諧音法對英語單字的學習與記憶影響效益研究－以台北市一所國中為例

全文

(1)國立臺灣師範大學英語學系碩. 士. 論. 文. Master’s Thesis Department of English National Taiwan Normal University. 諧音法對英語單字的學習與記憶影響效益研究－以台北市一所國中為例 The Effects of the Mnemonic Keyword Method on the Acquisition of Novel Vocabulary: A Case Study of a Junior High School in Taipei City. 指導教授：劉宇挺博士 Advisor: Dr. Yeu-Ting Liu 研究生：吳芷嫣. 中華民國一百零六年八月 August 2017.

(2) 摘要第二語言學習的研究者常指出，諧音法能夠幫助學習者記憶單字。然而，針對如何有效減緩學習者遺忘單字的速度之研究仍相當缺乏。因此，本研究旨在探討如何減緩諧音法學習者的遺忘速度，以及文字或圖像何者為更佳的視覺輔助。本研究包含為期五周的實驗、個別訪談，以及研究者在課堂中的觀察紀錄，研究對象為四個班共 111 名八年級生。研究結果顯示調整諧音法的教學順序能夠減緩學生對單字的遺忘速度，且文字為更佳的視覺輔助。總結而言，此研究希望能解釋諧音法如何促進學生單字學習並如何減緩諧音法的遺忘速度，英語教師也可藉此進一步了解在教室中使用諧音法的相關資訊。關鍵字：視覺輔助、遺忘速度、諧音法. i.

(3) ABSTRACT Researchers in second language learning have often suggested that the keyword method would likely facilitate learners’ vocabulary learning. However, research on how to effectively reduce the steepness of the forgetting rate is scant. Therefore, this study explores how to make the acquired vocabulary more robust and investigates which encoding cue, word or picture, facilitates learners’ later retrieval. This research involved a five-week experiment, several interviews, and the researcher’s classroom observations. Four classes of 111 eighth graders were recruited for the study. The results of this study revealed that the adjusted teaching sequence of the keyword method made the acquired vocabulary more robust, and that word encoding cues worked better than picture encoding cues. To conclude, this study may be of importance in explaining how the keyword method facilitates junior high school students’ English vocabulary learning, as well as in providing English teachers with a better understanding when adopting the keyword method in their classrooms. Key words: encoding cue, forgetting rate, the keyword method. ii.

(4) ACKNOWLEDGEMENTS I owe a debt of gratitude to many people whose devotion has been crucial to my success in completing this thesis. First, I would like to thank the Department of English at National Taiwan Normal University (NTNU) for its support, particularly the professors during the summer programs. Their sharing inspired me, not only in writing this thesis but also in my teaching. I would also like to thank all of my classmates at NTNU, whose constant encouragement cheered me up many times. I would like to express my profound gratitude to my advisor, Dr. Yeu-Ting Liu, who has given me extraordinary guidance throughout this thesis. It would not have been possible to complete this thesis without his full support and valuable advice. In addition to his patient guidance of my thesis writing, Dr. Liu is also an important mentor to me. In his class, he always urged us to walk out of our comfort zone and always have the faith to believe in ourselves. His words affected me greatly. Moreover, I would like to express my sincere appreciation to the other members of my dissertation committee, Dr. Mei-Zhen Wu from NTNU and Dr. Mae-Han Chen from National Taipei University of Education, for their attendance at my oral defense and for giving me valuable suggestions and comments. This thesis benefited greatly from their valuable advice. I would also like to thank my proofreaders, Andrew Stromme, Crystal Hsu, iii.

(5) Anya Lin, and Jo Yu, for reading every word in this thesis carefully and giving me many valuable suggestions. I want to thank them for helping me willingly and spending time discussing every detail with me. Finally, I want to express my deepest gratitude to my parents, who encouraged me to pursue my dreams and comforted me whenever I felt anxious and desperate.. iv.

(6) TABLE OF CONTENTS 摘要................................................................................................................................. i ABSTRACT ...................................................................................................................ii ACKNOWLEDGEMENTS ......................................................................................... iii TABLE OF CONTENTS ............................................................................................... v LIST OF TABLES ..................................................................................................... viii LIST OF FIGURES ...................................................................................................... ix CHAPTER ONE INTRODUCTION ........................................................................ - 1 Motivation and Background .............................................................................. - 1 Rationale and Purpose of the Study .................................................................. - 3 Theoretical Basis of the Study .......................................................................... - 4 Research Questions ........................................................................................... - 5 Organization of the Thesis ................................................................................ - 5 CHAPTER TWO LITERATURE REVIEW ............................................................ - 7 The Immediate Effectiveness of the Keyword Method .................................... - 7 Retention Remains an Unresolved Issue ......................................................... - 13 Changing the Procedure of the Keyword Method as a Possible Solution ....... - 23 Word or Picture Encoding Cue May Be Influential ........................................ - 27 Research Questions ......................................................................................... - 33 CHAPTER THREE METHODOLOGY ................................................................ - 34 Participants ...................................................................................................... - 34 Design of the Study ......................................................................................... - 36 Materials .......................................................................................................... - 47 Procedures for Data Collection ....................................................................... - 49 Assessments .................................................................................................... - 50 Data Analysis .................................................................................................. - 52 v.

(7) CHAPTER FOUR RESULTS AND DISCUSSION .............................................. - 53 The Adjusted Teaching Sequence ................................................................... - 53 Word Encoding Cues vs. Picture Encoding Cues ........................................... - 54 Results of the Delayed Vocabulary Tests ....................................................... - 55 Homogeneity of the four groups .............................................................. - 55 The result of the delayed test of all target words ..................................... - 55 The results of Delayed Tests 1 and 2 ....................................................... - 59 The results of Delayed Test 3 .................................................................. - 61 The results of Delayed Test 4 .................................................................. - 64 An Overall Observation and Reflection .......................................................... - 67 The link between keywords and definitions ............................................ - 67 The failure of picture encoding cues ....................................................... - 70 Participants’ attitudes toward the keyword method ................................ - 72 Discussion of the Teaching Sequence ............................................................. - 74 The adjusted teaching sequence .............................................................. - 75 The traditional teaching sequence ........................................................... - 77 Discussion of the Encoding Cues .................................................................... - 79 Picture encoding cues failed to facilitate long-term memory .................. - 80 The effectiveness of word encoding cues needs long-term tracking ....... - 82 CHAPTER FIVE CONCLUSION .......................................................................... - 84 Summary of the Major Findings ..................................................................... - 84 Adjusted teaching sequence facilitates later retrieval.............................. - 84 Keywords worked as mnemonics during later retrieval .......................... - 85 Pictures were not helpful in reducing the phonological similarity effect - 85 Word encoding cues made the novel L2 words more robust .................. - 86 Keywords can be used to learn other subjects ......................................... - 86 Pedagogical Implications ................................................................................ - 86 vi.

(8) Limitations of the Study .................................................................................. - 88 Suggestions for Future Research ..................................................................... - 90 REFERENCES ....................................................................................................... - 93 APPENDIXES ...................................................................................................... - 100 Appendix A. Vocabulary Chosen from Image System for Memorizing English Words ........................................................................................................ - 100 Appendix B. Questions Asked During the Interviews .................................. - 102 -. vii.

