Summary of the Findings

CHAPTER FIVE

DISCUSSION AND IMPLICATIONS

This chapter comprises four main sections. The first section summarizes the major findings of the present study. The second section presents detailed discussion of these findings with reference to the results in previous studies and relevant theories.

The third section provides both theoretical implications for future research and pedagogical implications for vocabulary instruction in the language classroom. The final section addresses limitations of the current study, followed by suggestions for future research on vocabulary learning.

Summary of the Findings

This study investigated the effects of different input treatments (i.e. reading, listening, and a combination of reading and listening) on vocabulary learning and retention across two text types (i.e. the narrative and the expository text). To research this topic, six subgroups of different input treatments and text types were compared for their performance in the vocabulary learning and retention. A post-experiment questionnaire was designed to measure different degrees of learning motivation, anxiety, and attention induced by different treatments. Findings of this study are summarized as follows.

According to the results in this study, different input treatments are confirmed to

have significant impacts on vocabulary learning and retention. Students who read or

simultaneously read and listened to the text appeared to learn significantly more new

words than those listening to the text. After one week, however, both the reading and

the combination groups displayed substantial loss of the learned words while the

listening group revealed a high rate of retention. In spite of the impressively high

retention of the listening group, the reading and the combination groups still retained much more new words in the recognition test (but not in the recall test) than the listening group on the delayed posttest. Reading and a combination of reading and listening seem to be more effective input treatments for second language vocabulary learning.

Another variable, genre, did not result in significant main effect on vocabulary learning in the current study. Participants treated with the narrative and the expository text appeared to learn and retain similar amount of vocabulary under different input treatments in the immediate posttest and the delayed posttest. Nor did genre interact with input treatment or time in influencing vocabulary learning. Due to poor power, measures to increase power should be taken in future studies to further clarify the role of text genre in vocabulary learning.

The design of the questionnaire was originally meant to help explain differences

in learning induced by different treatments. However, no significant correlation was

found between vocabulary scores and the three factors considered in this

study—motivation, attention, and anxiety. Consequently, they were not good

candidates for interpretation of differences in vocabulary learning though significant

differences in motivation and attention across groups were observed. In contrast,

significant correlation was found between vocabulary scores and self-reported

comprehension of the text. The better learners comprehended the text, the more new

words they learned during the treatment stage. However, this factor could not account

for the differences in vocabulary learning across groups either because there was no

significant difference in text comprehension among the six treatment groups. Another

significant correlation was found between retention scores and the attention learners

in the combination groups paid to the aural input. This interesting finding seems to

correspond to the high vocabulary retention of the listening group and leaves room for

more discussion on the effect of aural input on vocabulary learning in the next section.

Discussion

In this section, the effects of different input treatments on vocabulary learning and retention generalized from the current study will be discussed and compared with the findings from previous research. Then, the effect of genre as well as the different performance of input treatments across genres will also be briefly discussed. Finally, other findings from the analysis of the questionnaire and from the two vocabulary tests will be presented for reference.

The Effects of Different Input Treatments

In this study, different treatments of reading, listening, and the combination of reading and listening all produced more or less vocabulary gains and can be regarded as feasible ways to facilitate vocabulary learning. However, the three input treatments contributed differently to vocabulary learning and retention.

Concerning vocabulary learning, the reading and the combination treatments

induced significantly more vocabulary gains than the listening treatment in both the

recall and recognition tests. No significant difference was found between the

combination treatment and the reading group. The nonsignificant difference between

the combination treatment and the reading one is consistent with the finding in Lee’s

(2006) study, in which children simultaneously read and listened to a story did not

significantly outperform the silent reading group. As for the comparison between the

reading and listening treatments, the significantly more vocabulary gains for the

reading group over the listening one echoed with the findings in Baddeley’s (1988)

study but contradicted the results in Henderson’s (2001) study. Such comparisons

between previous research and the current study will be further elaborated in the next

section accompanied with possible explanations for any inconsistency.

In terms of vocabulary retention, both the reading and the combination treatments displayed significant decay of the learned words while the listening treatment revealed an impressively high rate of retention. However, owing to the limited vocabulary gains in the immediate posttest, the listening treatment still scored significantly lower than the other two groups in the recognition test for the delayed posttest. Such results seems interesting but find little support from previous research since few empirical studies compared learning retention from different input treatments for L2 learners. Thus, several theories reviewed in the second chapter will be incorporated here to help explain the different performance in retention resulting from different input treatments.

