Chinese (71%)
For those 15% of the participants who stated that they would rather choose English pre-reading input, their top three reasons can be summarized as (1) To maintain consistency, the supplementary materials should be in the same language as the academic article (36 times); (2) Reading in English would facilitate English learning (9 times). (3) English major students were used to reading texts written in English (3 times). For the participants who specified that language input did not make a difference, this group of the participants tended to indicate that both languages played equal importance and were accepted in reading (19 times).
In sum, the survey questionnaire found that L1 seemed to play an auxiliary role for many (71%). Realistically, L1 should be used as psychological and emotional support and also a time-saving devise. However, those who preferred English pre-reading input (15%) believed that L2 should be used whenever possible to increase their opportunity in English practice.
Statistical Analysis on Perceived Effects of Pre-reading Input
This study also explored whether the preference for pre-reading language input differed between the participants of varying content knowledge and English
proficiency levels. Focusing on the analysis of survey questionnaire, the researcher starts with reporting results from Chi-Square tests which compared the participants‘
perceived preferences for L1/L2 pre-reading input between H/L content knowledge level, and also between H/L English proficiency level. In the second section, the response analyses, which reflect the findings of chi-square results, are detailed qualitatively to answer the last research question of the present study.
These data, together with the findings from ANOVAs, can be converged in a cross-referenced fashion to see whether there is a gap between the actual effects and
the attitudinal differences for the pre-reading language inputs.
Learners’ Content Knowledge Levels and Their Preferences for Pre-reading Input Whether content knowledge level (high/low) and English proficiency level (high/low) were associated with the participants‘ preferred type of learning medium (Chinese vs. English) is one of the focuses of the study. Therefore, Pearson's
Chi-Square tests were applied to explore if there was a relationship between each of the two categorical variables: the preferred input language and content knowledge levels; and the preferred input language in the pre-reading stage and their English proficiency levels.
A few students did not answer the last survey question regarding their preference for L1/L2 pre-reading support; in total, 90 of them provided the written responses. The first Pearson Chi-Square indicates that for the variables between content knowledge levels and preferred reading, the result is χ² (2) = 0.812, p= .666.
This demonstrates that there is no statistically significant association between levels of content knowledge and preferred language. That is, these two levels of the participants equally prefer Chinese/English pre-reading materials. Table 24 is a cross-tabulation table which shows the counts and the percentage of the participants‘
reported choices. Since the designed survey questionnaire was open-ended in nature, apart from the language choice of L1 or L2, some participants also stated that both languages worked for them. Figure 16 below indicates the counts of the reported choices between the groups. The same figure also shows that both high and low content participants preferred Chinese input over English, as it has the statistically highest counts (percentage) of participants opted for this choice.
However, it was noted that the Chi-Square test might be inappropriate for tables with small expected cell frequency. One cell in Table 24 has the frequency count of 4. To avoid being unreliable, the researcher combined two cells to run
another Chi-square test. As some participants indicated that both L1 and L2 would work for them, therefore, those who indicated ―both‖ were combined with those who preferred only ―L2,‖ as these participants can be subsumed under the category of the participants who accepted L2 input.
Table 24. Cross-tabulation of Content High/Low x Preferred Language (N=90) Levels/Preferred Language
Both L1 L2 Total
Content Low Count 6 32 6 44
＊within 13.6% 72.7% 13.6% 100.0%
◎between 60.0% 46.4% 54.5% 48.9%
Content High Count 4 37 5 46
Within 8.7% 80.4% 10.9% 100.%
between 40.0% 53.6% 45.5% 51.1%
Total Count 10 69 11 90
% of Total 11.1% 76.7% 12.2% 100.%
＊Percentage of language choice within content level
◎ Percentage of language choice between content level
Figure 16. Preferred pre-reading language input between H/L content knowledge groups (N=90)
0 10 20 30 40 50 60 70
Both L1 L2
Content Low Content High
After the combination, another Chi-square test was run. Again, the result was insignificant, as the χ² (1) = 0.748, p= .387. This tells us that there is no statistically significant association between content and preferred pre-reading language input; that is, both higher and lower content knowledge participants equally prefer Chinese or English pre-reading input.