(9) LIST OF TABLES Table 1. Group’s Genders and Learning Conditions ............................................... - 35 Table 2. Test of Homogeneity of Variances (based on the result from Test 1) ....... - 55 Table 3. ANOVA for the Delayed Test of All Target Words ................................... - 57 Table 4. Descriptive Statistics of the Delayed Test of All Target Words ................ - 58 Table 5. Multiple Comparisons for the Delayed Test of All Target Words ............. - 59 Table 6. ANOVA for Delayed Test 1 ...................................................................... - 60 Table 7. ANOVA for Delayed Test 2 ...................................................................... - 60 Table 8. ANOVA for Delayed Test 3 ...................................................................... - 62 Table 9. Descriptive Data for Delayed Test 3 ......................................................... - 62 Table 10. Multiple Comparisons for Delayed Test 3 .............................................. - 63 Table 11. ANOVA for Delayed Test 4 ..................................................................... - 65 Table 12. Descriptive Statistics for Delayed Test 4 ................................................ - 66 Table 13. Multiple Comparisons for Delayed Test 4 .............................................. - 66 -. viii.

(10) LIST OF FIGURES Figure 1. Samples of Microsoft PowerPoint slides showing pictures of the target word (frugal, 節儉的) and its keyword (fermented soybean curd,豆腐乳 or 腐乳 in Chinese) .............................................................................................. - 37 Figure 2. Samples of Group A (phonological encoding followed by L1 semantic encoding cues) ........................................................................................ - 39 Figure 3. Samples of Group C (phonological encoding followed by picture semantic encoding cues) ....................................................................................... - 41 Figure 4. Samples of Group B (picture semantic encoding followed by phonological encoding cues) ...................................................................................... - 44 Figure 5. Samples of Group D (L1 semantic encoding followed by phonological encoding cues) ...................................................................................... - 45 Figure 6. Samples of Microsoft PowerPoint slides showing pictures of the target word (frugal means 節儉的 in Chinese) and its keyword (fermented soybean curd, 豆腐乳 or 腐乳 in Chinese) ...................................................... - 49 Figure 7. Administration timing of the learning sessions and the tests in the present study ....................................................................................................... - 50 Figure 8. The performance of the four groups throughout the delayed tests .......... - 54 -. ix.

(11) CHAPTER ONE INTRODUCTION. Motivation and Background Imagine a Taiwanese classroom during an English class. Instead of simply asking students to memorize words, the teacher uses the keyword mnemonic method to help students with their second-language (L2) vocabulary learning. She invited her students to associate the target (to-be-learned) word “feed” with the first-language (L1) keyword 肥的 fei de due to their phonological similarity. Following the above phonological association, the teacher proceeded to the second step of mnemonic instruction: “Imagine there is a bird. If you keep feeding it, it will become 肥的 ‘fat’, so feed means 餵食. The second step of mnemonic instruction, as can be seen from the above example, focuses on the semantic level. In response to the semantic association, a less successful student replied, “Oh, that stinks!” Despite this student’s grumbling, during the vocabulary quiz the next day, when he heard the word “feed,” he muttered “肥的、肥的…” and wrote down the correct Chinese translation 餵食. The above anecdote shows that even if the keyword method is not well-received by all students, it is indeed helpful in helping students with their later retrieval.. -1-.

(12) The teacher used the keyword mnemonic method invented by Atkinson (1975) because the method is suggested having an immediate advantage in vocabulary acquisition. Nation and Webb (2011) indicated that the keyword method “improves learning by at least 20 percent or more in comparison with learning in context, imagining the meaning of the word, rote learning, using pictures, and learning with several synonyms” (Nation & Webb, 2011, p. 47). In Levin, Levin, Glasman, and Nordwall’s (1992) study, the participants in the keyword group significantly outscored those in the comparison group. Carney and Levin (1998) found a significant advantage of the keyword method over the rote repetition method. Siriganjanavong (2013) reported the effectiveness of the keyword method on both immediate and delayed tests. Davoudi and Yousefi (2016) utilized the keyword method to teach vocabulary to teenagers and found that it was more effective than the traditional method. The immediate advantage of vocabulary acquisition with the keyword method is theoretically motivated by the dual-coding theory (Paivio & Desrochers, 1981). The dual-coding theory suggests that knowledge is encoded in two parts: verbal association and visual imagery, and “the keyword method enhances learning and recall because the method uses both the verbal [phonological] system and the image [semantic] system” (de Groot & van Hell, 2009, p. 11). Thus, when the student heard -2-.

(13) the target word “feed,” he first recalled the keyword 肥的 due to the acoustic similarity, and then the image of a fat bird being fed emerged in his mind. Hence, he remembered the Chinese translation 餵食 with the aid of both phonological and semantic associations.. Rationale and Purpose of the Study While the immediate effectiveness of the keyword method in vocabulary acquisition is generally recognized, its effect in retention remains an unresolved issue. Some researchers have claimed that students who used the keyword method forgot the new vocabulary sooner than mixed methods: contextual clues, word structure analysis, and antonyms (Siriganjanavong, 2013), and rote rehearsal (Wang, Thomas, & Ouellette, 1992). Carney and Levin (1998) reported that the significant advantage of the keyword method vanished within one week when the participants were not immediately tested. Wang et al. (1992) argued the reason why the immediate advantage of the keyword method vanished in later retrieval was because the instructed keywords may become a hindrance when learners recall the target words. They suggested that the provided episodic cues that benefit immediate performance may be forgotten over time, so even if participants remember the keyword, they fail to reconnect the keyword to the meaning of the target word. -3-.

(14) Theoretical Basis of the Study The hindrance caused by the keywords may due to the phonological similarity effect (Conrad & Hull, 1964). The phonological similarity effect occurs when a series of words are acoustically similar; confusion may result and therefore impair recall. Hence, this study aims to prevent the phonological similarity effect by letting students understand the meaning of the target word first rather than traditionally linking the target word to a keyword. To convey the meaning, the teacher will give either a vivid picture or the L1 translation of the target word, after which the teacher will instruct students who receive the translation to form a mental image related to the meaning. Finally, the teacher will connect the image to the target word. Besides examining whether preventing the phonological similarity effect benefits later retrieval, this study will also examine which stimuli, form [L1 translation] or semantic [vivid picture], better prime the target words. According to the revised hierarchical model of Dufour and Kroll (1995), providing a translation is better because the L2 lexical link to L1 is stronger than the L2 lexical to semantic link. However, Johnson (2015) found that the students who received picture stimuli not only reacted faster than those who received the semantic prime but also made fewer errors. McGivern and Levin (1982) argued that “different variations of the keyword method do interact with children’s abilities” (McGivern & Levin, 1982, p. 11), and -4-.

