Available online at www.sciencedirect.com
ScienceDirect Procedia - Social and Behavioral Sciences 232 (2016) 759 – 768
International Conference on Teaching and Learning English as an Additional Language, GlobELT 2016, 14-17 April 2016, Antalya, Turkey
Effects of Working Memory and Gloss Type on L2 Text Comprehension and Incidental Vocabulary Learning in ComputerBased Reading Burcu Varola,*, Gülcan Erçetinb a
YÕldÕz Technical University, Department of Foreign Language Education, Istanbul 34420, Turkey b Bo÷aziçi University, Department of Foreign Language Education, Istanbul 34342, Turkey
Abstract Current computer technologies can provide resources for the reader to employ both lexical and topic-level knowledge through glosses. As the reader is engaged in processing the text, s/he can easily access definitions of words or get more information about the topic through hyperlinks. However, there is little consensus regarding the effects of these resources especially on reading comprehension. Given the role of working memory (WM) in reading comprehension, the effects of these text resources may be mediated by WM capacity. As such, the current study, through a between groups experimental design, aims to compare the performance of low-knowledge readers on vocabulary and recall measures when they are exposed to glosses providing lexical information versus topic-level information. It also aims to examine whether WM mediates the effects of gloss content. Upon determining the participants’ WM capacity through a backward digit span task, they were randomly assigned to one of the experimental groups, which were exposed to an expository text with either vocabulary glosses or topic-level glosses or to the control group which received no glosses. After reading the text, the participants were given recall and vocabulary measures. Findings yield a significant relationship between WM and reading comprehension regardless of treatment condition and immediate positive effects of glosses on incidental vocabulary learning. © 2016 The TheAuthors. Authors.Published Published Elsevier © 2016 byby Elsevier Ltd.Ltd. This is an open access article under the CC BY-NC-ND license Peer-review under responsibility of the organizing committee of GlobELT 2016. (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of GlobELT 2016
* Corresponding author. Tel.: +90-212-383-5563 E-mail address:
[email protected]
1877-0428 © 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of GlobELT 2016 doi:10.1016/j.sbspro.2016.10.103
760
Burcu Varol and Gülcan Erçetin / Procedia - Social and Behavioral Sciences 232 (2016) 759 – 768 Keywords: working memory; definitional glosses; topic-level glosses; L2 reading; incidental vocabulary learning; computer-based reading
1. Introduction Successful text comprehension in a second language (L2) depends not only on fluent lexical decoding and syntactic parsing but also on activating relevant schemata and constructing schemata during reading. According to the construction-integration (CI) model of Kintsch and van Dijk (1978), text comprehension involves several levels of mental-model building which are surface representations, text-base, and situation model. The textbase consists of textdriven propositions that are primarily based on the meaning and rhetorical structure of the text, with minimal links to long-term memory (LTM). The reader mainly uses syntactic and semantic knowledge to arrive at textual meaning. The situation model, on the other hand, involves integration of propositional text-base and the relevant background knowledge about a specific topic (van Dijk & Kintsch, 1983). Thus, both vocabulary knowledge and background knowledge are significant predictors of reading comprehension in the L2 (Carrell & Wise, 1998; Coady, 1979; Grabe, 2004; Koda, 2005; Leeser, 2007). This means that for more advanced L2 reading, above 10,000 words at the recognition level are required (Grabe, 2014). A considerable amount of words is gained incidentally as a result of reading or listening tasks when the focus is not on word learning (Huckin & Coady, 1999). Lack of word knowledge affects reading fluency and looking up the dictionary interrupts the flow of reading. Those readers who do not know the meanings of many words in a text are inefficient readers (Anderson, 1999). Inability to recognize words may prevent and harm reading comprehension (Nation, 2001). Laufer (1996) reports a significant relationship (r=.71) between reading comprehension and vocabulary knowledge by comparing the results of vocabulary and comprehension test scores of the readers. Background knowledge, also dubbed as topic familiarity or prior knowledge, is defined as the reader’s existing knowledge about the topic (Bernhardt, 1991). Prior knowledge gains importance to maintain encoding and storage in LTM and to be able to make inferences to generate a successfully integrated representation. Learners’ background knowledge, namely domain knowledge, is an important factor in comprehension as the interaction of text content and domain knowledge