Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Watanabe thesis


Published on

Published in: Education, Technology
  • Be the first to comment

  • Be the first to like this

Watanabe thesis

  1. 1. 1University of Aizu, Graduation Thesis. March, 2013 s1170044AbstractEfficient use of cognitive and metacognitivereading strategies is important indicators of activelearning outcomes. These strategies suggest theefficacy with which content is construed andoutcomes are reached. Efficient use of thesestrategies are extremely important in an EFL contextwhere the language proficiency is low, and ability forcontent comprehension in the target language isminimal to moderate at best. This article discussed aspecific case study in this Japanese computer sciencecontext where readers were asked to comprehend thecontent in an English tourism website. This articledocuments the self-reported use of reading strategiesused by participants while making an attempt tounderstand the content of the English website.Metacognitive Awareness of Reading StrategiesInventory Questionnaire (MARSI) was used for thepurpose of understanding readers’ preferential use ofreading strategies and what it means for contentaccuracy. The self-reported use of strategies wasclassified into three groups: Global ReadingStrategies, Problem-solving Strategies and SupportReading Strategies. Results for mean values on anaverage show that most strategies in GLOB andPROB categories have been used consistently at ahigher level when compared to strategies in SUPsection. The findings are consistent with the literaturesuggesting that readers with low reading ability (e.g.,as in this EFL context as reported in the experiment)attached lower importance to support readingstrategies when compared to L1 readers [1].1 IntroductionThe literature is rich in suggesting how EFLreaders strategize towards reading, conceiving,selecting, comprehending, and recalling onlineinformation for target language learning or othersimilar purposes. Reading strategies are an indicator ofhow a task is conceived, the textual cues attendedduring reading, how a text is perceived and sensed,and how a non-comprehensible piece of text isunderstood [2]. Research on second language readingand comprehension might be a little different fromforeign language comprehension context, but theoverall use of reading strategies and its relation tosuccessful and unsuccessful second language readinghave wider implications and results in ourunderstanding of reading approaches in an EFLcontext [3][4].Research in a native context with much higher Englishreading and comprehension ability, when comparedwith ESL students displayed conscious awareness ofall metacognitive reading strategies commonly used.Further, readers in both contexts attribute similarlevels of importance to the reading strategies used.These strategies are cognitive strategies (the deliberateactions readers take when comprehension problemsdevelop), followed by metacognitive strategies(advanced planning and comprehension monitoringtechniques), and support strategies (the tools readersseek out to aid comprehension). Further, both L1 andESL readers demonstrated similar used of cognitiveand metacognitive reading strategies when comparedto EFL readers. Besides, readers with low readingability (close to being called EFL students) attachedlower importance to support reading strategies whencompared to L1 readers [1].The major purpose of this study is to briefly report onhow EFL readers comprehend website information inthe target language, reporting on the design of thewebsite in their own words, and then answering aquestionnaire wherein they identify diverseinformation that is available in the website but couldbe accessed through proper navigation.So, the major research question for this study that oneneeds to ask is as follows. What kind of cognitive, metacognitive and supportreading strategies have been used, as self-reportedby the reader in an EFL context when accessing awebpage in English for information comprehension,navigation and understanding how the website isdesigned? To what extent have these strategies been used? Is there any significant correlation between scoreson website information comprehension and use ofself-reported strategies or there is no pattern tosuggest how a strategy influenced performance onwebsite information comprehension?Metacognitive Reading Strategies for Website Analysisin an EFL ContextMasaya Watanabe s1170044 Supervised by Prof. Debopriyo Roy