To make clear the effect of each input treatment in contrast with the other two in the current study and also to generalize contributions of each treatment on vocabulary learning and retention, the following discussion will focus on one treatment at a time.

The Reading Treatment

According to the substantial vocabulary gains for the reading group in the immediate posttest, reading is indeed an effective source for EFL learners to learn new words, which echoes with the promotion of extensive reading programs in Elley’s (1991) and Lao and Krashen’s (2000) studies. The reading treatment in the current study not only significantly outperformed the listening treatment in vocabulary gains but almost equaled the performance of the combination group. Such outstanding performance of the reading treatment in facilitating vocabulary learning might be attributed to the availability of both visual and phonological encoding of the new words, as Baddeley (1988) suggested, when the texts were presented visually.

Although subjects in the reading treatment were not provided with sounds of the

words directly, those familiar with sound-letter correspondence could still mentally

decode the words and gained access to the auditory representations to facilitate

learning. The direct and indirect access to the visual and auditory representations of the text partly resembles the condition for the combination treatment and may thus account for the similar performance between the silent reading group and the reading-and-listening group in the current study as well as in Lee’s (2006). However, this explanation casts doubt on whether or not students with little knowledge on sound-letter correspondence could learn vocabulary well from silent reading. It was unknown in this study whether individuals’ learning performance correlated with the ability of phonological decoding when they were assigned to learn vocabulary from silent reading. Further studies can incorporate subjects of different abilities in phonological decoding to examine the effect of phonological decoding ability on learning vocabulary through silent reading.

Aside from the satisfactory vocabulary gains in the immediate posttest, the reading treatment also yielded the highest retention of the target words among the three treatments though the decay of the learned words over time also reached the significant level. This result was contradictory to the findings in Waring and Takaki’s (2003) and Cheng’s (2006) studies. In particular, in Cheng’s (2006) study, subjects receiving silent reading treatment learned slightly more target words but retained significantly less one week after the treatment than those under the listening treatment.

However, the reading groups in the current study learned and retained significantly

more target words than the listening groups. Two possible explanations were

suggested for this inconsistency. First, subjects in Cheng (2006) were tenth-graders in

senior high school while those in the current study were eleventh-graders. With one

more year of training in senior high school, participants in the current study might

have better ability to decode the words for auditory representations and might result in

better performance in vocabulary learning under the silent reading condition. Second,

the treatment for the reading group in Cheng (2006) was somewhat different from that

in the current study. Subjects for the reading group in Cheng (2006) were given a text in hand so that sentences were presented altogether. However, the text in the current study was presented sentence by sentence through PowerPoint. Based on the researcher’s in-class observation, the text presented through PowerPoint sentence by sentence appeared to involve the subjects in a highly-attentive learning condition and incidentally force them to follow every step in the treatment without distraction. The subjects in Cheng (2006), by contrast, might not follow the researcher’s instruction to review the text again and again during the treatment since the material was simple enough for them to comprehend within a single time of reading. Thus, the performance of the reading group in Cheng (2006) and the current study may differ as a result of the different treatments of silent reading. If a text has to be presented sentence by sentence to ensure a satisfactory learning retention, great caution should be exercised when students are given a text for silent reading on their own, a more common practice in a normal English class.

The outstanding performance of the reading group in the current study also contradicts the claim made by Waring and Takaki (2003) that vocabulary acquired from reading is only modest and cannot be retained effectively. The comparatively inferior performance of the reading group in Waring and Takaki’s (2003) study was not surprising since their subjects were required to learn incidentally from reading a much longer text (a graded reader) without explicit explanations for the target words.

In contrast, the current study adopted a much shorter text and provided learners with L1 translations of the target words to facilitate comprehension. These differences may contribute to the better performance of the reading group in the current study.

The above discussion has demonstrated how different designs of the experiments,

including the decoding skills of the subjects, materials used, procedures of treatments,

may affect the ultimate results of the reading treatment. Despite some inconsistency

with previous studies, reading has again been confirmed to be a dependable source for vocabulary learning for such intermediate learners as those in the current study.