can turn L2 comprehension into an L1-like efficient cognitive process (Erçetin & Alptekin, 2012). Studies have revealed that the amount of recall and comprehension is increased by activating or providing necessary content related knowledge compared to comprehension when no background knowledge is activated nor provided before reading even with two parallel texts (Carrell & Wise, 1998). Working memory (WM) is another significant predictor of text comprehension. WM's role in language comprehension has been well-established both in the L1 (see Daneman & Merikle, 1996 for a review) and L2 (see Linck et al., 2014 for a review). WM refers to the set of cognitive resources to encode, activate, store, and manipulate information while dealing with other mental activities (Baddeley, 2003). WM explains how readers process new information to place it in long-term memory (LTM) and how they search and retrieve the information when they need to remember it. WM classified into short-term WM (ST-WM) and long-term WM (LT-WM) by Ericsson and Kintsch (1995) plays a crucial role in the construction of the textbase and the situation model respectively: mental representations of successive sentences are generated in ST-WM. Elements of that representation are linked both to parts of the previously constructed text representation (the episodic text memory), which is already stored in LTM, and to the reader's knowledge. This linkage creates a LT-WM structure that provides direct access to relevant parts of these structures from the cues available in STM (p. 232). As such, text comprehension requires fluent lexical decoding, appropriate schemas and available WM resources. Thanks to current information technologies, easy access to text aids may facilitate both fluent lexical decoding and construction of appropriate schemas, which may in turn yield more WM resources available for higher-level text comprehension. One of these resources is glosses or annotations † that provide definitions of words or additional information about the topic through hyperlinks during web-based or electronic reading. Nation (2001) underscores
†
Glosses and annotations are used interchangeably.
Burcu Varol and Gülcan Erçetin / Procedia - Social and Behavioral Sciences 232 (2016) 759 – 768
that hypertext/multimedia‡ annotations can be beneficial for L2 learners in many ways. They provide immediate access to the meanings of unknown words, or help learners comprehend overall meaning of the text. Much of what we know about the effects of glosses in electronic text comprehension comes from the studies on multimedia glosses (see Abraham, 2008 for a meta-analysis). These studies generally investigated the effects of lexical glosses with visual versus textual content (e.g., Chun & Plass, 1996; Yoshii & Flaitz, 2002). While robust effects of lexical glosses on vocabulary learning have been observed, the question of whether textual-plus-visual glosses are better than textual-only or visual-only glosses is unresolved (Yun, 2011). Additionally, their effects on reading comprehension have also been inconsistent (e.g., AbuSeileek, 2011; Chen & Yen, 2013). Studies also explored the effects of the medium of language in the lexical glosses. Cheng and Good (2009) provided L1 definitions accompanied by L2 example sentences and L1-only definitions. The results revealed that overall L1 glosses are helpful in learning and reviewing words. Fang (2009) found that L1 glosses yielded increased learning in the short term, but L2 glosses led to better results in the long term. On the other hand, Yoshii (2006) found no significant effect for the medium of language, but the presentation mode of the annotations positively affected vocabulary learning. While both types of glosses enhanced vocabulary learning in Xu (2010), the benefits of L1 versus L2 glosses changed according to the proficiency level of learners in Vela (2015). As for glosses that provide extra information about the topic, few studies have investigated their effects. Ariew and Erçetin (2004) observed that access to visual or video-based topic-level annotations was negatively related to reading comprehension for intermediate ESL learners while no relationship was observed for advanced learners of English. Similarly, ùakar and Erçetin (2005) found that frequency of access to audio or video annotations providing extra information about the text was negatively correlated with reading comprehension for intermediate level learners of English. On the other hand, Akyel and Erçetin (2009), in a qualitative study investigating reading strategies of advanced EFL learners, have demonstrated the role of already existing background knowledge in reading a hypermedia text. Specifically, the low-knowledge readers consulted topic-level annotations more frequently than the highknowledge readers, suggesting that high-knowledge readers relied on their already existing prior knowledge, and they were cognitively and metacognitively more involved with the reading process. However, comparison of the groups in terms of text recall