(15) this study seeks to examine which stimuli lead to better later performance for junior high school students. In light of the aforementioned issues in existing research, the study aims to examine which stimuli may lead to better performance—vivid illustrations [semantic] or translation [verbal], and whether preventing the phonological similarity effect benefits students’ later retrieval.. Research Questions The questions below will be investigated: Q1 Will diminishing the phonological similarity effect facilitate later retrieval of vocabulary? Q2 Will students’ performances be different when provided with different encoding cues: picture or translation?. Organization of the Thesis The thesis is organized as follows. Chapter one provides background information and discusses the effectiveness of the keyword method and the unresolved issue of retention and its possible solutions. Chapter two will provide a literature review of the research topics, including (a) the phonological similarity effect, and (b) form and semantic stimuli in the process of second language learning. Chapter three details the methodology adopted in this study. Chapter four will present the -5-.

(16) results and discussion. Chapter five will provide a summary of the major findings and offer pedagogical implications and suggestions for future research.. -6-.

(17) CHAPTER TWO LITERATURE REVIEW This chapter reviews the literature relevant to investigating a possible solution to the retention problem of the keyword method. First, the procedure and theoretical framework of the keyword method will be briefly reviewed. Second, the issue of the retention problem will be discussed along with several studies that seek to prolong the immediate effectiveness of the keyword method. Third, possible solutions will be provided based on the findings of the studies.. The Immediate Effectiveness of the Keyword Method Atkinson (1975) invented the keyword method as a mnemonic strategy for vocabulary learning. The keyword method aims to bridge the gap between two languages with an L1 keyword and a mental image. Nation (2008) indicated that the keyword method is effective because it “makes learners process more than one feature of a word” (Nation, 2008, p. 112). When implementing the keyword method, the teacher will first present learners with the to-be-learned target word (e.g., feed), and then present the learners with an L1 formal associate (i.e., the keyword) that is reminiscent of certain phonological features of the target word (e.g., 肥肥的 fei fei de; this sound of this L1 keyword is reminiscent of the initial and final sounds of the target word, namely /f/ and /d/). Next, to entrench the linking between the keyword -7-.

(18) and the target word, which only resemble each other in certain acoustic features ([f] and [d]), learners need to form a mental image in which the meanings of the (L1) keyword (fat) and target word (feed) intersect (e.g., a fat bird being fed). Throughout the above two-step mental association process, the keyword neatly bridges the gap between the formal and semantic information of the target word. When learners later encounter the target word (feed), they can recall the keyword to be reminded of acoustic features of the target word ([f] and [d]→肥肥的) and recall the mental image to evoke semantic information about the target word (a fat bird being fed) (Atkinson, 1975). Accordingly, throughout the two-step association process, the target word is processed both linguistically and visually; learners are led to build up dual memory traces of the phonological and semantic codes of the target word—a pedagogical practice reminiscent of and motivated by the dual-coding theory (Paivio, 2013). According to the dual-coding theory, information is stored in two mental systems, verbal and nonverbal. The verbal system “contains visual, auditory, articulatory, and other modality-specific verbal codes,” whereas the nonverbal system “includes modality-specific images for shapes, environmental sounds, actions, skeletal or visceral sensations related to emotion, and other nonlinguistic objects or events” (Clark & Paivio, 1991, p. 151). The two systems may be internally connected but can -8-.

(19) function independently. For example, when we see a book, we can use the verbal code to tell everyone that that is a “book.” We can connect the image to its verbal code. Also, when we hear or see the word “book,” we can connect the verbal code to the mental image stored in our mind. Thus, when a new word is learned with the keyword method, the retrieval path can be multiple (verbal, visual, or both) because information is encoded both verbally (through the keyword) and nonverbally (through the construction of the mental image). Throughout the above dual encoding process, we can later retrieve the meaning of the word by either linking the new word to its L1 lexical counterpart (the verbal path) or recalling the meaning via mental image (the nonverbal path). The immediate effectiveness of the keyword method has been verified and explored by numerous studies. For example, Ashouri and Moghadam (2015) found a significant advantage of the keyword method on both immediate and three-week delayed tests. Their participants were EFL learners at pre-intermediate level. The immediate advantage of the keyword method was still observed three weeks later. This implied that the effectiveness of the keyword method lasted for at least three weeks. Davoudi and Yousefi (2016) also found the superiority of the keyword method over traditional ones. Their participants were EFL high school students. The results revealed that the keyword learners outperformed on both the immediate and the -9-.

(20) delayed tests. They stated the advantage of the keyword method may result from the connection between L1 and L2 and vivid visual stimuli (Davoudi & Yousefi, 2016). Likewise, Baleghizadeh and Ashoori (2010) reported that the keyword method led to better recall than rote learning due to the elaborative technique, namely, the formation of mental images. The immediate effectiveness of the keyword method was also reported by Little and Kobayashi (2015), who found that the keyword method helped college students to remember difficult English science vocabulary. They conducted a nine-week training for Japanese university science students, during which the participants learned six different vocabulary learning strategies: vocalization, imagery strategies, writing rehearsal, association, mnemonics (keyword method), and word cards. After the training, students had to complete a questionnaire to express their thoughts about the strategies they learned. Some students reported that the keyword method helped them to recall the form and meanings of vocabulary. For example, a low proficiency student stated that the keyword method allowed him to remember the target words better because by using the method, the words made a strong impression on him. Another low proficiency student thought this method suited for science vocabulary because mental images allowed him to remember the words better. A high proficiency student also reported that the keyword method made difficult words easier - 10 -.

(21) to recall due to the acoustic similarity to L1. However, this may merely be the learners’ perception, because no performance data was involved. While some students agreed on the immediate effectiveness of the keyword method, most participants expressed an unwillingness to use it due to the extra effort. Little and Kobayashi (2015) attributed the lack of consensus to two reasons. First, learners need time to come up with an L1 keyword. Second, beginners may have difficulty in forming mental images connected with the science terminology. No solution was provided for the two problems in the study. The researchers concluded that the keyword method seemed to be effective but also the least practical. The extra effort mentioned by Little and Kobayashi (2015) may be prevented if keywords and vivid images were provided by teachers. McGivern and Levin (1982) explored whether learners’ performance would be improved when keywords and pictures were provided by the researchers. In their study, they divided EFL learners into three keyword groups: fully-structured, semi-structured, and unstructured. For the fully-structured group, learners received the keyword, picture, and L1 definition. For the semi-structured group, learners only received the keyword and L1 definition; they needed to form mental images by themselves. For the unstructured group, learners only received the target word and L1 definition; they were taught and encouraged to use the keyword method. In other words, they needed to generate keywords and - 11 -.

(22) mental images. The data revealed that all keyword groups significantly outperformed the control group (learners could use any strategy they preferred) and the fully-structured group outperformed the other two keyword groups. This result suggests that the keyword method was effective with or without keywords and pictures being given to the learners. Moreover, it could be more advantageous when both keywords and pictures were provided. Though they reported better performance on the immediate test, they did not track whether the superiority persisted over time. Some researchers explored whether the immediate advantage of the keyword method persisted over time when learners generated their own keywords. For instance, Levin et al. (1992) conducted four experiments to examine the immediate and delayed effect of the keyword method. Their participants were EFL middle school students, who were divided into two groups: keyword and context. The keyword learners only received the target words and L1 definitions. That is, they needed to generate their own keywords and form their own mental images. The context learners made L2 sentences with the target words. The results revealed that the keyword learners outperformed the context learners. To track if the advantage persisted, the researchers asked the participants to take the delayed test 10‒14 days after the treatment. At that time also, the keyword learners statistically outscored the context learners. They reported that the keyword groups outperformed on both immediate and delayed tests. - 12 -.