Learners’ English Proficiency Levels and Their Preferences for Pre-reading Input The Pearson Chi-Square result for the variables between English proficiency levels and preferred pre-reading input was significant, χ² (2, N = 138) = 18.845, p= .000. This result indicated that there was a significant relationship between levels of English proficiency and the preference for pre-reading language input. The strength of association between the variables was moderate (0.36) as calculated by Phi, which means that the participants‘ English proficiency was moderately related to their preference for the language of the pre-reading input.
Table 25 shows that a difference exists between English proficiency levels and the preferred language. Eighty-one percent of the participants at the higher English proficiency level preferred the L2 pre-reading material; however, only 19% of the participants at the lower English proficiency level preferred English pre-reading input.
Sixty-one percent of the participants at the lower proficiency level preferred the L1 pre-reading input; but only 39% of the participants at the higher English proficiency level preferred Chinese (See comparison also on Figure 17). However, notice that in Table 25, two cells have low frequency counts (less than 5). To avoid being unreliable, the researcher ran another Chi-square test by collapsing ―both‖ and ―L2‖ as shown in Table 25, as these participants can be subsumed under the broader category of
―accepting L2 input.‖ This Chi-square test is also significant, χ² (1) = 18.842, p= .000.
The results from these Chi-square tests might be intuitive as it confirmed the notions that less proficient participants might be more reluctant to read English while the higher proficient participants accepted the challenges of receiving more materials in the target language.
Table 25. Cross-tabulation of English High/Low x Preferred Language (N=138) Levels/Preferred Language Both L1 L2 Total
Eng Low Count 2 65 4 71
＊within 2.8% 91.5% 5.6% 100.%
◎ between 18.2% 61.3% 19.0% 51.4%
Eng High Count 9 41 17 67
within 13.4% 61.2% 25.4% 100%
between 81.8% 38.7% 81.0% 48.6%
Total Count 11 106 21 138
% of Total 8.0% 76.8% 15.2% 100%
＊Percentage of language choice within English proficiency level
◎Percentage of language choice between English proficiency level
Figure 17. Preferred pre-reading language input between H/L English proficiency groups (N=138)
0 10 20 30 40 50 60 70
Both L1 L2
Content Low Content High
Analysis on Written Response for Preferred Language
This section will start with analyzing answers on the questionnaire regarding the participants‘ perception toward the pre-reading input of different languages. The researcher placed more weight on detailing certain experiences mentioned by a few participants to reveal what troubled the EFL students when reading an academic text.
Their answers might provide a portal into understanding what challenges in academic readings were faced by the participants and how pre-reading materials were utilized by them to make sense of the domain-specific article.
Participants Preferring L1 Pre-reading Input
The following examples illustrated why the participants made the choices of selecting pre-reading input in L1. The data indicated different tendencies regarding the reasons between the high/low English proficiency participants. Among the higher proficiency participants, for those who preferred L1, the reasons provided for their choice were associated with the efficiency, or as one participant put it, for the sake of expediency in absorbing new information. For the higher-proficiency participants, because not all academic L2 reading could be attuned to an understandable level, reading efficacy, i.e., to save time, is their justification to use L1; the lower proficiency participants, on the other hand, were likely to give priority to the L1 pre-reading because they claimed L1 lowered their anxiety. It appeared that the use of mother tongue could provide a certain degree of comfort and confidence to the
participants. For lower proficiency participants, as L2 academic reading involved a variety of psychological and emotional considerations, L1 served to reduce their affective barriers that may inhibit comprehension.
The following examples are responses from two of the higher proficiency participants. The first participant put focus completely on the efficiency of time. In
the second example, the participant put more emphasis on the efficiency of content knowledge absorption.
Reading Chinese materials saves my energy and time (MCHE-28) 中文讀補充資料省時省力。
It works better if we apply our mother tongue to read something we do not quite understand. Compared with English supplementary materials, Chinese could be mediation for fuller knowledge absorption of new things, and more effective, too (LCHE-210).