revealed that the mean difference was not statistically significant, suggesting that low-knowledge readers may have benefited from the use of annotations that provided background knowledge about the topic. WM's role in relation to access to glosses has been investigated only by Chun and Payne (2004). They observed a robust relationship between phonological WM (PWM), as measured by a recognition-based nonword repetition task, and frequency of access to translations of words while reading an L2 text. Specifically, participants with low-PWM capacity, on average, looked up three times more words than participants with high-PWM capacity. However, the vocabulary scores of the high- and low-PWM groups did not differ significantly. Measurement of WM through a recognition-based reading span task did not reveal a significant relationship between WM capacity and vocabulary, reading comprehension, and text recall measures either. The fact that Chun and Payne did not find a significant relationship between text comprehension and WM contradicts the large body of evidence suggesting a robust relationship between WM and text comprehension. The use of recognition-based WM tasks might have resulted in a measurement artifact, which makes it difficult to reliably interpret their findings. The discussion thus far suggests that most of the studies examining the effects of glosses in computer-based reading have focused on the mode (verbal or visual) and the language (L1 vs. L2) of input. Few studies have focused on the type of input (lexical vs. topic-level). Additionally, the role of WM as a mediator of the effects of glosses has been under investigated. 2. The Present Study Based on the above considerations, the research underpinning the present study aimed to investigate the combined effects of WM and the use of glosses, especially in relation to gloss content (lexical vs. topic-level). Reading in the L2 may involve high cognitive load depending on the linguistic difficulty of the text for the learners' proficiency level
‡
Hypertext and multimedia are used interchangeably.
761
762
Burcu Varol and Gülcan Erçetin / Procedia - Social and Behavioral Sciences 232 (2016) 759 – 768
and the availability of the appropriate schemas. Difficulty with lexical decoding or syntactic parsing would overload WM resources, especially for low-proficiency learners which would lead to problems with forming the textbase. As such, easy access to lexical glosses is expected to facilitate the reading process (Alptekin & Erçetin, 2011). Similarly, lack of appropriate schemas will also lead to difficulty with forming the situation model which requires inferences generated through the integration of the learners' background knowledge and the textbase. Thus, it was hypothesized that access to glosses that provide explanation about unfamiliar concepts during reading would also support the reading process. 3. Methodology 3.1. Participants The data came from ninety students studying at English Language Teaching Department of a Turkish state university participated in this study. The participants' proficiency level was identified to be advanced level based on Oxford Quick Placement Test. They were randomly assigned to one of the three groups: control, lexical annotation, and topic-level annotation groups. 3.2. The treatment A multimedia reading environment was created based on an expository text about ancient civilizations. The text was comprised of 2622 words including the annotated parts. It was divided into seven hypertext pages, so the readers did not have to scroll down to move through the text. There was a navigation map on top of each page showing the subtitles of each of those seven pages. In the control condition, there were no annotations in the hypertext. The lexical annotation condition provided access to L2 definitions of twenty-eight target words in the text. In the topic-level annotation condition, there were explanations or information about twelve possibly unfamiliar themes, concepts, or issues. 3.3. Instruments 3.3.1. Digits backward task From Wechsler Adult Intelligence Scale (WAIS), Digits Backward measure was used to determine the participants’ WM capacity. Lange (2011) indicates that WAIS is among the most commonly used scales to measure “general intellectual ability in adults” (pp. 2731). The task was implemented to all participants individually by the researcher. 3.3.2. Reading comprehension test The test consisted of fourteen multiple choice questions. Four of them targeted literal understanding while ten of them tapped inferential understanding. The test was self-paced, and while answering the questions, the readers could access the text. 3.3.3. Vocabulary tests The participants were given two unannounced vocabulary tests: (a) a recognition task which involved matching L2 definitions with the correct target words; (b) a production task which required the participants to provide definitions of the target words either in L1 or in L2. The tests were given immediately after the comprehension test and three weeks later.