(23) This suggests that the immediate effectiveness of the keyword method persists for at least two weeks. In this study, Levin et al. (1992) also explored whether generating keywords individually or in groups affects the learning outcome. Keyword learners generated their own keywords individually, in pairs, or in heterogeneous groups. Likewise, context learners made their sentences individually, in pairs, or in heterogeneous groups. All participants took two kinds of tests: definition recall and story comprehension. The results revealed that all keyword groups (individual, paired, and heterogeneous) outperformed the context groups on definition recall and story comprehension tests. This implies that grouping does not affect the effects of generating keywords: Keyword learners outperformed context learners whether they were grouped or not. The findings of previous research collectively suggest that the keyword method successfully helps learners remember the form-meaning knowledge of the L2 target words and strengthens form-meaning encoding (Karpicke & Smith, 2012).. Retention Remains an Unresolved Issue Though the immediate effectiveness of the keyword method is supported by many studies, Wang et al. (1992) claimed that the performance on the delayed test may be improved by the immediate test. That is, Wang et al. (1992) suggested that the - 13 -.

(24) immediate test may serve as a potential learning trial and thereby boost learners’ performance on the delayed test. They speculated that the keyword learners’ superiority may not last over time if they were not tested immediately. To explore whether an immediate test boosts learners’ later performance, Wang et al. (1992) conducted an experiment using a between-subjects design. The purpose of the research design was to prevent any effects of the immediate test by having half of the participants take the immediate test and the other half the delayed test. The participants were adult beginners who learned the vocabulary through either rote repetition or the keyword method. The researchers then compared the scores on the immediate and delayed tests. The results revealed that the keyword learners forgot newly learned words faster than the rote learners did. The keyword learners significantly outperformed the rote learners on the immediate test, but among the participants who only took the delayed test, no significant difference was found. They concluded that the keyword learners forgot things faster than the rote rehearsal learners. Later, Wang and Thomas (1995) conducted another study to test if the faster forgetting rate also occurred when comparing the keyword method to context learning. The participants learned vocabulary either with the keyword method or in context. The keyword learners were given keywords and then asked to form mental images, - 14 -.

(25) while context learners were presented flashcards with a two-sentence context and the L1 translation of the target word. Participants were tested either immediately or after a two-day delay. The keyword learners in the study outperformed the context learners on the immediate test; however, on the delayed test the keyword learners performed worse than the context learners. The results thus revealed that the effectiveness of the keyword method vanished within two days. The reason that the keyword method could not maintain its advantage over time was not explored in the study. The faster forgetting rate of the keyword method was confirmed by Carney and Levin (1998). The participants in their study were undergraduates. A faster forgetting rate for the keyword learners over rote learners was also observed in their study. They separated the target words into two categories, some of which were tested on the immediate test and the others not. The results indicated that the tested words were better recalled during the delayed test, and it was found that the advantage of the keyword method was not seen on the delayed test when the target words were not on the immediate test list. This echoes Baddeley’s (2012) assertion—phonological coding is easy to build and the effect is rapid, but phonological coding is also readily forgotten [in the short-term memory]. That is, the immediate advantage of the keyword method was built by the phonological encoding connected to the keyword but rapidly disappeared over time. - 15 -.

(26) Researchers and practitioners have sought to maintain the immediate advantage of the keyword method. Several ways have been used to reinforce the connection between the target word and its semantic link based on the keyword. These techniques include, but are not limited to: encouraging learners to generate their own keywords (Shapiro & Waters, 2005) or form vivid and interactive images (Campos, González, & Amor, 2003, Experiments 1 and 2; Wang & Thomas, 1995, Experiment 1), repetitively providing stimuli of the target words (Karpicke & Smith, 2012; Wang & Thomas, 1995, Experiment 3), and attaching a relevant sound to the keyword (e.g., lamb → “baa, baaing”) in a picture related to the meaning of the target word (Carney & Levin, 1998, Experiment 3). Though many techniques have been examined, none of them led to a significant advantage in later retrieval when learners were not tested immediately. The retention issue remains unresolved. While some researchers have used different techniques to reinforce the link between a target word and its semantic link, Wang et al. (1992) speculated that the problem may stem from forgetting the keywords provided by the researchers. They conjectured that when the keywords were provided by the researchers instead of generated by the learners, they would be forgotten easily. If learners forgot the keyword, it would be unlikely for them to rebuild the mental image or link the image to the meaning of the to-be-learned target word. However, the results revealed that the - 16 -.

(27) participants could still recall most of the keywords during later recall (Wang et al., 1992; Wang & Thomas, 1995). Though the participants could recall the keywords linked to the target words, they failed to reconnect the keyword to the meaning of the target words. Wang et al. (1992) thus claimed that the keywords remembered by the participants may become a hindrance during later recall. They speculated that when learners recalled the keywords to draw on their acoustic similarity to the target words, irrelevant associations may occur. For instance, in their study, their participants were asked to connect the keyword “egg” to the mental image of a church by encoding the keyword as a white, oval hen’s egg. The French word église (church in English) is acoustically similar to the English word egg. Hence, “egg” worked as a keyword and its white figure was connected to the church. However, participants might recall different kinds of eggs (e.g., scrambled eggs) one week later. Though Wang et al. (1992) speculated that the keyword may become hindrance, they did not further investigate the reason causing the hindrance. Wang et al. (1992) believed the acoustically similar keywords may introduce confusion over time and thus impair the participants’ performance. The hindrance of the keywords may be caused by the phonological (acoustic) similarity effect. Lian, Karlsen, and Winsvold (2001) discovered that when words were connected to - 17 -.

(28) irrelevant semantic information, the phonological similarity effect was observed. In their study, they divided words into those with high and low associative values. To define the associative value, they presented a list of words one by one to the participants and asked them to link a word to a semantic associate as fast as possible and then calculated the time to make the connection. If the reaction time of a word was shorter than the mean reaction time, the word was defined as having a high associative value. On the other hand, if the reaction time of a word was longer than the mean reaction time, the word had a low associative value. The results revealed that when the words had a high associative value, the phonological similarity effect was observed. In other words, if a word could be linked easily to irrelevant semantic information, the phonological similarity effect would occur. This echoes Wang et al.’s (1992) speculation. They speculated that the keywords became linked to irrelevant semantic information over time (e.g., eggs linked to scrambled eggs instead of the shape of a mosque) and hindered the retrieval of the meaning of the target words. Nevertheless, Wang et al. (1992) did not explore whether the hindrance was caused by the negative phonological similarity effect. The phonological similarity effect was first observed by Conrad and Hull (1964). They discovered that when learners are provided with a list of acoustically similar words, it is very likely that the words may cause interference and confusion - 18 -.

