Papers by Chikako Nakagawa
The present study examined how Japanese EFL readers revise their situation models, fbcusing on tw... more The present study examined how Japanese EFL readers revise their situation models, fbcusing on two factors: (a) learners' reading sub-ski11s, and (b) the causal structures of stories. A total of 76 Japanese undergraduates read stories which required them to revise their interpretation of the stories during reading (i,e,, situation model revision). then they perfbrmed sentence verification and recall tasks, The readers' success with the situation model revision was assessed using the combined data on those tasks, 'l'he results showed that reading skills needed for successfu1 revision can be explained by the two textual factors specified by causal network analysis: the explicitness of the revision clues ( ± cxplicitness) and the complexity of the causal structures in each text ( ± simplicity). Specifically, higher level reading processing (e.g., making inferences) was needed for readers to revise their situation models when they read a story which did not include explicit revision clues (i,e., -explicitness), whereas lower level processing (e.g., paraphrasing) was needed when reading a story with a complicated causal structure (i.e., - simplicity). It is suggested that teachers should carefu11y take into account students' reading sub-ski11s and the causal structure ofa text when requiring them to flexibly interpret the text,
The present study irrvesdgates he efii cts of morphological and contextual clues on Japanese EFL ... more The present study irrvesdgates he efii cts of morphological and contextual clues on Japanese EFL reader's lexical inferences. A total of 148 undergraduates participated in the experiment. GToup A (}i = 70) cornpleted the word-form based lexical inference test which examines the efiix ts of morphological clues, while Group B (n = 78) completed the context based lexical inference test which examines the eifems of contextual clues on deterrnining the word class and inferring the meaning. After thcy completed each testl both groups took part in a common lexical inference test which aims at examining the effects of two types of inference processes: morphology based, and conte)a based precesses. The results demonstrated that: (a) contextual clues had strvnger effi)cts on the success of the judgrnent of word classes; fo) the knowledge ofderivational sufiixes helped the word classjudgrnent; (c) contextual clues had advantages over morphological clues on inference of the meaning; and (d)...
A recall test is usefu1 in that the test score is not affected by question items; however, this t... more A recall test is usefu1 in that the test score is not affected by question items; however, this test is frequently criticized as being simply a memory test that does not measure pure reading comprehension. Although retrieval cues are sometimes provided to readers in order to decrease the elifects of the memory load, it is unclear how these cues are related to the ideas recalled. The aim of this study is to examine what types of retrievai cues can promote recall perfbrmance of Japanese EFL learners. In Experiment 1, we investigated the relationship between learners' L2 reading proficiency and the importance level of recalled ideas. In Experiment 2, we examined the effects of cue types on recall tests. In addition, the results of three types of recall testsfree recall, recaLl with main cues, and recall with detailed cueswere compared in order to elucidate the validity of cued recall tests. The results showed that retrieval cues did not promote recall perfbrmance. Moreover, we fbund that cued recall tests had lower correlations with L2 reading proficiency sceres than free recall tests, which indicates that retrieval cues reduced the concurrent validity.
This study investigated how a flashback in a passage affects Japanese EFL readers' narrative comp... more This study investigated how a flashback in a passage affects Japanese EFL readers' narrative comprehension. The order ofevents written in a narrative story is sometimes different from that in the real world; thus, readers need additional processes which fi11 in the gap between the linguistic description of the event order and the event order in their mind, In the present study, 37 Japanese university students read a narrative text written in chronological order, and another 37 students read the same text written in non-chronological order, In the ]atter version, the first event (E) was put irrto the third position (E2->E3-)El->E4), That is, El was presented as a flashback. im imrnediate and a delayed recal1 test revealed that the participants who read the chronological text successfu11y recalied more infbrmation than those who read the non-chronological text, which was also supported by the participants' evaluation of text difficulty. In addition, the error analysis showed that both Iecal and g]obal errors increased when participants read the non-chronological text, Furthermore, more proficient readers werc able to arrange the text infbrmation m chronological ordcr in their mental representations even when thcy read the fiashback text. Although the flashback did increase the cognitive demand on readers, these readers could fo11ow the events in chronological order, and successfully corrected their situation model.
This study investigated how a flashback in a passage affects Japanese EFL readers' narrative ... more This study investigated how a flashback in a passage affects Japanese EFL readers' narrative comprehension. The order ofevents written in a narrative story is sometimes different from that in the real world; thus, readers need additional processes which fi11 in the gap between the linguistic description f the event order and the event order in their mind, In the present study, 37 Japanese university students read a narrative text written in chronological order, and another 37 students read the same text written in non-chronological order, In the ]atter version, the first event (E) was put irrto the third position (E2->E3-)El->E4), That is, El was presented as a flashback. im imrnediate and a delayed recal1 test revealed that the participants who read the chronological text successfu11y recalied more infbrmation than those who read the non-chronological text, which was also supported by the participants' evaluation of text difficulty. In addition, the error analysis sho...