763
Burcu Varol and Gülcan Erçetin / Procedia - Social and Behavioral Sciences 232 (2016) 759 – 768
4. Data analysis and results 4.1. Reading comprehension Table 1 presents the descriptive statistics for comprehension scores across the treatment conditions and WM groups. A 3 x 2 between groups ANOVA with treatment condition (control, lexical gloss, topic-level gloss) and WM capacity (low, high) as between groups factors revealed a significant main effect of WM, F(1, 84) = 4. 363, p < .05, indicating that high-WM participants have significantly higher comprehension scores (M = 9.46, SD = 2.24) compared to low-WM participants (M = 8.41, SD = 2.12), Cohen’s d = .48. The main effect of treatment condition, F(2, 84) = . 748, p > .05 and its interaction with WM, F(2, 84) = . 833, p > .05 were not significant. Table 1. Descriptive statistics for comprehension scores Condition
WMC
Mean
SD
N
Control
Low
8.82
2.128
17
High
9.46
1.613
13
Total
9.10
1.918
30
Low
8.00
1.581
17
High
9.00
2.309
13
Total
8.43
1.960
30
Low
8.40
2.875
10
High
9.75
2.573
20
Total
9.30
2.706
30
Lexical
Topic-level
4.2. Incidental vocabulary learning Table 2 provides the descriptive statistics for the vocabulary recognition and production scores. A 3 x 2 x 2 mixed ANOVA with treatment condition (control, lexical gloss, topic-level gloss) and WM capacity (low, high) as between groups factors and time (post-test, delayed test) as the repeated measures factor was conducted separately for recognition and production. Table 2. Descriptive statistics for vocabulary learning Post-test
Recognition
Condition
Mean
SD
M
SD
N
Low
Control
2.415
.6987
2.496
.7097
17
Lexical
2.988
.4338
2.719
.3805
17
Topic-level
2.734
.1558
2.772
.5915
10
Control
2.464
.6485
2.664
.6881
13
Lexical
2.935
.4100
2.645
.4079
13
Topic-level
2.977
.3777
2.850
.5857
20
Control
2.829
.8982
2.813
.8057
17
Lexical
3.833
.7955
3.029
.4991
17
Topic-level
3.035
.3068
2.602
.6015
10
Control
3.138
.8618
3.127
.5666
13
Lexical
3.676
.7803
3.209
.7132
13
Topic-level
3.425
.4364
3.026
.7244
20
High
Production
Delayed Test
WM
Low
High
764
Burcu Varol and Gülcan Erçetin / Procedia - Social and Behavioral Sciences 232 (2016) 759 – 768
The ANOVA results for vocabulary recognition revealed a significant main effect for treatment condition, F(2, 84) = 4.782, p < .05, partial eta2 = .102. The main effects of time and WM were not significant. As for the interactions, the only significant interaction was that between time and condition, F(1, 84) = 3.821, p< .05, partial eta2 = .083. This interaction (Figure 1) was probed through independent samples t-tests, which revealed that the gloss conditions had significantly higher post-test means (M = 2.97 for lexical gloss, M = 2.89 for the topic-level gloss) than the control condition (M = 2.44). However, the delayed test scores were not significantly different among the groups (M = 2.57 for control, M = 2.69 for lexical gloss, M = 2.82 for topic-level gloss).
Figure 1. Interaction between time and treatment conditions for vocabulary recognition.
