includes two sections. In the first section, the results of data analysis are reported, including pretest results and posttest results, the latter of which separate analyses are done on word production and word recognition in immediate and delayed posttests. In the second section, findings based on the results in the first section are discussed to further explain the effectiveness of different combinations of multimedia annotations on PDAs. 4.1 Results In this section, analyses of data are reported as follows. First, in the results of the pretest, participants’ pre-knowledge of the target words are revealed. Then, the scores of the posttests, separated into two parts—word recognition and word production, are reported to make clear whether combined annotations have different effects on word production and word recognition. In each of these two parts, the scores of the two parallel posttests—immediate and delayed posttests—are examined so that participants’ performances immediately after the experiment and one month after the experiment can be investigated. 4.1.1 Pretest Results As mentioned in Chapter 3, the pretest was designed to assess the participants’ vocabulary knowledge about the target words prior to the experiment. The results of the pretest analyzed by one-way ANOVA ensured the equivalence of the four groups’ pre-knowledge of the target words (F (3, 117) = 1.517, p > .05), as shown in Table 2,. 35.