Although the reading groups displayed significantly decay of the learned words one week after the treatment in the current study, the reading treatment still facilitated learners in retaining the most target words among all the treatments. Most importantly, in light of the fact that the groups given the reading treatment (be it a sole treatment of reading or a combined treatment of reading and listening) outperformed the groups merely treated with listening, reading appears to be an effective means of facilitating L2 learners’ vocabulary learning. Accordingly, language teachers are advised to promote reading among students.

The Listening Treatment

The listening treatment in this study, regardless of the text used, yielded significantly less vocabulary gains than the reading and the combination treatments.

The result contradicts the strong support for learning vocabulary from listening to stories in previous studies (Brett et al., 1996; Elley, 1989; Penno et al., 2002; Robbins

& Ehri, 1994). The result even contradicts the finding in Henderson’s (2001) study, where listening to stories read loud contributed to a much greater vocabulary gain than reading stories silently. A potential explanation for this discrepancy is that previous studies listed above mainly researched the effect of learning vocabulary from storytelling on L1 children, who are generally believed to have more spoken words than L2 learners. When treated with aural presentation of a text, L1 children may have less difficulty comprehending the words and the text than the L2 learners. In other words, nonnative learners may not process the auditory text as easily as native learners and they may need textual support to facilitate learning, as suggested by Baddeley (1988).

Despite inferior vocabulary gains, the listening treatment in the current study

produced a high retention on the learned vocabulary. This result conforms to Elley’s (1989) suggestion that learning vocabulary from listening was relatively permanent.

Henderson (2001) once ascribed the effect of story telling on vocabulary learning to the learners’ promoted interest and persistent attention during the treatment stage.

However, the current study did not detect a significant correlation between the level of attention or learning motivation and their performance in vocabulary gains and retention. Thus, other explanations are provided here for reference. First, the modest vocabulary gains induced by the listening treatment were more likely to be retained well in comparison with the substantial gains induced by the reading and the combination treatments due to the floor effect. Second, the high retention of the listening treatment may correspond to Conrad’s (1964) and Baddeley’s (1992) theories. According to Conrad (1964), linguistic material, whether presented auditorily or visually, has to be retained in the speech form to be recalled afterwards.

Similarly, Baddeley (1992) proposed a phonological loop specific for new words acquisition, in which phonological forms were stored temporarily and rehearsed for more permanent memory to be constructed. From the perspectives of Conrad (1964) and Baddeley (1992), the aural input might result in better retention due to its ready speech form or a direct access to the loop as Baddeley (2003) suggested.

To sum up, aural presentation of the text alone cannot assure a satisfactory

vocabulary gain for L2 learners, who lack enough spoken words of the target

language and need more support from textual input. However, aural input may

function well in retaining the learned words. Although the listening treatment

produced the lowest scores in both the immediate and delayed posttests in the current

study, it may still be a feasible learning source for advanced learners with more

spoken words.

The Combination Treatment of Reading and Listening

The combination treatment of reading and listening significantly outperformed the listening treatment and produced the most vocabulary gains among all the treatments in the immediate posttest although the difference between it and the reading treatment did not reach a significant level. The result seems to support the claim in the dual coding theory (Sadoski, 2005) and the multimedia learning theory (Mayer & Sims, 1994) that a combination of visual and auditory representations of the words facilitates vocabulary learning better. This result obviously differs from the redundancy effect described in Kalyuga et al.’s (2004) study, probably because L2 learners in the current study were not proficient enough to learn well from aural input alone but required support from both the written and aural input to facilitate learning and retention. As shown in Kalyuga et al. (2004) and Yeung et al. (1998), whether or not the simultaneous presentation of written and aural input is redundant depends on the expertise of the learners.

Concerning the retention of the learned words, the combination treatment of

reading and listening scored significantly higher than the listening group in the test of

vocabulary recognition for the delayed posttest. Despite its higher score over the

listening group in the delayed posttest, the combination treatment resulted in the

lowest rate of retention of the learned words among all the treatment groups. The

above results may well suggest that a group of combined input treatment does not

necessarily reflect the sum of all the strengths from both types of input. One possible

explanation for this phenomenon is that some students, when exposed to both input at

the same time, may fail to focus on both of them and yield an ideal learning outcome

(Beneli, 1997; Hsueh, 2005). This assumption also echoes with the claim in Kalyuga

et al. (2004) that split-attention to multiple input sources may result in none of the

input processed adequately. Moreover, as Jacoby et al. (1993) suggested, divided

attention may not establish a sustainable long-term memory trace or facilitate retention as well as full attention. All of these may help explain the lower retention rate of the combination treatment. Despite the comparatively lower rate of retention, it should be noted that students under the combination treatment of reading and listening still learned and retained significantly more target words than those under the listening treatment; besides, they performed as well as those receiving the reading treatment. The combination treatment could thus be seen as an effective treatment for L2 learners.