Regarding vocabulary production, the ANOVA results revealed a significant main effect of condition, F(2, 84) = 4.859, p < .01, partial eta2 = .104, and time F(1, 84) = 27.229, p < .001, partial eta2 = .245 as well as a significant interaction between time and condition, F(2, 84) = 7.41, p < .01, partial eta2 = .15 (Figure 2). In order to probe the interaction, independent samples t-tests were conducted. The results indicated that the lexical gloss condition had significantly higher post-test mean (M = 76) than both the topic-level gloss condition (M = 3.29) and the control condition (M = 2.96), with no significant difference between the latter conditions. On the other hand, the delayed test scores were not significantly different among the conditions (M = 2.95 for control, M = 3.11 for lexical gloss, M = 2.88 for topic-level gloss).
Burcu Varol and Gülcan Erçetin / Procedia - Social and Behavioral Sciences 232 (2016) 759 – 768
Figure 2. Interaction between time and treatment conditions for vocabulary production.
To sum up, the results suggest a significant relationship of WM with reading comprehension but not with vocabulary learning. Additionally, the effect of access to glosses was not observed in terms of reading comprehension. However, its effect was significant on vocabulary learning. It was observed that access to both lexical and topic-level glosses had immediate facilitative effects on word recognition, with no facilitative delayed effects. As for production level, access to lexical glosses had facilitative immediate effects, which disappeared on the delayed test. 5. Discussion Within the present study, reading comprehension is not affected by annotation use, whereas vocabulary learning is. Hence, our hypotheses presuming that easy access to either type of glosses will lead to increased text comprehension are not confirmed. This finding supports Cheng and Good’s (2009) study, which reported that access to glosses, regardless of the type, enhanced vocabulary acquisition while the same effect was not observed for reading. Likewise, the content of annotations did not affect the participants’ reading comprehension in the present study. In other words, providing either vocabulary glosses or background information glosses did not change the results in terms of passage comprehension. These results are in line with Ariew and Erçetin’s (2004) findings, which did not report any relationship between annotation use and reading comprehension for high proficiency level learners. The significant relationship between WM and reading comprehension partially confirms Alptekin and Erçetin’s (2011) findings which found a significant contribution of WM to inferential comprehension, which requires more
765
766
Burcu Varol and Gülcan Erçetin / Procedia - Social and Behavioral Sciences 232 (2016) 759 – 768
controlled and demanding processing even for proficient readers compared to more automatic and less demanding processing for literal comprehension (Alptekin & Erçetin, 2011). The dominance of textually and scriptally implicit questions in the comprehension test of the current study must have triggered inferential reading which impose a high intrinsic cognitive load (Sweller, 1994). It follows, then, that low WMC advanced readers had difficulty with forming a coherent situation model compared to high WMC advanced readers when faced with inferential reading tasks. On the whole, this finding lends support to Harrington and Sawyer (1992) and Leeser (2007) studies as they also observed a meaningful interaction between WMC and reading comprehension. In terms of incidental vocabulary acquisition, both lexical and topic-level glosses led to better learning in recognition task while WMC was not influential in lexical acquisition. Contrary to Mendonça’s (2003) findings which revealed an interplay between memory span and L2 vocabulary retention, no relationship was detected in the current study. Additionally, the content of the gloss did not have a discriminatory effect on recognizing newly learned words. That is, no matter what the content of glosses are, vocabulary learning is enhanced. This may have stemmed from the nature of the matching activity. When there is a list to choose from, it is easier for learners to supply the correct word no matter which type of annotation they get during reading. In the production task where participants had to supply the words’ definitions in either L1 or L2, being exposed to lexical annotations seems to