  2. 2. 2University of Aizu, Graduation Thesis. March, 2013 s1170044 Is there any pattern (or significant correlation) tosuggest that a person who used a specific strategyalso used other strategies to a similar extent? Is there any significant difference in the self-reported scores within global, problem solving andsupport reading strategies?This study reports results from an experimentalanalysis, wherein different cognitive, metacognitiveand support reading strategies have been usedtowards web analysis and web informationcomprehension from an English travel website.2. A Review of the LiteratureRecent research on L2 learning strategies haslargely focused on language learners metacognitiveknowledge or awareness of strategies. The study foundthat in China, EFL readers’ metacognitive knowledgeof reading strategies had close links to their EFLproficiency [5]. However, some studies have de-emphasized the importance of metacognitive readingstrategies. The results of the investigation on readingstrategies and extensive reading of EFL studentsindicate that reading a lot in both L1 and L2/EFLbecomes basically the most important factor forimproving reading skills rather than just teachingreading strategies. Extensive reading gives learnersrich background knowledge, vocabulary recognition, ahigh motivation for more reading, and becomes thebasic skill of rapid reading, discovery of readingstrategies by learners themselves, and increasesguessing ability in context [6].Understanding cognitive and metacognitive readingstrategies in a web-based EFL reading context isimportant for judging how readers might haveapproached the task of website reading and analysis.Brown (1994) defines cognitive strategies (e.g.,bottom-up and top-down) as specific learning tasksand is used to help an individual achieve a particulargoal (for example, understanding a text) [7].Goodman, K (1998), refers to the bottom up model asthe common sense notion [8]. In this approach,reading is meant to be a process of decoding;identifying letter, words, phrases and then sentences inorder to get the meaning. On the other hand, Carrell(1998) stated that, reading process as “an activepsychological guessing game [9]. Top down rejects thenotion that identification of letters to form words, andthe derivation of meaning from these words is efficientreading. On the contrary, it is assumed that efficientreading requires the readers to make predictions andhypotheses about the text content by relating the newinformation to their prior knowledge and by using asfew language clues as possible. Both bottom-up andtop-down theories formed the models of reading, andare important for web-based language learning in anEFL context.Metacognitive reading strategies in an EFL contextclearly have shown that strategies like skimming andscanning often are practiced for extracting the requiredinformation [10], without a deeper-level analysis.Further, research projects observed learner-task-dictionary interactions and suggested the importanceof learners’ interactions while carrying out specificlanguage learning tasks [11]. A different research onthe use of metacognitive reading strategy shows thoseparticipants who received training and/or read the textwith headings remembered text topics and theirorganization better than participants who received notraining and read the text without headings [12].Analysis of time data showed that participants withheadings spent more time per word reading the pageswith headings than participants who did not haveheadings [13].3. MethodSample and Context: Participants (N=59) arejunior level students (age group: 18-20 years) in theirthird year undergraduate program specializing incomputer science in a Japanese technical university.For the purpose of closely observing the participantsover a prolonged period of time and reporting on theiractivities, we felt that this sample size representedsizes commonly used in usability tests. Students at thisuniversity are required to take 15 credits ofundergraduate English in order to graduate with acomputer science major. At the time when studentstook this course, they already had passing grades orbetter in several English courses offered during thefirst two years of their program. With this specificelective course named Writing and Design for WorldWide Web, students mostly focused on the process ofonline writing, designing and analyzing websites basedon design principles, besides designing concept mapson websites they analyzed.Japanese Context: Detailed class observations at thisuniversity over a period of time and across studentbodies have shown extensive use of metacognitivereading strategies like skimming, scanning, readingheadlines, using dictionaries, translation etc in
