CHAPTER FOUR RESULTS
S- G d Non-G e Overall
Note: **Correlation is significant at the 0.01 level (2-tailed).
a TA I = Transaction Belief I – Reader Interpretation
b TA II = Transaction Belief II – Reader Engagement
C TM = Transmission Belief
d S-G = General Reading Strategy Factor
e S-Non-G = Non-General Reading Strategy Factor
As can be seen from the table, similar to the results obtained from the L1 data, in L2, all three types of beliefs, TA I, TA II, and TM, were found to be significantly correlated with the overall strategy scales as well as the two strategy subscales, with correlation scores ranging from .37 to .66 (p<.01). Again, the results also suggested that when reading in L2, students’ beliefs about reading were significantly correlated with their reported use of reading strategies. In addition, the correlation scores in Table 13 also revealed that among the three beliefs types, while the TA II – Reader Engagement type appeared to have the highest correlation scores with all the strategy categories (.64 with the General, .61 with the Non-General, and .66 with the overall use), the TM belief type was found to have the lowest correlation scores with all the strategy categories (.37 with the General, .38 with the Non-General, and .40 with the overall use). The results seemed to indicate that when reading in L2, there was a
closer link between the subtype of transaction belief, the reader engagement, and the reported use of reading strategies in all categories. On the other hand, the results also suggested a possibly less strong link between the transmission belief type and the use of reading strategies in general.
One-way ANOVA Test
As described in the previous section, in order to further investigate the relationship between the students’ beliefs about reading and their reported use of strategies across the languages, the study examined a group of students identified as having distinctive beliefs about reading in L1 and L2. A total of 68 students were selected from the whole sample group (N = 432). These students consisted of four groups of readers with different belief profiles: Type A (L1TA-L2TA), Type B (L1TA-L2TM), Type C (L1TM-L2TM), and Type D (L1TM-L2TA). The number of subjects contained in each group was 41 in Type A, 12 in Type B, 11 in Type C, and 4 in Type D. Note that a considerably higher number of students (n = 41) were found in Type A group, whereas fewer students were included in Type D group (n = 4).
After the selection, One-way ANOVA was employed to examine the differences in strategy use among these four groups of readers. Table 14 shows the mean scores and standard deviations of reported use of reading strategies in L1 by the four groups of readers. As revealed from Table 14, there were observed differences in the reported strategy use among groups. In L1, the mean scores of Types A and B readers appeared to be higher than those of Types C and D readers in all strategy categories. It should be noted that concerning reading in L1, Types A and B readers were those who scored high on the transactional subscale, whereas Types C and D were those who had high scores on the transmission subscale. These observed differences seemed to suggest that readers with strong transactional (TA) beliefs tended to apply more reading strategies than those with strong transmission (TM) beliefs when reading in L1.
Table 14.
Mean scores and standard deviations of reported use of reading strategies in L1 by the four types of readers 15, a similar situation was found regarding the four group of readers’ reported strategy use in L2. It was found that Types A and D readers had higher mean scores than the rest of the two groups (i.e. the Types B and C readers) in all strategy categories. Note that when reading in L2, Types A and D readers consisted of those who scored high on the TA subscale, whereas Types B and C comprised those who had high scores on the TM subscale. Again, a similar conclusion can be drawn based on these observed differences. There appeared to be a tendency for readers with strong TA beliefs to reportedly use more reading strategies than those with strong TM beliefs in either L1 or L2.
Table 15.
Mean scores and standard deviations of reported use of reading strategies in L2 by the four types of readers
However, the results of ANOVA test revealed that none of the differences were found to be statistically significant. More specifically, in L1, no significant differences were found among groups in the General Reading Strategy Factor (F (3, 63) = .74, p
= .53), Non-General Reading Strategy Factor (F (3, 63) = .50, p = .69), and overall use of reading strategies (F (3, 62) = .71, p = .55). Likewise, in L2, no significant differences existed among groups in the General Reading Strategy Factor (F (3, 62) = 2.62, p = .06), Non-General Reading Strategy Factor (F (3, 61) = .41, p = .75), and overall use of reading strategies (F (3, 61) = 1.25, p = .29).
The Relationship Between Readers’ Beliefs and Actual Use of Reading Strategies in L1 and L2
This section presents results obtained through the use of qualitative approach.