(29) during subsequent recall. Later, Nimmo and Roodenrys (2005) conducted an experiment to examine if the phonological similarity effect occurred when the students were provided with a list of acoustically similar words. The result revealed that acoustically similar words could not be recalled as well as phonologically dissimilar words. In other words, greater impairment occurred as the similarity in sound increases (Nimmo & Roodenrys, 2005). The phonological similarity effect may explain the failure of the immediate effectiveness of the keyword method to be retained. According to Baddeley (2012), there are two storage systems involved in memory: phonological and semantic. The phonological system mainly involves short-term memory (working memory) and the semantic system involves long-term memory. When the target words are connected to acoustically similar keywords, phonological encoding is involved and the information is stored temporarily and immediately “using a process that is rapid, and requires minimal attention” (Baddeley, 2012, p. 12). This explains the immediate effectiveness of the keyword method. However, when learners need to recall the learned information over time, their performance might be impaired by the phonological similarity effect caused by the keywords. According to Lian et al. (2001), the phonological similarity effect might occur when “the irrelevant lexical clusters and their neighbors” in long-term memory - 19 -.

(30) were activated during recall (Lian et al., 2001, p. 297). In other words, if the target words made the participants recall the keywords over time, this might trigger the phonological similarity effect and thus cause confusion. The confusion caused by the phonological similarity effect impaired the participants’ recall of the semantic meanings of the words. This explains why the keyword learners failed to recall the definitions of the target words over time, even though they could still recall almost all of the target words (Carney & Levin, 1998; Wang & Thomas, 1995; Wang et al., 1992). That is to say, the keyword learners successfully retrieved the phonologically similar keywords but failed to retrieve the definitions due to the negative phonological similarity effect. Some researchers have suggested that by encouraging semantic encoding prior to phonological encoding, it may be possible to make the learned information more robust and hence facilitate long-term retention (Campoy & Baddeley, 2008; Savill, Ellis, & Jefferies, 2016). In Campoy and Baddeley’s (2008) study, they discovered that the phonological similarity effect may be alleviated when the phonologically similar words were encoded semantically. When the participants connected the phonologically similar words with word meanings, the phonological similarity effect was inhibited and the participants’ performance was improved. They stated that “semantic encoding appeared to protect participants from the effects of phonological - 20 -.

(31) similarity” and thus made the learned information more robust (Campoy & Baddeley, 2008, p. 334). In other words, the phonological similarity effect may be alleviated by changing the implementation procedure of the keyword method. Encouraging the semantic encoding prior to the phonological encoding might inhibit the phonological similarity effect caused by the keywords, hence facilitating later retrieval. Moreover, Savill et al. (2016) discovered that newly acquired words were more robust when semantic encoding was built prior to phonological encoding. This implies that if semantic encoding was first encouraged during the learning session, learners’ later performance might be improved. They concluded the semantic information stored in long-term memory may promote the reintegration of missing phonological elements (kept in short-term memory). Hence, encouraging semantic encoding prior to phonological encoding might facilitate the retention of the newly acquired words. In the aforementioned studies, which aimed to retain the immediate effective of the keyword method (Carney & Levin, 1998; Wang & Thomas, 1995; Wang et al., 1992), the participants were given word lists that show target words and keywords simultaneously and were then instructed to directly link the target word to the keyword. The form or semantic meaning of the target word was provided next to allow learners to build mental images linking the keyword to the meaning of the target - 21 -.

(32) word. Throughout the above procedure, a negative phonological similarity effect may occur before semantic encoding is built. When this phonological similarity effect occurs, semantic encoding may be harder because the phonological similarity effect would lead to inaccurate memory traces, which may lead in turn to later recall problems (Baddeley, 2012). Hence, to avoid the negative phonological similarity effect, semantic encoding should be encouraged prior to phonological encoding. In other words, the L1 translation or the meaning-related image should be provided before the keyword association. When students are familiar with the form or semantic meaning, the keyword may cause less interference because the semantic encoding was encouraged prior to the provision of the keyword. In this way, the target word can be recalled with the dual memory traces without the hindrance of the phonological similarity effect. Though previous studies verified that encouraging the semantic encoding prior to phonological encoding led to better performance (Campoy & Baddeley, 2008; Savill et al., 2016), none of the studies connected the issue of the encoding sequence to the keyword method. This study aims to examine if the failure of later retrieval for keyword learners was caused by the phonological similarity effect. If so, preventing such an effect might be a possible solution.. - 22 -.

(33) Changing the Procedure of the Keyword Method as a Possible Solution To prevent the phonological similarity effect resulting from sound similarity between the keywords and the target words, changing the procedure of the keyword method might be a solution, i.e., encouraging semantic encoding of the target word (e.g., L1 translation) prior to phonological encoding. Traditionally, a keyword will be presented right after a target word. When participants accept a set of acoustically similar words, the phonological similarity effect may occur and therefore impair later retrieval. According to Baddeley (2012), separating acoustically similar words with a dissimilar word (i.e., L1 translation or vivid illustration, which shows the semantic meaning of the target word) may diminish or dissipate the phonological similarity effect and “show semantic and other long-term-memory-based effects” (Baddeley, 2012, p. 8). If the phonological similarity effect is inhibited by adjusting the sequence, confusion may be prevented and learners can further maintain long-term retention (semantic encoding). The technique of changing the encoding sequence in the keyword method was tested with undergraduate students in Karpicke and Smith’s (2012) study. Karpicke and Smith (2012) argued that the effect of repeated retrieval practice is more powerful - 23 -.

(34) than repeated study. Vocabulary learning was separated into two periods: study and recall. For the study period, the participants saw the target words and the definitions in pairs on a computer screen. For the recall period, only the target words appeared, and learners typed in the correct definitions. Each study period was followed by a recall period. There were six study/recall periods in total. All of the target words and the definitions were presented in pairs during the first study period. The participants were divided into three groups: drop, repeated study, and repeated retrieval practice. Under the drop condition, a target word would be deleted once it was correctly recalled. That is, if the learners typed in the correct definition during the recall period, the target words would be removed during the following study/recall periods. The learners would not be tested or presented the definition of a correctly recalled target word again. Under the repeated study condition, all target words and their definitions appeared during all six study periods; the learners would keep seeing all target words and their definitions. The correctly recalled target words would be removed from the recall periods; the learners would still repeatedly see the correctly recalled target words and their definitions but would not be tested on them. Under the repeated retrieval practice condition, all the target words and their definitions would appear during all six recall periods. Thus, the learners would need to type in the definitions of. - 24 -.

(35) all target words throughout the six recall periods. However, the learners would not see the definitions of the correctly recalled target words again. In Karpicke and Smith’s (2012) study, the keyword learners were provided a keyword after the first successful recall of a definition. They were encouraged to form vivid mental images regarding the definitions of the target words. The control group was given only the target words and their definitions. The participants were informed that they were to take a one-week delayed test before the study trial. The results revealed no significant difference between the keyword group and the control group under drop, repeated study, or repeated retrieval conditions. The keyword learners might not have outperformed the control learners because of the ceiling effect. In their study, Karpicke and Smith (2012) suggested that at the end of the learning trial, both the keyword group and control group’s performances reached a ceiling. That is, the participants in both the control group and the keyword group could correctly recall almost all of the definitions of the target words. Hence, no significant difference was found on the immediate recall test. On the one-week delayed test, no faster forgetting rate of the keyword groups over the control groups was found among any of the three conditions, which contradicts the claim that the keyword method has a faster forgetting rate than rote learning (Carney & Levin, 1998; Wang et al., 1992). This outcome implied that when keyword learners - 25 -.