  3. 3. 3University of Aizu, Graduation Thesis. March, 2013 s1170044analyzing technical texts in an EFL context and resultshave shown superficial and often surface level abilityto comprehend, analyze and write about a technicaltext [14] and similar results were observed in otherEFL situations [15].Stages of the experiment:1st Week: During the 1st week of the actualexperiment, students (N = 59) analyzed the Belizetourism website based on the 8 open-ended questionsasked during the same assignment that happened overthe previous weeks (but with different websites eachweek) as reported above.Second Week:1stStage: During the 2nd week, readers (N = 59) weregiven a set of questionnaires to answer on the sameBelize tourism website. The websites focused on threedifferent aspects related to information comprehensionand online reading strategies for an EFL context. Thequestionnaires focused on three different aspects. Their ability to navigate through and look forinformation from the website. Their ability to meaningfully self-report on theusability of the website in terms of navigation,content, and organization. Their ability to meaningfully self-report on theirreading strategies when analyzing the Belizetourism website.The website information comprehension with theabove-mentioned focus was completed in class.Results on the website information comprehensionquestionnaire are beyond the scope of this study.2ndStage: During this stage, participants completedthree usability questionnaires of their impression aboutthe website, how it helps with navigation, informationsearch, learning, etc.Self-Reporting of Reading Strategies (Final Stage):This article will specifically focus on the self-reporteduse of cognitive, metacognitive and support readingstrategies, and a summary report explaining the extentto which readers were able to complete the websitedesign and comprehend the website information.3.1 InstrumentsThe purpose during the final stage was to ascertain towhat extent these EFL readers are aware of theirmetacognitive reading strategies on the web, whileanalyzing a website in the target language. This studyis exploratory to some extent because this MARSI 1.0questionnaire has never before being used forunderstanding metacognitive use of reading strategiesfor a web analysis context in an EFL setting. Studentsself-reported on the degree and extent to which variousreading strategies were used. Questions related to thedifferent strategies as mentioned in the questionnairewere in three different groups:· Global Reading Strategies (GLOB)· Problem Solving Strategies (PROB)· Support Reading Strategies (SUP)This questionnaire session went on as in-class activityand three undergraduate students closely watchedreaders as they completed the assigned activities. Aspart of an earlier pilot study (not reported in thisarticle) with the same questionnaire, three graduatestudents took part in actively translating thequestionnaires in Japanese, explained the class aboutthe activity in Japanese, and also completed anobservation checklist explaining how readersapproached the task of completing the questionnaires.Three coders were used for recording the participantchoices from the Likert scale in a range of 1 – 5, asused in the Metacognitive Awareness of ReadingStrategies Inventory (MARSI. Version 1.0)questionnaire; 1 suggesting a minimal use of the givenstrategy and 5 indicating the highest and probablyaccording to the reader the most successful use of thereading strategy.4. ResultsThe following section provides an in-depth summaryof how the different metacognitive reading strategieswere used when analyzing websites in the targetlanguage. Table 1 shows the mean and standarddeviation of the self-reported scores suggesting theextent to which each of the 13 approaches under“global reading strategies” was adopted. Resultsindicate that the mean varied between 2.5 and 3.8 onan average across all global reading strategies, withmost scores in the range of 2.5 and 3.5. The “GLOB”reading strategy mean score for 13 questions is 3.024.Table 1: Descriptive Statistics on Self-ReportedUse of Global Reading StrategiesQuestion Mean Std.Deviation1 3.83 1.1013 3.54 1.0394 2.14 1.1377 3.42 1.27610 2.76 1.27814 2.95 1.265
  4. 4. 4University of Aizu, Graduation Thesis. March, 2013 s117004417 3.31 1.22119 3.71 1.01822 2.36 1.39923 2.36 1.22825 3.39 1.09926 2.95 1.18129 2.56 1.249Table 2 provides the self-reported scores on the use of7 problem-solving strategies during website analysis inthe target language. For question # 13, we see a highself-reported mean score of 4.08 suggesting it to be apopular strategy (I adjust my reading speed accordingto what I am reading). The “PROB” reading strategymean score for 8 questions is 3.574.Table 2: Descriptive Statistics on Self-ReportedUse of Problem Solving StrategiesQuestion Mean Std.Deviation8 3.71 1.09911 3.2 1.2713 4.08 .91516 3.64 1.09518 3.54 1.05621 3.14 1.35827 3.83 1.10130 3.44 1.134Table 3 provides the self-reported score on the use of 5support reading strategies during English websiteanalysis as reported in the experiment undertaken. Inmost cases we see a self-reported score of less than oraround 2.5. The “SUP” reading strategy mean scorefor 9 questions is 2.762.Table 