用母語讀不懂的東西，效果會比較好。比貣英文的輔助，中文是比英文更 好的媒介來學新的東西，效果應該更好。
The responses below are from participants whose English proficiency level was relatively lower. As shown by the results in Table 25, the reading attitude exhibited by the four lower proficiency readers were influenced, at least to some extent, by
affective factors.
Table 26. Reasons of Why L1 Is Preferred by Lower Proficiency Participants Participants Complete response Affective factors identified (MCLE-76) I think Chinese provides me
with more information. The Chinese materials have empowered me. Reading Chinese helps me engaged with the later content with a relatively ease.
我覺得中文會給我比較多，中 文賦予的力量比較強，讀中文 後感覺後面比較容易懂。
Feeling empowered, feeling at ease
Table 25. (continued)
(HCLE-16). (I think) the advantage of Chinese is giving me more confidence. As Taiwanese, we should read more Chinese 中文的好處是給我信心，台灣 人本來就應該讀中文。
Gaining confidence,
assuring the identity as Taiwanese
(HCLE-16). Reading Chinese provides a deeper grasp of the English article. When reading Chinese, I understand every word which is comparatively comfortable for me.
讀完中文能幫我深入(了解)原 文，沒有看不懂的字讀來比較 輕鬆。
Clarity in understanding, feeling more comfortable
Feeling empowered, confident, and less pressured is the psychological merit of adopting materials in Chinese, as it created a supportive atmosphere in reading. For the lower proficiency participants, this marked predisposition towards preferring their pre-reading materials in L1 has made Chinese material an additional resource to their comprehension in L2 academic reading. The above analysis on the opinions of lower proficiency participants indicates the interconnection between psychological
components and readers‘ reading proficiency level.
Participants Preferring L2 and Participants with No Special Preferences The results of the survey questionnaire showed that among the total 300 participants, even though the majority of the participants (71%) opted for L1
pre-reading input, there were still 15% of the participants who preferred English input,
and about 11% of participants who mentioned both inputs would be beneficial for academic reading comprehension. Compared with the lower proficiency readers, a relatively larger number of higher proficiency participants preferred L2 as their pre-reading input, and also a high percentage of higher proficiency participants stated that either L1 or L2 would work for them. The reasons for L2 preference were that they would like to maximize L2 input or interact with L2 reading materials whenever they could. These English majors had higher expectations for their English
proficiency. They thought that the greater the exposure to L2 texts, the quicker they will be able to master English.
We need to read the academic article in English, it is better to think in English all the time by reading everything English (HCHE-27).
文章是用英文寫的，應該用英文思考，所有的閱讀都寫英文比較好。
I major in English, English reading should be natural for me. Besides, I need to read more English to prepare for studying abroad (LCHE-200).
英文系應該習慣用英文，也方便我準備留學考試。
And for those who claimed that both languages worked, they seemed to be comfortable working in two languages. However, some still mentioned that they would prefer the assigned L2 text to be written in easier English. The following are examples of the higher proficiency participants who preferred L2 and of those who did not have a special preference.
As long as I obtain the necessary concepts from the supplementary materials, it does not matter which language the concepts are written in (MCHE-49).
只要有補充到概念，什麼語言皆可。
I think it does not matter whether it is Chinese or English, but if it is English, I would like it to be written in easier English (LCHE-208).
我覺得中英文都沒差，但是如果是英文希望是簡單一點的英文。
Reading a domain-specific article in English was a complex cognitive activity.
When asked how to effectively improve reading performance and what was the optimal pre-reading language input, the results indicated that many students had a preferred language which they believed would benefit them in the pre-reading stage.
For participants of different English proficiency levels, the answers appear to be different. The analysis has, to a certain extent, clarified the effects between L1 and L2 pre-reading for the comprehension of academic reading perceived by participants of different English proficiency levels.
Summary and Conclusion
In summary, answers to research questions under investigation were reported in this chapter. In the beginning, results of descriptive statistics were detailed. The description included the basic features of the participants‘ performances on the political content knowledge test and the GEPT reading proficiency test. The descriptive statistics helped to pave the way for deeper interpretations of the data.