(36) were encouraged to do the semantic encoding prior to accepting the keywords, the forgetting rate might be less steep. Though the technique of adjusting the encoding sequence was implemented in the study, the researchers did not compare the performance of the traditional teaching sequence (first keyword then definition) and that of the adjusted teaching sequence (first definition then keyword). Furthermore, though the researchers utilized the adjusted teaching sequence, they did not link the adjusted teaching sequence to the issue of phonological similarity effect. This study aims to compare whether the participants’ performance improve when they are taught in the adjusted sequence. In Karpicke and Smith’s (2012) study, all keyword learners learned the vocabulary in the adjusted sequence. The present experiment will examine whether adjusting the teaching sequence prescribed in the keyword method (i.e., first semantic then phonological) enhances and maintains students’ learning outcomes over time. If the students could learn better through such an adapted practice than through the traditional keyword method on the delayed tests, adjusting the sequence may be the solution to the retention problem commonly seen in keyword method research. Namely, by preventing the phonological similarity effect, learners may be able to maintain the learning outcome of the keyword method.. - 26 -.

(37) Word or Picture Encoding Cue May Be Influential Encouraging learners’ semantic encoding of the target words prior to giving the keywords may be beneficial to later retrieval. Semantic encoding of a target word can be fostered through the following two possibilities: presenting learners with an L1 translation or with vivid images that are conceptually/semantically associated with the target word. Some researchers have explored whether learners’ performance is enhanced when they are provided with vivid illustrations. It was thought that providing vivid illustrations might make the target words more concrete and semantically transparent, and hence reduce the effort in forming mental images. According to Campoy, Castellà, Provencio, Hitch, and Baddeley (2015), “word concreteness had an effect on memory for item identity” (Campoy et al., 2015, p. 10). In other words, the more concrete the words to the learners, the more easily the words can be recalled. For example, Siriganjanavong (2013) found when the keywords and the definitions were presented along with illustrations, the keyword learners outperformed the learners taught with mixed methods. The mixed methods were contextual clues, word structure analysis, and opposite word-pairs. The keyword learners statistically outperformed on both immediate and two-week delayed tests. Also, McGivern and Levin (1983) found the keyword learners provided with keywords and illustrations - 27 -.

(38) performed statistically better than those who simply received keywords or the introduction of the method. The advantage of providing pictures was also found by Levin et al. (1984), who reported that vivid pictures facilitated both beginners and high flyers’ vocabulary learning. Furthermore, Shapiro and Waters (2005) found the imagery level of the target words has a significant effect on both immediate and delayed retrieval. Based on the assumption that vivid images may lead to better immediate and later performance, Dolean (2014) conducted a study to explore if presenting pictures related to keywords may facilitate learners’ performance on definition recall. Note that the pictures related to the keywords have no interaction with the definitions of the target words. The participants were middle school students, who were divided into three groups: keyword, picture, and word. The vocabulary was read aloud by the teacher. The keyword participants were presented with two pictures, one related to the target word and the other to the keyword. Two pictures were simultaneously presented on the same PPT slide; the picture for the definition was bigger than that for the keyword. For the picture group, only pictures related to the definitions of the target words were presented. For the word group, only an L1 translation was shown on the slide. The results indicated that the keyword learners outperformed the other groups. The better performance remained on immediate, one-week delayed, and even - 28 -.

(39) one-month delayed tests; the participants were not informed about the delayed tests in advance. The maintenance of the learning outcome of the keyword method over time may be attributed to how the keyword was presented in this study. According to Jarrold and Citroën (2013), when acoustically similar words were presented with pictures, the phonological similarity effect could be effectively inhibited. In other words, when the target word and the keyword were presented with pictures rather than verbal codes, the phonological similarity effect was prevented, so learners directly accepted the semantic knowledge of the target word through the picture without the hindrance of the phonological similarity effect. The semantic encoding of the to-be-learned target words effectively helped learners’ later retrieval. Though it would appear that providing pictures related to keywords is more powerful than simply providing the L1 translations (Dolean, 2014, Experiment 2), this may be because the participants in the translation group were not instructed about the keyword method. More research should be done to explore if word participants would perform better with instruction in the keyword method. While some researchers agreed that pictures would facilitate vocabulary learning (Dolean, 2014; Levin et al., 1984; McGivern & Levin, 1983; Peters & Levin, 1986), others, such as Wang and Thomas (1992), observed faster forgetting rates for - 29 -.

(40) learners provided with vivid images. The participants who accepted word visual stimuli performed better in recognizing Chinese characters than those who accepted picture stimuli. They attributed the result to encoding variability, assuming that “the mnemonic description accompanying each ideograph may differ from the spontaneous, pre-experimental encoding of individual subjects” (Wang & Thomas, 1992, p. 373). Hence, they speculated that only when the mental images were formed by learners could the effectiveness persist over time. In other words, to maintain the immediate learning outcome, encouraging learners to form mental images rather than providing vivid illustrations might be more beneficial to later retrieval. Considering provided pictures may cause interference, Wang and Thomas (1992) suggested that learners should be encouraged to form their own mental images regarding the word forms, which was confirmed in Campos et al.’s (2003) study. The keyword learners presented with pictures performed significantly worse than the rote learners (Campos et al., 2003, Experiment 3). The researchers conjectured that the reason might be that “drawings do not induce a direct association between the second-language word and the first-language word” (Campos et al., 2003, p. 408). When learners accept pictures as visual stimuli, it might be challenging for them to translate the target words into an L1 definition via the recall of images. Learners may need to transfer the nonverbal code (picture) into verbal code (L1 translation) during - 30 -.

(41) retrieval to complete the word definition test. Hence, directly providing L1 translation as visual stimuli might be more effective than providing pictures related to definition. Providing L1 translation as visual stimuli for EFL learners is theoretically motivated by Dufour and Kroll’s (1995) revised hierarchical model, which indicates that the lexical link from L2 to L1 is stronger than L2 to the conceptual link. The stronger the link, the more easily the information can be connected and retrieved. For instance, when learners see the target word “fire,” recalling its L1 translation might be easier than the presented image of fire. Because the link from L2 to L1 is stronger, learners can recall the L1 translation (form) more easily than the image (semantic). Dufour and Kroll (1995) suggested that beginning EFL learners tend to initially connect the L2 target word to its L1 translation. If the target word is presented with its L1 translation, learners may immediately realize the form-meaning knowledge of the target word. When they need to make the semantic association of the target word, its L1 translation may be easily retrieved due to the stronger link from L2 to L1. Based on the revised hierarchical model, EFL learners’ conceptual link to L2 is weaker than the lexical link from L2 to L1. Luna and Peracchio (1999) further explained that the conceptual link to L2 is weak because learners needed to achieve “a high level of proficiency in their second language so that they rely less on their L1 to gain access to meaning” (Luna & Peracchio, 1999, p. 307). When the definition of the - 31 -.