3: Descriptive Statistics on Self-ReportedUse of Support Reading StrategiesQuestion Mean Std.Deviation2 2.39 1.2875 2.31 1.3296 3.05 1.2929 2.69 1.36812 2.66 1.52715 3.14 1.33220 3.24 1.27824 3.34 1.12428 2.05 1.09Figure 1 shows a line diagram showing comparativemean scores on all the 3 subscales. This data shows arelatively wider use and acceptability of the PROBmetacognitive reading strategy with an overall meanvalue of more than 3.5.Figure 1: Line Diagram Showing ComparativeMean Scores on all 3 SubscalesData in Figure 2 shows a relatively wide variabilityamong the self-reported use and acceptability of theGLOB reading strategies, with question 1 (I have apurpose in mind when I read) and 19 (I use contextcues to help me better understand what I am reading)showing relatively higher values.Data in Figure 3 for problem-solving strategies show arelatively lesser variability in self-reported scores withall scores above 3.0, showing relatively moreacceptability and use of problem-solving strategies onan average.Figure 2: Line Diagram Showing ComparativeMean Scores on Global Reading StrategiesFigure 3: Line Diagram Showing ComparativeMean Scores on Problem Solving Strategies01234GLOB PLOB SUPMeanValueineachsubscaleSubscale Type02461 3 4 7 101417192223252629MeanValueQuestion in GLOB Subscale02468 11 13 16 18 21 27 30MeanValueQuestion in PROB Subscale
  5. 5. 5University of Aizu, Graduation Thesis. March, 2013 s1170044Figure 4 shows that all the mean scores on supportreading strategies are around 3.0 with one strategyshowing as little as 2.05 (I ask myself questions I likedto have answered in the text) and is the minimum forall strategies combined.Figure 4: Line Diagram Showing ComparativeMean Scores on Support Reading StrategiesPearson correlation results show a strong patternbetween the reported use of different strategies inGLOB, PROB and SUP categories. More importantly,a significant correlation is visible across GLOB,PROB and SUP categories suggesting similar use ofdifferent strategies across the spectrum.Data also shows a large number of significantcorrelations involving each approach (question). Thisdata will give us an indication of the extent to whicheach strategy usage as reported, were found to besimilar in usage to another strategy, across GLOB,PROB, and SUP approaches.Another important research question that needed to beexplored in this context is whether there is anystatistically significant correlation between websiteInformation Comprehension Scores and self-Reportedscores on Different Reading Strategies.Table 4: Significant Correlation between WebsiteInformation Comprehension Accuracy Score andMetacognitive Reading StrategiesCorrelationBetweenPearson CorrelationValueSig. (2 –tailed)Q 16 ~ Q 9 .279 * .032Q19 ~ Q5 -.271 * .038Q20 ~ Q8 .345 ** .007Data suggested that in most cases, there is notmuch of a pattern when it comes to inferring a directrelation between accuracy scores and metacognitivereading strategy, suggesting the use of readingstrategies might be generally contextual and somethingthat readers prefers to use/practice. However, somestatistically significant results were obtained. These areas shown in Table 4.The Friedman non-parametric statistical analysiswas performed to find out if there is an overallstatistically significant difference between the meanranks of the related reading strategies. This test tells uswhether there are overall differences between relatedreading strategies but does not pinpoint whichstrategies in particular differ from each other. To dothis we need to run post-hoc tests, but post-hocanalysis was not considered as part of this study. Datashows us a Friedman analysis for all 3 sub-scales(GLOB, PROB, and SUP) combined.Null Hypothesis: Mean Ranks for all the readingstrategies are equal;Alternative Hypothesis: Not all the mean ranks areequal.Results suggest that there was a statisticallysignificant difference in perceived use of readingstrategies for all the three sub-scales combined,depending on which type of strategy was used whileanalyzing the English website, χ2(2) = 382.491, P =0.000. Since p-value = 0.00 ≤ 0.01 = α, we reject thenull hypothesis.Friedman Test values for the GLOB subscalesuggest that there was a statistically significantdifference in perceived use of reading strategies withinthe GLOB subscale, depending on which type ofstrategy was used while analyzing the Englishwebsite, χ2(2) = 165.016, P = 0.000. Since p-value =0.00 ≤ 0.01 = α, we reject the null hypothesis. Datashows the Friedman Test values for exclusively thePROB subscale. Results suggest that there was astatistically significant difference in perceived use ofreading strategies within the PROB subscale,depending on which type of strategy was used whileanalyzing the English website, χ2(2) = 38.308, P =0.000. Since p-value = 0.00 ≤ 0.01 = α, we reject thenull hypothesis. Finally, data shows the Friedman Testvalues for exclusively the SUP subscale. Resultssuggest that there was a statistically significantdifference in perceived use of reading strategies withinthe SUP subscale, depending on which type of strategywas used while analyzing the English website, χ2(2) =67.819, P = 0.000. Since p-value = 0.00 ≤ 0.01 = α, wereject the null hypothesis.012342 5 6 9 12 15 20 24 28MeanValueQuestion in SUP Subscale