Interestingly, the analysis of the final item in the questionnaire for the combination group revealed that the level of attention the subjects paid to the aural presentation of the text was significantly correlated with their scores in the delayed posttest but not with the scores in the immediate posttest. That is, the more attention the subjects paid to the aural input, the better they retained the learned words in the delayed posttest. Such a result corresponds to the high learning retention of the listening group and again confirms the importance of aural input in facilitating retention of the learned words.

As a whole, the combination treatment of reading and listening can elicit an immediate pleasing effect on vocabulary learning, which conforms to many previous studies (Dubois & Vial, 2000; Lee, 2006; Lightbown, 1989, cited in Elley, 1991) and theories (Mayer & Sims, 1994; Sadoski, 2005). However, there were few studies that compare the retention effect of dual input sources with single input source. Thus, whether concurrent presentation of written and aural input generally produces a lower retention rate than single input alone, as found in the current study, deserves more research in the future.

The Effect of Genre

The analysis of three-way ANOVA indicated that subjects given different text

types in the present study did not differ significantly from each other in vocabulary learning and retention. However, a significant effect of genre on vocabulary learning was found in Chao’s (2003) study, which demonstrated that students learned and retained significantly more new words from the narrative text than from the expository text. The failure to detect a significant effect of genre in the current study might be the result of a relatively small sample size. Another possible explanation is that the supply of L1 translations in the current study may reduce the effect of genre on vocabulary learning. With L1 translations for each target word provided, the participants did not have to determine the word meanings from the context but only to consolidate word meanings provided by the teacher. The nonsignificant effect of genre in this study may suggest that the effect of genre is better observed during the process of meaning determination from the context rather than during the stage of meaning consolidation. In view of this, future studies may examine vocabulary learning from different text types with and without L1 translations to further clarify the effect of genre. Besides, larger sample size and longer texts that can better reflect the features of each text type should also be considered in the future to increase power of the statistical tests when examining the effect of genre on vocabulary learning.

The Effects of Input Treatments Across Genres

In this study, genre was not found to interact significantly with input treatments

through the analysis of three-way ANOVA. Wilcoxon-Mann Whitney tests also

revealed that subjects given different texts under each input treatment learned and

retained a similar amount of vocabulary. In other words, the effect of each input

treatment remained the same across the two text types. Thus, concerning the initial

intention to incorporate genre in the study, that is, to examine whether learning

vocabulary from listening can only function well with a narrative text, it is tentatively

concluded that the listening treatment can be applied to both the narrative and the

expository text. However, the lack of interaction between genre and input treatment may be a result of poor power of the statistical test. Future studies with more satisfactory power are required to further clarify the effects of input treatments across genres.

Different Performance Across Two Test Types

In this study, two types of vocabulary tests were used to measure the participants’

knowledge of the target words. One was a test of passive recall requiring participants to translate the target words into L1 equivalents, and the other was a test of passive recognition which asked the participants to match the target words with their corresponding L1 translations. According to Laufer and Goldstein (2004), the passive recall is considered a more advanced measure of vocabulary knowledge than the passive recognition. They claimed that “Recall of information indicates a better memory trace than recognition of the same information …. A person who can recall the meaning of a given word is also likely to be able to recognize it among several options.” (p. 408) Such a claim about the difference between the two vocabulary tests is supported in the current study. All the participants in this study scored higher in the recognition test than in the recall test, suggesting that the recall test is more difficult and requires a deeper memory trace of the words. Moreover, the significantly better performance of the reading and the combination groups over the listening group dwindled evidently in the delayed recall test, resulting in no significant difference among the three treatment groups. However, when measured with the recognition test, the superiority of the reading and the combination groups over the listening group in the immediate posttest sustained in the delayed posttest. This result may suggest that limited effort from a single learning task may not be sufficient enough to establish a permanent memory trace of the words to be retrieved in a recall test one week later.

More times of encounters or practice of the target words might be required to help

score high in a recall test. In contrast, a recognition test is prone to elicit both the explicit and vague impression of the words and may better reflect the effects of different treatments.