After the presentation of descriptive statistics, the results from three ANOVAs to investigate the effects of pre-reading inputs, content knowledge, and English
proficiency on academic reading comprehension were reported. The results from the ANOVAs showed the factors that impacted the actual performance of academic reading comprehension. In the last section, the analysis of written responses from the participants on the survey questionnaire was presented. Several quotes from the participants were applied to illustrate their perceptions and emotions. The findings of the present research can be summarized into the following major points.
Equal Effects of L1 and L2 Pre-reading Support
The first ANOVA conducted to investigate the effects of pre-reading inputs indicated that there was no significant difference between the performances of the participants who received L1 or L2 treatments. These two groups performed equally well and significantly better than those who did not receive any treatment.
In addition, the survey questionnaire regarding the effects of pre-reading support revealed that the participants were generally positive about the pre-reading treatment.
The majority of participants expressed positive attitudes toward the beneficial effects of the support in pre-reading on improving academic reading comprehension and provided their own explanations.
Overwhelming Preference for L1 Pre-reading Input
The finding from the participants‘ reported preference was in conflict with their actual comprehension performance. The majority of participants in this study
preferred and believed that L1 was more beneficial than L2 pre-reading input. In the higher English proficiency group, the participants preferred L1 mainly because, compared with L2, L1 was more efficient in absorbing new information. In the lower English proficiency group, L1 is preferred mainly for its emotional support in L2 academic reading.
More Participants in Lower English Reading Proficiency Preferred L1
The results of Chi-Square test indicated that L1 was much more preferred by the participants whose English proficiency was lower than those whose English
proficiency was higher. Although the two-way ANOVA showed that no matter which language input the participants at the lower English proficiency level received, they performed equally well in domain-specific academic reading. The participants in the
lower English proficiency group believed that L1 input could benefit their L2 reading more than L2 for various reasons. Detailed analysis of the direct quotes extracted from the participants‘ responses found that most reasons focused on the fact that L1 could bring psychological merits and reduce anxiety in L2 academic content area reading.
Significant Effect of Content Knowledge and English Proficiency
Statistically, the participants with higher content knowledge performed
significantly better than those with lower content knowledge. Also, the participants with higher English proficiency performed significantly better than those with lower English proficiency. Content knowledge and English proficiency play major roles in domain-specific reading comprehension.
On the survey questionnaire, the findings revealed that the participants perceived content knowledge to be important mainly because they believed it entailed a deeper understanding of the academic text. As for why they considered the English
proficiency to be important, aside from English proficiency compensated for their lack of background knowledge, many pointed out that the language skills involved seem to be indispensible to L2 academic reading.
Lower Content Knowledge/Higher English Proficiency Learners Attached More Importance to Background Knowledge
The analysis of the survey found that most participants were cognizant of their utilization of different knowledge bases, i.e., content knowledge and English
proficiency, during reading. In general, there was no universal consensual view among the participants on the roles the two knowledge bases played in
domain-specific academic reading. The degree to which why one knowledge base
dominated in an individual‘s reading experience over the other, however, varied and was contingent upon the individual‘s understandings of their own strengths and weakness in L2 reading.
However, the descriptive statistics also indicated that the participants with the combinations of higher English proficiency and lower content knowledge are the group who attached more importance to background knowledge. These participants realized that English itself was not sufficient to comprehend academic reading, the next step to learn more in content area relied heavily on their content knowledge.
To conclude, this chapter reported findings of Taiwanese EFL university students‘ performances after receiving the pre-reading treatments and also their attitudes toward pre-reading input in different languages. As the participants had different levels of content knowledge and English proficiency, the researcher sought to identify the beneficial effects due to the interactions. However the interactions between knowledge bases and pre-reading input were not statistically significant in their actual academic reading performance.
The statistical results showed that the participants could benefit from two language sources equally, and the higher their English and content knowledge, the better the performance in domain-specific academic reading they had. The
participants‘ self-reported perception toward the effects of L1/L2 pre-reading input, however, was at odds with their actual comprehension performances. With the
participants‘ self-reported perception toward the effects of L1/L2 pre-reading input, however, was at odds with their actual comprehension performances. With the