  6. 6. 6University of Aizu, Graduation Thesis. March, 2013 s11700445. DiscussionOverall results suggest relatively higher levels ofagreement with the statement that the strategies havebeen used in more cases than not. However, meanvalues on an average show that most strategies inGLOB and PROB categories have been usedconsistently at a higher level when compared tostrategies in SUP section. These findings are consistentwith the literature suggesting that readers with lowreading ability (e.g., as in this EFL context as reportedin the experiment) attached lower importance tosupport reading strategies when compared to L1readers [1].6. ConclusionFuture studies could be designed to map metacognitiveawareness of reading strategies with actual use ofreading strategies using screen capture software, eye-tracking data and formal usability testing methodswhich forwards direct evidence of how each readingstrategy might have been used. However, some of itwill still be an inference because mental informationprocessing might not always have direct actionevidence in terms of interaction with the software orwebsite. However, it might still be a worthwhile studyto conclude.This study helped in forwarding direct evidence ofhow far readers are aware of their use ofmetacognitive reading strategies. The fact that thisMARSI questionnaire was filled out right after theirworking with the website provided more reliabilitybecause chances are higher that students rememberedwhat they exactly did when interacting with thewebsite. Further, separate findings related to GLOB,PROB and SUP subscales helped us understand howthese reading strategies differ from each other andhow important they are for the reader in acomparative scale.7. References[1] Sheorey and Mokhtari, “Differences in themetacognitive awareness of reading strategies amongnative and non-native readers”, System, Vol. 29, Issue4, Dec. 2001, pp. 431–449[2] Ellen Block, “The Comprehension Strategies ofSecond Language Readers”, TESOL Quarterly, Vol.20, Issue 3, Sep. 1986, pp. 463–494[3] Hosenfeld, Carol, “A Preliminary Investigation ofthe Reading Strategies of Successful andNonsuccessful Second Language Learners”, System,Vol. 5, No. 2, May 1977, p110-123[4] Jimenez, Garcia, & Pearson, “Three Children, TwoLanguages, and Strategic Reading: Case Studies inBilingual/Monolingual Reading”, AmericanEducational Research Journal, Vol. 32, No. 1, Mar.1995, pp. 67-97[5] Lawrence Jun Zhang, “Effects of task complexityon the fluency and lexical complexity in EFL students’argumentative writing”, Journal of Second LanguageWriting, Vol. 19, Issue 4, Dec. 2010, pp. 218–233[6] Chiyo Hayashi, “Extensive Reading and IntrinsicMotivation to Read”. [Online][7] John Seely Brown, “Situated Cognition”,Perspectives on Situated Learning, EducationalTechnology; spring 1994.[8] Kenneth S. Goodman,Yetta M. Goodman, “To ErrIs Human: Learning about Language Processes byAnalyzing Miscues,”Reconsidering a BalancedApproach to Reading, Constance Weaver (ed.), Urbana,IL: National Council of Teachers of English, 1998, pp.101-123.[9] Patricia L. Carrell, "Metacognition and EFL/ESLreading", Instructional Science March 1998, Volume26, Issue 1-2, pp 97-112[10] Nebila Dhieb-Henia, “Evaluating theeffectiveness of metacognitive strategy training forreading research articles in an ESP context”, Englishfor Specific Purposes, Vol. 22, Issue 4, 2003, pp. 387-417[11] Marie-Josée Hamel & Catherine Caws, “UsabilityTests in CALL Development: Pilot Studies in theContext of the Dire autrement and FrancotoileProjects”, CALIC Journal, Vol 27, No. 3, May 2010,pp. 491-504[12] Sanchez and colleagues. (2001). Effects ofheadings on text processing strategies. ContemporaryEducational Psychology. v26. 418-428.[13] John R. Surber & Mark Schroeder, “Effect ofprior domain knowledge and headings on processingof informative text”, Contemporary EducationalPsychology, Vol. 32, Issue 3, Jul. 2007, pp. 485-498[14] Debopriyo Roy, “Reading strategies forprocedural information in EFL business writingenvironment: An exploratory analysis”, ProfessionalCommunication Conference (IPCC), 2010 IEEEInternational, Jul. 2010, pp. 143- 151[15] Maghsudi, M., and Talebi, S. H. “The Impact ofLingualuity on the Cognitive and MetacognitiveReading Strategies Awareness and ReadingComprehension Ability”, Journal of Social Sciences,18 (2), 2009, pp. 119-126