Findings From the Post-Experiment Questionnaire

The post-experiment questionnaire in this study contained two main sections: (1) self-reported comprehension of the text; and (2) three dimensions of learning motivation, attention, and anxiety. Findings from the questionnaire will be briefly discussed in the following.

Concerning text comprehension, most participants reported good comprehension of the texts and there was no significant difference among groups in self-reported comprehension of the text. The result suggests that the difficulty of the text is appropriate for the subjects in this study. Moreover, the self-reported comprehension was found to be significantly correlated with vocabulary gains. Thus, the comprehensibility of a text may function as a predictor of how much vocabulary one learns from a text.

Though few participants considered the texts too difficult to comprehend, those

who found them difficult mainly ascribed the difficulty to the unknown words rather

than to the complex structure in the text. This result suggests that lack of knowledge

of the words in the text may be a main cause for less comprehension of the text in the

current study. Besides, it should be noted that more students from the listening

treatment, Group LE in particular, admitted difficulty in comprehending the text

owing to the unknown words than those under the combination treatment. Actually,

participants in each treatment group possessed similar English proficiency and were

assumed to have similar vocabulary size. Why participants in the listening treatment

found it more difficult to recognize the words in the text than the other treatment

groups might be attributed to their inability to decode the auditory text efficiently

rather than lack of knowledge of the words. That is, they might not have had so much difficulty recognizing the words if the texts had been presented visually as in the other groups.

This explanation corresponds to the assumption addressed in the previous subsection that students treated with aural input alone may learn less owing to limited spoken words to deal with the auditory texts.

The second section in the questionnaire measured the states of learning motivation, attention, and anxiety aroused by different treatments. On the one hand, no significant difference was found among groups in anxiety state. One the other hand, there were significant differences in the dimensions of motivation and attention among groups; however, the different degrees of learning motivation and attention were not significantly correlated with vocabulary gains and retention. These results suggest that input treatments aroused different degrees of motivation and attention but differences in these two aspects did not significantly affect the final learning outcomes.

A specific example is Group LE. Students in Group LE appeared to be highly motivated and attentive to the aural presentation of the text, but they failed to comprehend the auditory text as well as the other groups. Neither did they gain substantial vocabulary from it. Thus, the different performance among groups in the current study should be ascribed more to the treatments than to the “temporary”

cognitive and affective “states”. However, this study cannot determine whether different cognitive or affective traits, such as general learning motivation and trait anxiety, are related to vocabulary learning.

Though few implications on vocabulary learning can be generalized from the analyses of the three cognitive and affective states, they have at least helped explain

3

This assumption was later confirmed by an informal interview with 20 subjects from the two listening

subgroups. Ten nontarget words selected from the two texts for the experiment were presented aurally

first and then visually to the 20 students. All of them revealed more knowledge on the words when they

were presented visually than when they were presented aurally. This phenomenon is especially marked

with some common words like “blood,” “chance,” “disease,” “effect,” “smoothly,” and “amount.”

the discrepancy between Henderson’s (2001) study and the current one. Henderson (2001) once suggested that the interest level of storytelling contributed to the better performance of the listening treatment over the silent reading. However, participants assigned to learn vocabulary from listening to the narrative text in the current study did not reveal significantly more interest in the learning task nor learned vocabulary more effectively than the other treatment groups. When accompanied with the expository text, the listening treatment, however, revealed high learning motivation but still yielded limited vocabulary gains. Therefore, it is suggested that the positive effect of learning vocabulary from listening to stories and the increased interest aroused by the story telling activity may be confined to native speakers and cannot be applied to the L2 learners in the current study.

Implications

According to this study, learning vocabulary from reading, from listening, and from concurrent treatment of reading and listening contribute differently to vocabulary learning and retention. Some findings about the input treatments correspond to the claims made in previous studies and theories while some need further clarification from future research. In the following, theoretical implications from these findings are presented first, followed by pedagogical implications on vocabulary instruction for EFL learners.

Theoretical Implications

The different effects of the three input treatments on vocabulary learning in the present study have lent support to the assumptions of several theories (e.g., Mayer &

Sims, 1994; Sadoski, 2005) and echoed with the findings in previous studies (e.g.,

Baddeley, 1988). For example, the highest gain scores obtained from the combined

treatment of reading and listening in the current study have supported the dual coding

theory (Paivio, 1986; Sadoski, 2005) and the multimedia learning theory (Mayer &

Sims, 1994) that a combination between verbal and visual presentations of the words enhances learning. Besides, the better performance found for the reading treatment over the listening treatment corresponds to Baddeley’s (1988) suggestion that visual presentation facilitates vocabulary learning more effectively than auditory presentation for L2 learners.