The think-aloud method was utilized to collect information about learners’ actual use of reading strategies when reading across the two languages. The verbal reports were collected from a selected sample of 22 students. As described in the previous chapter,
these participants comprised three types of readers (after the exclusion of the only type D reader) classified as having distinctive beliefs about reading in Chinese and English (i.e. high TA or high TM in both languages). These three types of readers were: Type - A (L1TA-L2TA), Type- B (L1TA-L2TM), and Type - C (L1TM-L2TM).
In the following part, I will first present the frequency of strategy use across the two languages. Then I will begin to report general patterns of strategy use in terms of the three reader types. After that, a more in-depth report of the strategy patterns for each group of readers will be presented.
General Patterns of Strategy Use across Languages
A total of 1,042 reading strategies were coded in the present study using the classification scheme developed in the study (see Appendix N). There are 25 types of strategy in the present scheme divided into two broad categories: General Strategies (15 types) and Local Strategies (10 types). The actual occurrences of individual strategy type in L1 and L2 can be found in Appendices O and P. Note that the two levels of strategy categorization (i.e. general versus local reading strategies) utilized here should be distinguished from the two strategy factors (i.e. general versus
non-general reading strategy factor) identified from the Chinese and English Reading Strategy Questionnaires (RSQs). As described in Chapter Three of this paper, the present coding scheme was expanded from the work of Cheng (1999), which was based on Block’s (1986) two-level conceptual framework of reading strategies. It is, therefore, considered reasonable to follow the original and more conventional distinction of general versus local reading strategies.
Results of Chi-square tests
Table 16 presents the frequencies of strategy use across the two languages. As shown in the table, there were a total of 399 strategies coded in L1, among which 246
(62%) belonged to the general strategies and 153 (38%) belonged to the local strategies. In L2, 643 strategies were found including 115 (18%) general strategies and 528 (82%) local strategies.
Table 16.
Frequencies of reading strategy used across the two languages Strategy
Language
Type Frequency %
General 246 62
L1 Local 153 38
TOTAL 399 100
L2
General Local TOTAL
115 528 643
18 82 100 Note: N = 22
Two general observations can be made from these findings. First, concerning the total number of strategy used across the languages, the total number of strategies used was found to be considerably higher when reading in L2 than in L1 (643 versus 399). The result of the Chi-square test revealed that the difference was significant (χ2
= 57, df = 1, α = .05). Second, regarding the distribution of strategy within each language, the table shows that in L1 the frequency of general strategies appeared to be higher than the frequency of local strategies (62% versus 38%), whereas in L2, the reverse seemed to be true in that the frequency of local strategies was higher than that of general strategies (82% versus 18%). Again, the results of subsequent Chi-square tests confirmed that while in L1, readers tended to use significantly more general than local reading strategies (χ2 = 21.62, df = 1, α = .05), in L2, they used a significantly greater proportion of local than general reading strategies (χ2 = 265.36, df = 1, α
= .05).
The five most frequently used reading strategies across languages
The participants’ apparent different use of strategy patterns between reading in
L1 and L2 was further supported by the data presented in Table 17, which lists the five most frequently used strategies in the two languages.
Table 17.
The five most frequently used strategies across the two languages
Language Strategy (Type) Frequency %
L1 (1) Question meaning of a word or phrase (L a) (2) Re-read (L)
(3) Use background knowledge (G b) (4) Question information in the text (G) (5) Comment on text structure (G)
58
(2) Question meaning of a word or phrase (L) (3) Word solving behavior (L)
(4) Question meaning of a sentence (L) (5) Comment on behavior or process (G)
231
a L = Local Reading Strategies
b G = General Reading Strategies
As shown in Table 17, in L1, three out of the top five frequently applied strategies (i.e.
to use background knowledge, to question information in the text, and to comment on text structure) belonged to the general strategy category. More specifically, these three strategies accounted for approximately 30% (11% + 10% + 9%) of the total strategy use in L1. Furthermore, although both of the top two strategies in L1 (i.e. to question the meaning of a word or phrase and to re-read) belonged to the category of local strategies, their total occurrences were only about 28% (15% + 13%). This indicated that among the top-five strategies in L1, the occurrences of the two strategy types, general versus local strategies, appeared to be rather evenly distributed (30% versus 28%).
In contrast, in L2 (see also Table 17), four out of the top five most frequently used strategies were local strategies, which accounted for nearly 74% (36% + 22% +
8% + 7.9%) of the total strategy use. The occurrences of the top two strategies alone (i.e. translation and questioning the meaning of a word or phrase) were found to be over 50% (36% + 22%) of the total strategy use. The only type of general strategy that was frequently applied in L2 was to comment on the reading behavior or process, which accounted for 7.6% of the total use. The results seemed to suggest that when reading in L2, students tended to heavily use several local reading strategies.