(42) target word is presented with pictures, learners accept the concept of the target word. Beginning EFL learners may find it difficult to link the concept to the L2 verbal code because of the weaker link. For instance, if the target word “fire” is only presented with a picture, learners may find it hard to connect the image of fire to the L2 verbal code during retrieval. However, they may connect the image of fire to concepts like “hot,” “dangerous,” or “do not touch” based on their prior knowledge. The connection formed by the participants is not directly related to the L1 definition of the target word and being irrelevant may hinder learners’ retrieval. The hindrance of pictures was observed during later retrieval (Wang & Thomas, 1992), and during both immediate and delayed retrieval (Campos et al., 2003). Though some studies claimed that pictures may become a hindrance, other studies suggested providing vivid pictures to the learners helped them retain the learned information over time (Ashouri & Moghadam, 2015; Dolean, 2014, Experiments 1 & 2; Levin et al., 1984; McGivern & Levin, 1983; Rummel, Levin, & Woodward, 2003; Siriganjanavong, 2013). Hence, which learning or encoding cue (word or picture) helps facilitate vocabulary learning remains an unresolved issue. Given the results of the aforementioned studies, the goals of the present experiment were twofold. First, the study described in this thesis aims to examine whether adjusting the teaching sequence of the keyword method—which may - 32 -.

(43) eliminate the phonological similarity effect—helps learners retain their vocabulary gains over time. Second, this study explores whether providing a word prime or picture prime leads to different effects on the (vocabulary) learning outcomes.. Research Questions In light of the aforementioned issues, the research questions are: 1. Will encouraging semantic encoding (L1 translation or pictures related to both L1 translation and keyword) prior to phonological encoding (keyword) benefit the learning and retention of novel L2 words? 2. Will different learning or encoding cues (L1 translation vs. picture) have different effects on the learning and retention of novel L2 words?. - 33 -.

(44) CHAPTER THREE METHODOLOGY In this chapter, the design of the present study will be described. Detailed information of the entire research method will include the background of the participants, the design of the study, the materials, the procedures of data collection, the assessments utilized in this study, and data analysis of this study.. Participants The participants in the present study consisted of 111 eighth graders in a junior high school in Taipei City, recruited from four classes (Class A, Class B, Class C, and Class D). There were 29 students in Class A, 26 in Class B, 28 in Class C, and 28 in Class D. The students recruited to participate in the present study had five English classes per week; each class lasted for 45 minutes. Their English teacher was the researcher in this study. All of them started to learn English when they were third graders in elementary school. Of the students, 55 were male and 56 female, and their mean age was 14 years. During the learning sessions, the four classes were randomly assigned to four groups: Experimental Group A (phonological encoding followed by L1 semantic encoding cues), Experimental Group B (picture semantic encoding followed by phonological encoding cues), Experimental Group C (phonological encoding - 34 -.

(45) followed by picture semantic encoding cues), and Experimental Group D (L1 semantic encoding followed by phonological encoding cues) (see Table 1). The procedures (phonological encoding → semantic encoding) in Groups A and C were consistent with the traditional keyword mnemonics and were described as “traditional” procedures, while the procedures (semantic encoding → phonological encoding cues) in Groups B and D were inconsistent with the traditional keyword mnemonics and were therefore described as “adjusted” procedures.. Table 1. Group’s Learning Conditions Participants. Sequence. Initial Encoding Cue. Secondary Encoding Cue. Group A. 29. Traditional. Phonological. Semantic (Word). Group B. 26. Adjusted. Semantic (Picture). Phonological. Group C. 28. Traditional. Phonological. Semantic (Picture). Group D. 28. Adjusted. Semantic (Word). Phonological. As for the participants’ English proficiency, all four classes were in the normal class grouping (常態分班) based on the entrance exam they took when they entered junior high school. In other words, the four classes had similar proficiency - 35 -.

(46) distribution profiles and all were heterogeneous. Each participant would have no problem understanding the semantic meanings of the target words, for all target words were presented with vivid pictures or L1 translations. The instruction was given in Chinese, the participants’ L1, to make sure all of the participants understood the researcher’s instruction. The reasons that these students were invited to join the present study are as follows. First, students at different proficiency levels could be found in each class. The keyword method was suggested as being effective regardless of students’ proficiency (Peters & Levin, 1986). Hence, the keyword method might be helpful for students at different proficiency levels. Second, the participants were all eighth graders. In other words, unlike ninth graders they were not under the pressure of taking the Comprehensive Assessment Program for Junior High School Students (國中教育會考). Therefore, they would be more willing to learn supplementary materials.. Design of the Study A 2 (encoding sequence: traditional vs. adjusted) × 2 (nature of semantic encoding cue: word vs. picture) design was used, resulting in four experimental conditions. The four classes, which had similar proficiency distribution profiles, were randomly assigned to the four teaching conditions. The participants with different - 36 -.

(47) encoding cues would see different PPT slides (see Fig. 1).. Keyword (L1 Semantic Encoding Cue). Keyword (Picture Encoding Cue). Figure 1. Samples of Microsoft PowerPoint slides showing pictures of the target word (frugal, 節儉的) and its keyword (fermented soybean curd,豆腐乳 or 腐乳 in Chinese). For the traditional sequence groups (Groups A and C), a target word was connected to its phonologically similar L1 keyword first. Before providing the keyword, the researcher read out loud the target word and the participants were then asked to repeat the target word. There were two reasons for reading and asking for repetition. First, by knowing the pronunciation of the target word, the participants could link the target word to the phonologically similar keyword more easily. Second, by reading out the target word, the participants would not mistake the pronunciation of the keyword for the pronunciation of the target word. This was to prevent the - 37 -.

(48) participants from mispronouncing the target words afterwards. In Group A, the participants first saw the PPT slide with only the target word (see Fig. 2). The researcher read the target word aloud to present the participants with the pronunciation of the word, and then told the participants what the keyword was. Next, the researcher connected the target word to the phonologically similar keyword orally (e.g., Please connect the word “frugal” to its keyword 豆腐乳). The researcher instructed the participants to pay attention to the phonological similarity between the target word and the keyword (e.g., The pronunciation of “frugal” sounds like 豆腐乳) to encourage them to build the phonological encoding. The researcher also told the participants that the keywords worked as mnemonics. After building the phonological encoding, the semantic meaning of the target word was presented with the L1 translation. The L1 translation appeared next to the target word. The researcher then directed the participants to form a vivid mental image to connect the keyword to the L1 translation (e.g., 吃豆腐乳過日子的人很節儉, People who eat fermented soybean curd every day live a frugal life. “Frugal” means 節儉).. - 38 -.

(49) Connect the target word “frugal” to 豆腐乳 without the image of the keyword.. Connect 豆腐乳 to the L1 translation of “frugal.” Group A Figure 2. Samples of Group A (phonological encoding followed by L1 semantic encoding cues). In Group C, the picture of the keyword was presented prior to the picture of the target word. This was to avoid the participants building the semantic encoding prior to the phonological encoding by looking at the picture of the semantic meaning of the target word. The participants first saw the PPT slide with the target word and - 39 -.