Although the differential effects of the three input treatments on vocabulary learning in this study are supported by previous research, some of the results concerning vocabulary retention in this study are beyond expectation and lack direct support from previous research. For example, the combined treatment of reading and listening revealed a lower rate of vocabulary retention in comparison with the reading treatment and the listening treatment alone. This surprising result, however, was seldom addressed or explained in previous research. A tentative explanation for this result was proposed in the discussion section. With reference to the theory of attention and memory (Jacoby et al., 1993) and the cognitive overload theory (Kalyuga et al., 2004), it is argued that split-attention to two input sources in the combination treatment may result in neither input fully processed and in turn hinder the construction a deeper memory trace of the learned words. The finding of a lower retention for the combination group is interesting in that it suggests the more is not always the better. A combination of multiple learning sources does not always ensure an expected ideal result.

The listening treatment, on the other hand, reveals a consistent high rate of

retention on the learned words in both the present study and some previous studies

(Cheng, 2006; Elley, 1989). Therefore, the facilitating effect of aural input on

vocabulary retention appears to stand for both the L1 and L2 learners. Previous

studies (Elley, 1989; Henderson, 2001) tended to attribute the effectiveness of

learning vocabulary from listening to the interest level of story telling and learners’

persistent attention to the activity. However, this explanation was not supported in the current study. Subjects in the two listening groups in this study both revealed high retention of the learned words but only the group with the expository text reported significantly higher learning motivation over the other treatment groups. It is thus suggested that the high rate of retention should not be a product of promoted motivation or attentive learning. Another explanation for the effect of aural input was advanced in the discussion section. According to Baddeley (2003), aural input has a direct access to the phonological loop, a device for vocabulary acquisition, in which the unfamiliar sounds are rehearsed for a permanent memory to be established.

Conrad (1964) also claimed that linguistic materials have to be transformed into auditory representations to be retained in memory. All of these theories may lend support to the high rate of vocabulary retention of the listening treatment. However, in light of the initial low vocabulary gains of the listening groups, “floor effect” may also be a plausible explanation for the listening groups’ high retention rate. More research is required to further clarify this issue.

The reading treatment, in contrast, displayed a second-highest rate of retention of

the learned words and even outperformed the listening treatment in vocabulary

retention in terms of raw vocabulary scores. This result is contradictory to Cheng’s

(2006) as well as Waring and Takaki’s (2003) claim that vocabulary learned from

reading cannot be retained effectively. The finding is also different from that of Cheng

(2006), where the listening group retained significantly more vocabulary than the

reading group in the delayed posttest. Such inconsistent performance of the reading

treatment in vocabulary retention across studies is also confirmed by Hunt and Beglar

(2005). As suggested in the discussion section, the different performance in

vocabulary retention across studies could be related to differences in the

implementation procedures of the treatment or to the subjects’ different L2 proficiency levels and thus decoding skills. Future research may need to incorporate these variables into research design.

In sum, the effects of different input treatments on vocabulary retention have not yet been substantially examined. Only a few theories and empirical studies could be introduced to help interpret the results of vocabulary retention in the current study.

Therefore, more research is needed to clarify the effects of different input treatments on vocabulary retention and to develop potential explanations for their differences.

Pedagogical Implications

Different input treatments of reading, listening, and a combination of reading and listening are confirmed, in the current study, to contribute differently to vocabulary learning and retention. While the reading treatment and the concurrent treatment of reading and listening ensure better vocabulary gains, the listening treatment and the reading treatment alone seem more effective in retaining the learned words. To make the most of the strengths of different treatments in vocabulary instruction and learning, several suggestions about the implementation of each treatment are provided in the following for reference.