be more effective. The facilitative effects of lexical and topic-level glosses, however, were not repeated in the delayed posttests. This may suggest that provision of either word-level or topic-level glosses does not have a lasting effect in incidental vocabulary learning. On the whole, annotating a lengthy reading text is found to play an influential role in vocabulary learning. The overall effect of glosses on vocabulary learning in the present advocates a myriad of studies so far (Abuseileek, 2008, 2011; Chen & Yen, 2013; Chun & Plass, 1996). When readers encountered with unknown words have access to their meanings immediately through glosses rather than being left alone on their own, they acquire a great amount of words. As Poel and Swanepoel (2003) put forward vocabulary learning through contextual guessing can be a slow and erroneous process as readers can reach wrong inferences about the meaning. Hence, as this study has also shown, assisting readers with glosses during the reading process can prove to be an efficient way of vocabulary learning. 6. Pedagogical implications It is apparent from the above findings that learners can build on their vocabulary knowledge with the help of glosses. It is advisable, then, that material developers annotate reading materials even at higher proficiency levels. Furthermore, in order to increase vocabulary learning at production level, word definitions should be given rather than topic-level information. On the other hand, glossing does not affect reading comprehension whereas WM does. This suggests that the content and complexity of the passages should be chosen with care in the case of low capacity learners. Additionally, providing glosses did not affect reading comprehension negatively even if a significant positive effect was missing. To this end, annotating such lengthy passages with lexical and topic-level information can be an alternative material design solution considering its beneficial effects on vocabulary gains. 7. Limitations and suggestions for further research The results of the present study should be interpreted cautiously since there are some limitations. First, this study did not integrate visual or audio annotations into the design due to its experimental focus, thus it would add to our understanding to see the contributions of different annotation formats to such a design. Furthermore, short-term effects of lexical and topic-level glosses in this study can be extended to long-term with the inclusion of supplementary multimedia, such as pictures or videos. Next, the participants of this study included only advanced level students majoring in English Language Teaching. Hence the results of this study cannot be generalizable to other proficiency levels. It would be interesting to see the effects of gloss type in such lengthy passages on lower-level readers’ reading and vocabulary performance.
Burcu Varol and Gülcan Erçetin / Procedia - Social and Behavioral Sciences 232 (2016) 759 – 768
767
Acknowledgments The research underlying the study was supported by a Bogaziçi University Research Fund grant (Project No. 10901).
References Abraham, L.B. (2008). Computer-mediated glosses in second language reading comprehension and vocabulary learning: A meta-analysis. Computer Assisted Language Learning, 21, 199–226. Abuseileek, A. F. M. (2008). Hypermedia annotation presentation: Learners’ preferences and effect on EFL reading comprehension and vocabulary acquisition. CALICO Journal, 25(2), 260-275. Abuseileek, A. F. M. (2011). Hypermedia annotation presentation: The effect of location and type on the EFL learners’ achievement in reading comprehension and vocabulary acquisition. Computers & Education, 57(1), 1281-1291. Akyel, A., & Erçetin, G. (2009). Hypermedia reading strategies employed by advanced learners of English. System, 37(1), 136-152. Alptekin, C., & Erçetin, G. (2011). The effects of working memory capacity and content familiarity on literal and inferential comprehension in L2 reading. TESOL Quarterly, 45(2), 235-266. Ariew, R., & Erçetin, G. (2004). Exploring the potential of hypermedia annotations for second language reading. Computer-Assisted Language Learning, 17, 237-259. Carrell, P. L. & Wise, T. E. (1998). The relationship between prior knowledge and topic interest in second language reading. Studies in Second Language Acquisition, 20, 285-309. Chen, I. and Yen, J. (2013). Hypertext annotation: Effects of presentation formats and learner