Moreover, there was a considerable difference in the distribution between the two types of strategy, general versus local strategies (7.6% versus 74%), among the top-five strategies in L2.
General Patterns of Strategy Use by Reader Types
As aforementioned, the qualitative analyses of the study focused on the three groups of readers with different belief profiles across the languages. Their use of reading strategies collected from the think-aloud tasks were analyzed and compared.
The first group of readers, Type A readers, consisted of ten students who were found to score high on the transaction scales of both Chinese and English RBQs. They were classified as having transaction beliefs about reading in L1 and L2 (TA – TA). The second group of readers, Type B readers, comprised six students who had particular high scores on the transaction scale of Chinese RBQ but were also found to score high on the transmission scale of English RBQ. In other words, they held very different beliefs about reading between reading L1 and L2 (TA – TM). The last group, Type C readers, was made of six students who had particular high scores on the transmission scales of the two RBQs (TM – TM). In contrast to their Type A counterparts, these readers were classified as having transmission beliefs when reading across the
languages. This section will briefly report the general strategy patterns utilized by the three groups of readers. In the section that follows, the strategies frequently used by each group of readers will be explicated in greater details.
Results of Chi-square tests
Table 18 shows the frequencies of reading strategies used by the three groups of readers across the languages. First, concerning reading in L1, when comparing the these groups in terms of the frequencies of the two broad strategy types, general versus local, a clear difference was found between readers from the Type C group and the other two groups. More specifically, for both Type A and Type B readers, in L1, the frequencies of general strategies were much higher than those of local strategies (67% versus 33% for Type A; 68% versus 32% for Type B). However, for Type C readers, the difference appeared to be shown in the opposite direction in that the frequency of local strategies was found to be slightly higher than that of general strategies (General 48% versus Local 52%).
Table 18.
Frequencies of reading strategies used by the three reader types and across the two languages
To examine the observed differences, Chi-square tests were again conducted.
The results showed that while both Type A and Type B readers used a significantly greater amount of general than local strategies when reading in L1 (χ2 = 8.94, df = 1, α = .05 for Type A; χ2 = 16.68, df = 1, α = .05 for Type B), no significant difference was found between the two types of reading strategies used by Type C readers (χ2
= .46 < 3.84, df = 1, α = .05). Since what differentiated the Type C readers from the rest of the reader groups was their strong endorsement to the transmission beliefs when reading in L1, it was reasonable to suggest that their distinctive pattern of strategy use was related to their underlying implicit models of reading. Furthermore, the results seemed to indicate that when reading in L1, there was a connection
between the stronger transaction beliefs and the more frequent use of general reading strategies.
Next, as for reading in L2, Table 18 shows a completely different picture. For all three groups of readers, the frequencies of local strategies were found to be much higher than those of general strategies (23 % versus 77% for Type A; 14% versus 86%
for Type B; 12% versus 88% for Type C). The results of subsequent Chi-square tests also revealed that all the differences reached the significant level (χ2 = 94.48, df = 1, α
= .05 for Type A; χ2 = 88.92, df = 1, α = .05 for Type B; χ2 = 88.36, df = 1, α = .05 for Type C). In other words, when it comes to reading in L2, these readers showed a predominant use of local reading strategies regardless of belief types.
Type A (L1TA-L2TA) Readers’ Use of Reading Strategies
Type A readers consisted of 10 students who were classified as having strong transaction beliefs about reading in both languages. However, as previously presented, they were found to behave differently when reading across languages. More
specifically, these readers tended to apply significantly more general strategies than local strategies in L1, but they tended to show predominant use of local reading strategies in L2.
The top-five strategies used by Type A (L1TA-L2TA) readers
To further demonstrate the strategy pattern adopted by this particular group of readers, the study examined their uses of individual strategies across the languages.
Table 19 shows the five most frequently applied strategies by Type A readers in L1
and L2. As it can be seen from the table, three out of the top five strategies in L1 were general reading strategies, to use background knowledge, to question information in the text, and to make an inference with each accounting for 15%, 11%, and 10 % of the total strategy use. The other two frequently applied strategies in the L1 list were two local reading strategies, re-reading and questioning the meaning of a word or phrase, with the percentages of 13.5% and 12% respectively. In general, as for the five most frequently used strategies in L1, the total percentage for the three general
strategies was found to be slightly higher than that for the two local strategies (36%
versus 25.5%).