Learning from a text presented both visually and auditorily is a common practice

in a language classroom. Although the concurrent visual and auditory presentation of

a text facilitates vocabulary learning effectively, split-attention to the two input

sources presented simultaneously may be detrimental to the retention of the learned

words. Since both written and aural input are essential in facilitating vocabulary

learning for L2 learners, a separate presentation of the two input sources may be an

alternative. For example, students can be encouraged to process the written text by

themselves before the teacher’s explanation, which is usually conducted with both

written and aural input of the text presented simultaneously. After the teacher’s

explanation, the students can be required to review the text by silent reading or listening to the text read aloud. These procedures are expected to integrate the benefits of both types of input and minimize the chance to learn under split-attention.

On the other hand, since reading has been confirmed to be an effective source for vocabulary learning, learners are encouraged to learn vocabulary from extensive reading of appropriate materials such as magazines, newspaper, or graded readers, etc.

However, considering Conrad’s (1964) suggestion that linguistic materials is better retained when transformed into auditory representations, L2 learners should be reminded of the importance of letter-sound correspondence to ensure a better retention of the learned words. In a language classroom, teachers can also design tasks of silent reading once in a while for students to engage in active processing of the text, which may also serve as a variety to the common practice of simultaneous visual and auditory presentation of the text imposed by the teacher. In an activity of silent reading, however, a teacher should beware of whether students are attentively involved. To avoid distraction, presenting the text sentence by sentence through PowerPoint as conducted in the current study may be a feasible alternative.

The listening treatment, though less effective in facilitating vocabulary learning for students with limited spoken words in L2, can enhance retention of the learned words. Therefore, the activity of listening to stories read aloud can also be considered as part of teaching activities. To minimize the difficulty caused by lack of spoken words, a teacher may present the text slowly and pause to check students’

comprehension when necessary. Follow-up activities such as oral comprehension

questions or discussion can also be designed to accompany the implementation of

story telling. Once the words are learned from this source, as this study suggests, they

may be retained for a long time. For more advanced L2 learners with sufficient

spoken words, learning from auditory materials should be an ideal option since they

shall have little problem comprehending the text and could even benefit from the potential strength of this treatment, i.e., to have a high retention of the learned words.

Limitations and Suggestions for Future Research

The present study has attempted to examine the effects of different input treatments on vocabulary learning and retention across two text types for EFL learners.

Previous theories and studies were also elicited to help interpret the findings in this study. However, several limitations were found during analyses of the test results. In the following, the limitations of this study are presented respectively followed by potential solutions for future research. Moreover, other suggestions for future studies are also provided for further clarifying the effects of different input treatments on vocabulary learning.

In this study, genre did not have a significant effect on vocabulary learning. Nor did genre interact significantly with input treatment or time. These results, however, should be treated with caution owing to poor power found for the tests of genre effect.

Therefore, further studies are needed to increase the power so as to clarify the effect of genre as well as its interaction with input treatment on vocabulary learning. To increase the power, studies in the future may incorporate a larger sample size for the experiment or to improve the effect size of genre by adopting longer texts which may better reflect the features of each genre, among others.

According to the Discussion section, subjects under the listening treatment might

learn less because of lack of spoken words in L2. Whether more advanced L2 learners

with better L2 proficiency or ability in letter-sound correspondence can learn well

from listening deserves further investigation. On the other hand, Baddeley (1988)

suggested that subjects learned well from visual presentation of the words alone since

both visual and auditory encoding of the words are available when they are presented

visually. Conrad (1964) also proposed that linguistic materials presented visually have to be transformed into phonological forms to be retained in memory. Based on these two studies, whether students’ performance in vocabulary learning and retention under the reading treatment is correlated to their language proficiency or ability in letter-sound correspondence requires further examination. In short, further studies may include subjects of different levels of L2 proficiency or ability in letter-sound correspondence to examine their performance under each input treatment.

Previous studies have attempted to examine the effects of different input modalities on vocabulary learning (Baddeley, 1988; Henderson, 2001; Lee, 2006;

Militante, 2006). However, few studies went further to compare the retention from these learning sources. Although this study was aimed to fill this gap, some unexpected results found here need further clarification. For future studies, several issues can be further explored, including whether the dual input treatment reveals lower retention than each single input source, or whether the listening treatment always ensures a high rate of retention.

In addition, the simultaneous presentation of the written input and aural input was found to facilitate vocabulary retention less effectively. Whether separate presentation of written input and aural input would outperform the concurrent presentation of the two input sources deserves further examination.

Finally, studies concerning longer periods of retention or the differential effects

of longitudinal reading and listening programs are also expected in the future to better

clarify the effects of different input treatments and to glean insights into vocabulary

learning.