Words having multiple meanings cause dithculties in text comprehension because readers possibly i... more Words having multiple meanings cause dithculties in text comprehension because readers possibly interpret such words as the primary meaning even when the words are used as other meanings. The current study examined how Japanese EFL learners interpreted those words that were used as the secondary meanings in contexts when the panicipants knew the primary but not the secondary meanings. The results showed that there was no significant difllerence in the correct rate of interpreting homonyms between r.eading projiciency levels. However, the typcs of error they produced were different: the upper group wrote more answers that fitted the context than the lower group who stuck to the primary meanings even when they understood the target contexts. These results suggest that the bcttcr readers had grcater flexibility in changing their interpretation to one that fitted the contexts, but the poor readers persisted in assuming the meanings they already knew and could not change thcir interpretation.
The present study irrvesdgates the efii cts of morphological and contextual clues on Japanese EFL... more The present study irrvesdgates the efii cts of morphological and contextual clues on Japanese EFL reader's lexical inferences. A total of 148 undergraduates participated in the experiment. GToup A (}i = 70) cornpleted the word-form based lexical inference test which examines the efiix ts of morphological clues, while Group B (n = 78) completed the context based lexical inference test which examines the eifems of contextual clues on deterrnining the word class and inferring the meaning. After thcy completed each testl both groups took part in a common lexical inference test which aims at examining the effects of two types of inference processes: morphology based, and conte)a based precesses. The results demonstrated that: (a) contextual clues had strvnger effi)cts on the success of the judgrnent of word classes; fo) the knowledge ofderivational sufiixes helped the word classjudgrnent; (c) contextual clues had advantages over morphological clues on inference of the meaning; and (d) difirerences oftwD lexical int])rence pTocesses did not have much efibct on the success ofinference. Based on these results, some implications for fhrther research and instnJctions of lexical infl)rence strategy are provided. 1 , lntroduction Learners' vocal)ulary knowiedge is important to be a proficient reader. There are several strategies for readers to cope with unknown words, such as consulhng a dictionary, ignoTing, and infetTing the meaning (e.g., Fraser, 1999; Natfoq 2001). Infemhg word meanings is an efilective strategy for development of reading ski11s because it niinirnizes the risk of impeding the readmg process (Camine, Karneenui & Coyle, 1984). Moreoveg inferc nce is recommended in the Course of Study for high sehools (Ministry ofMucatio4 Culture, Spons, Science, and 'Ilechnology} 1ep9) as a strategy to foster readers' abMties and the poshive attitudes toward reading. in this recommendaion for instmctiog lexical inference is defined as an activity in wiiich we utilize context infomiation and knowledge of sentence stmcture, werd formation and derivatives. Although the use ofmany clues is,recommended in the definitioq it is not clear how each of the clues works in the infererice process and what type ofinfererice processes wcnk better than others. As Nagy and Scott (1997, p. 275) state "comext and morpbology <word parts) are the two major sources Japan Society ofEnglish Language Education
The relationship between a learner's first language (Ll) and second language (L2) has been di... more The relationship between a learner's first language (Ll) and second language (L2) has been discussed by many rcsearchers. Among them, Kroll and Steward (1994) preposed the Revised Hierarchical Model. According to this model, lexical entries in one's Ll lexicon are connected directly "ith relevant concepts whereas these in the L2 lexicon eed to access Ll equivalents in order to activate concepts. Some researchers proposed the theory that links between the L2 lexicon and concepts are developed with the gro"・tii of learner's proficiency, etc, but results of the previous studies on this developmerrtal hypothesis are not conclusive. The present study, focusing on "rord firequency and concreteness, investigated how Ll and L2 lexicons and concepts are interrelated, A tota1 of40 first year university students participated in the experiment. lhe r¢ sults of reaction time and recall rate ofthe target xi,Tords showed that words with concrete concepts constructed a di...
A recall test is usefu1 in that the test score is not affected by question items; however, this t... more A recall test is usefu1 in that the test score is not affected by question items; however, this test is frequently criticized as being simply a memory test that does not measure pure reading comprehension. Although retrieval cues are sometimes provided to readers in order to decrease the elifects of the memory load, it is unclear how these cues are related to the ideas recalled. The aim of this study is to examine what types of retrievai cues can promote recall perfbrmance of Japanese EFL learners. In Experiment 1, we investigated the relationship between learners' L2 reading proficiency and the importance level of recalled ideas. In Experiment 2, we examined the effects of cue types on recall tests. In addition, the results of three types of recall tests free recall, recaLl with main cues, and recall with detailed cues were compared in order to elucidate the validity of cued recall tests. The results showed that retrieval cues did not promote recall perfbrmance. Moreover, we fb...