proficiency on reading comprehension and vocabulary learning in foreign languages. Computers and Education, 63, 416–423. Cheng, Y-H. & Good, R. L. (2009). L1 glosses: Effects on EFL learners’ reading comprehension and vocabulary retention. Reading in a Foreign Language, 21(2), 119-142. Chun, D. M. & Payne, J.S. (2004). What makes students click: Working memory and look-up behavior. System 32, 481–504. Chun, D. M. & Plass, J. L. (1996). Effects of Multimedia Annotations on Vocabulary Acquisition. The Modern Language Journal, 80(2), 183198. Coady, J. (1979). A psycholinguistic model of the ESL reader. In Mackay, R., Barkman, B., & Jordan, R.R. (Eds.). Reading in a second language (pp. 5-12). Rowley, Mass: Newbury House. Daneman, M. & Merikle, P. M. (1996). Working memory and language comprehension: A meta-analysis. Psychonomic Bulletin & Review, 3 (4), 422-433. Erçetin, G., & Alptekin, C. (2012). The explicit/implicit knowledge distinction and working memory: Implications for L2 reading comprehension. Applied Psycholinguistics, 34, 727-753. Ericsson, K.A., Kintsch, W. (1995). Long-term working memory. Psychological Review 102, 211–245. Fang, S. (2009). Chinese gloss or English gloss: Which is more effective for incidental vocabulary acquisition through reading? Retrieved at 8 April 2016 from http://www.diva-portal.org/smash/get/diva2:228734/FULLTEXT01.pdf Grabe, W. (2004). Research on teaching reading. Annual Review of Applied Linguistics, 24, 44-69. Grabe, W. (2014). Key issues in L2 reading development. Centre for English Language Communication, 8-18. Harrington, M., Sawyer, M. (1992). L2 working memory capacity and L2 reading skill. Studies in Second Language Acquisition 14, 25–38. Huckin, T., & Coady, J. (1999). Incidental vocabulary acquisition in a second language. Studies in Second Language Acquisition, 21(2), 181-193. Koda, K. (2005). Insights into second language reading: A cross-linguistic approach. New York: Cambridge University Press. Kintsch, W. & Van Dijk, T. A. (1978). Toward a model of text comprehension and production. Psychological Review, 85 (5), 363-394. Lange, R. T. (2011). Working memory index. Encyclopedia of Clinical Neuropsychology, 2731-2732. Laufer, B. (1996). The lexical threshold of second language reading comprehension: What it is and how it relates to L1 reading ability. In K. Sajavaara & C. Fairweather (Eds.), Approaches to second language acquisition (pp. 55-62). Jyvaskyla, Finland: University of Jyvaskyla. Leeser, M. J. (2007). Learner-based factors in L2 reading comprehension and processing grammatical form: Topic familiarity and working memory. Language Learning, 57, 229–270. Linck, J. A., Osthus, P., Koeth, J. T., & Bunting, M. F. (2014). Working memory and second language comprehension and production: A metaanalysis. Psychonomic Bulletin & Review, 21 (4), 861-883. Mendonça, Daniela M. (2003). Working memory capacity and the retention of L2 vocabulary. Unpublished Master's Thesis, Universidade Federal de Santa Catarina, Florianópolis. Nation, I. S. P. (2001). Learning vocabulary in another language. Cambridge: Cambridge University Press. Poel, K., & Swanepoel, P. (2003). Theoretical and methodological pluralism in designing effective lexical support for CALL. Computer Assisted Language Learning, 16(2–3), 173–211. Sakar, A., & Ercetin, G. (2004). Effectiveness of hypermedia annotations for foreign language reading. Journal of Computer Assisted Learning, 21, 28-38. Sweller, J., 1994. Cognitive load theory, learning difficulty, and instructional design. Learning and Instruction 4, 295–312. Van Dijk, T. A. & Kintsch, W. (1983). Strategies of discourse comprehension. New York: Academic Press.
768
Burcu Varol and Gülcan Erçetin / Procedia - Social and Behavioral Sciences 232 (2016) 759 – 768 Vela, V. (2015). Using glosses for incidental vocabulary acquisition. Procedia - Social and Behavioral Sciences, 199, 305-310. Xu, X. (2010). The effects of glosses on incidental vocabulary acquisition in reading. Journal of Language Teaching and Research, 1(2), 117120. Yoshii, M. (2006). L1 and L2 glosses: Their effect on incidental vocabulary learning. Language Learning and Technology, 10 (3), 85–101. Yoshii, M. and Flaitz, J. (2002). Second language incidental vocabulary retention: The effect of picture and annotation types. CALICO Journal, 20 (1): 33–58. Yun, J. (2011). The effects of hypertext glosses on L2 vocabulary acquisition: A meta-analysis. Computer Assisted Language Learning, 24 (1), 39-58.