(50) the picture of the keyword (see Fig. 3). The researcher read out the target word, and then directed the target word to the phonologically similar keyword. The keyword was presented with a picture. The researcher instructed the participants to build the phonological encoding by connecting the target word to the keyword (e.g., Look at the picture. This is 豆腐乳. The pronunciation of “frugal” sounds like 豆腐乳. 豆腐乳 is a keyword to help you remember the semantic meaning of “frugal.” Notice the phonological similarity). After that, the picture of the semantic meaning was presented. The researcher asked the participants to link the keyword to the semantic meaning of the target word by directing to the picture of the semantic meaning (e.g., Look at the bigger picture. The money is tied tight. Hence, the money should be spent sparingly. Frugal means 節儉. 吃豆腐乳過日子的人很節儉. People who live a frugal life eat fermented soybean curd every day).. - 40 -.

(51) Connect the target word “frugal” to 豆腐乳 with the image of the keyword.. Connect 豆腐乳 to the semantic meaning of “frugal.” Group C Figure 3. Samples of Group C (phonological encoding followed by picture semantic encoding cues) - 41 -.

(52) As shown in Fig. 3, the picture of the semantic meaning is bigger than the picture of the keyword. Based on Lawson and Hogben (1998), the picture of the definition has to be bigger than the picture of the keyword. There are two reasons for making the sizes different. First, making the picture of the definition bigger encourages semantic encoding. Second, it allows the participants to distinguish the picture of the semantic meaning from the picture of the keyword. To examine whether avoidance of the negative phonological similarity effect may improve learning outcomes during later retrieval, the other two groups (Groups B and D) were instructed in the adjusted sequence (encouraging semantic encoding prior to phonological encoding). Before providing the semantic meaning, the researcher read the target word aloud, and the participants were then asked to repeat the target word, after which the researcher immediately presented the semantic meaning. This was to encourage the semantic encoding of the target word. The reason the researcher asked for repetition was to make sure the participants grasped the correct pronunciation of the target word. By reading the target word aloud, the participants would not mistake the pronunciation of the keyword as the pronunciation of the target word. This was to prevent the participants from mispronouncing the target words afterwards. In Group B, the participants saw a PPT slide with the target word and the - 42 -.

(53) picture of the semantic meaning (see Fig. 4). The researcher read the target word to let the participants know the pronunciation, then the researcher directed the target word to the picture of the semantic meaning. The semantic meaning was explained by the researcher (e.g., Look at the picture. The money is tied tightly. Hence, the money should be spent sparingly. In other words, “frugal” means 節儉). After knowing the semantic meaning of the target word, the participants were instructed to link the semantic meaning to the keyword, which was presented with a picture after the explanation of the semantic meaning of the target word. The reason the two pictures did not appear simultaneously was to prevent the participants from building the phonological encoding prior to the semantic encoding. The keyword was presented both with a picture and orally by the researcher (e.g., Look at the smaller picture. This is 豆腐乳). Then the participants were instructed to form a vivid mental image to connect the semantic meaning to the keyword (e.g., Frugal means 節儉. 吃豆腐乳過日子的人很節儉. People who live a frugal life eat fermented soybean curd every day).. - 43 -.

(54) Connect the target word “frugal” to its semantic meaning with a picture.. Connect the semantic meaning of “frugal” to the keyword 豆腐乳 with the picture of the keyword. Group B Figure 4. Samples of Group B (picture semantic encoding followed by phonological encoding cues) - 44 -.

(55) For Group D, the participants saw a target word and its L1 translation (see Fig. 5). After the participants received the definition of a target word, the keyword was orally provided. Then the participants were instructed to form a vivid mental image to connect the definition to the keyword (e.g., Frugal means 節儉. 吃豆腐乳過日子的人很節儉. People who live a frugal life eat fermented soybean curd every day).. Connect the target word “frugal” to its L1 translation.. Connect the L1 translation 節儉的 to the keyword 豆腐乳 without the picture of the keyword. Group D Figure 5. Samples of Group D (L1 semantic encoding followed by phonological encoding cues) - 45 -.

(56) For the design of the PPT slides, there were two reasons the L1 translations were not shown in the slides for picture groups (Groups B and C). First, this study aims to examine if pictures may become hindrance during delayed retrieval. According to Wang and Thomas (1992), pictures provided by the researcher may become a hindrance over time. They suggested the researcher should only provide the target word and its L1 translation. In other words, the participants should generate their own mental images without seeing any pictures. By comparing the performance between the picture groups and the word groups, the result may reveal whether provided pictures would become hindrance over time. Second, there are already three elements (a target word, a keyword picture, and a picture of the semantic meaning of the target word) on one PPT slide. Presenting L1 translations on PPT slides might result in information bombardment. The participants in the picture groups already had one more element on the slide (a target word and two pictures) than the word groups did (a target word and its L1 translation). Adding one more element would give the picture groups twice as many elements as the word groups. In addition, the L1 translations would be presented orally by the researcher along with the description of how to form mental images. In other words, the participants would still be able to understand the meaning of the target word in their L1. Providing L1 translations on PPT slides of picture groups - 46 -.

(57) might have been redundant and unnecessary. Each week, only five target words were presented to the participants. Based on Farley, Ramonda, and Liu (2012), under normal teaching conditions teachers should teach fewer target words each session and later evaluate learners’ later retrieval. They suggested that the total amount of to-be-taught words should be less than twenty-four. After teaching the five target words, cued recall was tested after a seven-day delay. Because the students needed time to prepare for their final exam, the experiment lasted for only one month. That is, within one month, the participants were presented twenty target words total (five per week). Four delayed tests were implemented because of the number of the to-be-taught words. For each delayed test, only the five target words learned a week before were tested. After the four delayed tests, all target words were reviewed. Seven days later, the participants took the last delayed test, during which all target words were tested. There were five delayed tests in total. During the tests, all target words were read loudly by the researcher. The participants needed to write down the L1 definitions of the target words to test if the participants remembered the form-meaning knowledge of the novel L2 words.. Materials Twenty-five target words were selected from Image System for Memorizing English Words (Yao, 1992), Part 3 (英语单词形象记忆法, Part 3 谐音形象记忆法) - 47 -.

(58) (see Appendix A). The target words had either two or three syllables. There are two reasons the target words were chosen from this book. First, as the vocabulary in the book was more difficult than that in the participants’ English textbooks, the participants were less likely to have learned the target words previously. Second, the vocabulary in the book was presented with Chinese keywords. The description of how to connect the keyword to the definition was also provided. Hence, it would be easier to build the connection with the keywords provided. Each target word was read by the researcher aloud in class before the four learning sessions began. The participants could raise their hands if they thought they knew the word. Then the researcher would ask them to provide its definition individually (if there were more than one participant raising their hands). Once a word was correctly answered by any participant, the target word would be eliminated. This was to ensure none of the target words had been learned by the participants before the sessions. Finally, twenty words were chosen as the target words in this experiment. All target words were presented with PPT slides. For the participants who received the word encoding cue, they saw the English target word and its L1 definition (see Fig. 6). For the participants who received the picture encoding cue, they saw the English target word and two pictures. One was related to the semantic meaning of the target word; the other to the keyword. The picture of the semantic - 48 -.