Table 19.
The five most frequently used strategies by TypeA (TA-TA) readers in L1 and L2
Language Strategy (Type) Frequency %
L1 (1) Use background knowledge (G) (2) Re-read (L)
(3) Question meaning of a word or phrase (L) (4) Question information in the text (G) (5) Make an inference (G)
(2) Question meaning of a word or phrase (L) (3) Comment on behavior or process (G) (4) Word solving behavior (L)
(5) Question the meaning of a sentence (L)
102 Note: N=10; L = Local Reading Strategies, G = General Reading Strategies
However, concerning the top-five strategies in L2, the results appeared to parallel the previously found patterns of strategy use across languages in which there was a predominant use of local reading strategies in L2. As shown in Table 19, readers tended to heavily use the same four local strategies (c.f. Table 17), to use translation, to question the meaning of a word or phrase, to use word-solving strategy, and to question the meaning of a sentence when reading in L2. These strategies alone accounted for nearly 68% of the total strategy use in L2. The only general strategy in
the L2 list was also found to be the strategy of commenting on the reading behavior or process, accounting for about 10% of the total strategy use.
In addition to the frequency counts in Table 19, Type A readers’ use of these strategies will be exemplified through instances from their think-aloud protocols.
The first three excerpts (Excerpts 1 to 3) provide examples of how Type A readers used these general reading strategies: to use background knowledge to help interpret text meanings, to question information in the text, and to make an inference when reading in L1.
From Excerpt 1, it can be seen that the reader (Participant #10) tried to interpret the meaning of the sentence (i.e. Lao-Zi’ s emphasis on the peace without active human intervention) by drawing on her background knowledge gained from watching TV shows, which often implicitly delivered the moral doctrine of making peace with people. Although she did not quite understand the contrast of “peace with and without active human intervention” described in the text, she did utilize relevant background knowledge to help her better understand the text.
Excerpt 1. (Participant #10)
(English translations are provided and shown in italics).
She read (in L1),
“可見,重要的是「無為」的和, 而非「有為」的和”
“Therefore, what considered important (in Lao-Zi) was the peace without human intervention rather than the peace with human intervention.”
“This seems to say that we should value peace among people. This is just like what we see on TV. They often ask people to make peace and don’t fight with one another.
It’s better to value peace among people.”
In the second excerpt, the reader (Participant #13) explicitly expressed her doubts about the truthfulness of the sentence she had just read. The sentence she just
read mentioned one of Lao-Zi’s theories of being tender and weak, which was obviously different from her own conception of the theory of Lao-Zi. Instead of passively taking the information presented in the text, she first pondered on the truthfulness of information by asking herself twice and then explicitly expressed her doubts.
Excerpt 2. (Participant #13)
(English translations are provided and shown in italics).
She read (in L1),
“就倫理觀而言,儘管老子與聖 經均強調柔弱或軟弱的價值,但 兩者在本質上有極大差異”
“As far as the views of ethics are concerned, although both Lao-Zi and Bible emphasized the
importance of being tender and weak, they were essentially different in concepts.”
She said (in L1),
“老子? 是強調柔弱或軟弱的價 值? 我有點懷疑”
“Lao-Zi? Did he emphasize the importance of being tender and weak? I doubt it.”
In Excerpt 3, we see an example for the strategy of making an inference. In her think-aloud protocols, the reader (Participant #14) first identified the purpose of the text she just read (i.e. to make a comparison between Christianity and Daoism) and then she made a hypothesis about the writer’s intention by pointing out the author’s position of favoring the ethics of Daoism. Although during the entire text, the author did not explicitly express in words about his personal preference regarding the two theories of ethics, based on what she read, the reader was able to reasonably infer about the author’s intention.
Excerpt 3. (Participant #14)
(English translations are provided and shown in italics).
She read (in L1),
“由此我們可以看出：基督教的 倫理學不同於道家的倫理觀,它 從不以尋求人與人之間的和諧 為依歸”
“From this we can see that the ethics of Christianity are different from those of Daoism. The
purpose of Christianity was never to seek peace among people.”
She said (in L1),
“看到這邊我就想說他是在拿基 督教倫理學和道家倫理學做比 較,然後作者本身好像比較偏向 道家的倫理觀”
“I think he (the author) was trying to compare the ethics of
“I think he (the author) was trying to compare the ethics of