There has been much research cenducted to compare open-ended and multipie-choice tests from the v... more There has been much research cenducted to compare open-ended and multipie-choice tests from the viewpoints of construct and difficulty. However, almost no studies have examined the eifects of question types in relation to test forrnats. Through two experiments, this study investigated how question types influence the dienculty of these two test fbrmats. 1'he results of Experiment 1 showed that question types affected item dithculty in open-ended tests; more specifica]ly, thematic questions were the most difficult, fo11owed by inference questions, and paraphrase questions were the easiest. In contrast, the result of Experiment 2, in which the same tests were conducted in the multiple-choice test fbrmat, revealed that item difficulty did not ditfer significant]y by question type, In addition, we fbund that predictability of the results of the multiple-choice test vv'as low compared to the open-ended test, Comparison ot' these two experiments suggested the importance of choice in the multiple-choice test. Close examination of choices indicated that overlap of words in correct choices and the text influenced the item difficulty,
JLTA Journal Kiyo, 2009
Graduate Sthool, Uhii ersity of7lrukuba Abstcac't, 'lhe relationship between reading comprehensio... more Graduate Sthool, Uhii ersity of7lrukuba Abstcac't, 'lhe relationship between reading comprehension and vocabulary knowledge has been exarniiied in many studies, However, few studies se far have distinguished the vocal)utary knewledge in isolation and in c.ontext. Addressing this point, this study considered two types of vocal)ulary tests: Mochizuki (1998) Vbcal)ulary Size [[bst (VST) and TOEFL vocabulary items, and tried to clarify the relationship further. Adding to this, we classified question types in the reading test and exarnined eagh relationship with two types ofvocabulary knoyvFledge in order to differeniiate reading items by the level of reading comprehension, The results of Experiment I suggested that regardless of the presence or absence of context, there was no significant diffbrence between the correlation coethcients of the two vocabulary tests and TOEFL Reading scores, In addition, no correlation coeracients were significantly different in any question type, remaining the issue about whether the vocabulary tests in context and without context really differ or not. For the sake of investigating this point further, the word inference test was newly introduoed in Experiment II to exarnine what
大学英語教育学会 …, 2006
... 卯城 祐司 USHIRO Yuji; University of Tsukuba. 清水 真紀 SHIMIZU Maki; Graduate School, University of ... more ... 卯城 祐司 USHIRO Yuji; University of Tsukuba. 清水 真紀 SHIMIZU Maki; Graduate School, University of Tsukuba. 土方 裕子 HIJIKATA Yuko; Graduate School, University ... 佐藤 剛 SATO Tsuyoshi; Owani Junior High School. 印南 洋 IN&#x27;NAMI Yo; Graduate School, University of Tsukuba ...
Effbctive methods for vocabulary instruction based on the rnodel of vocabulary growth haiie not y... more Effbctive methods for vocabulary instruction based on the rnodel of vocabulary growth haiie not yet been proposed, The present study investigated how three types of lexical networks (paradigmatic, syntagmatic, and phonological) are developed. Experiment 1 used learners' reaction time to measure link strengths between two words, The results shosNJed that phonological networks develop first, followed by syntagmatic and paradigmatic networks, Based on the findings of Experiment 1, Experiment 2 hypothesized that the etlbcts of three types of cue (paradigmaticall>r., syntagmatically, and phono]ogically related words) on retrieval of newly 1earned wQrds difler according to encoding conditions, The results shovvied that: (a) all types of cuc were eflective in retrievai of known words; and (b) phonologically related words were the most effective cues in retrieving unlanovil'n werds. However, close examination of unexpected responses revealed that a disadvantage ofproviding phonologically related words in vocabulary instruction was that it could impede the use ofsemantic information. 1. Introduction Finding effective methods fbr vocabulary learning has been one of the most important issues in second language learning. }vlan}, studies have been conducted on various methods such as incidental vocabulary learning (e.g., de Bot, Paribakht, & Wesche, 1997) and eff'ects ofword retrieval on retention (e.g,, Barcroft, 2007), However, we still lack exidence that shows what the most effk ctive way to introduce a ne"J word is, Understanding the process ofvocabu]ar}, gro-rth would help us to flnd such evidence. Basically, "'ords do not stand alonc but make connections with other words based on their meanings, the phonological and erthographic fbrms, collocation, and colligation (Singleton, 1999). Lyons (1995. p, 124) called the network between words with the same meaning and grammatical category "paradigmatic:' (e.g., choose-select) and called the network between v,'ords that have different grammatical categories but can be put together in gratnmatically well-fonned combinations "syntagmatic" (e,g., take-place). Many studies haxie assessed how these lexical networks are organized and developed {e.g.,
Uploads
Papers by Chikako Nakagawa