CHAPTER FIVE
DISCUSSION AND IMPLICATIONS
This chapter is composed of five parts. First of all, the summary of the findings will be presented. Discussion and explanation of the findings form the second part. The third section illustrates pedago gical implications. What follows is a presentation of implications for future study. Conclusions will be provided in the end.
Summary of Findings
The present study aims to investigate the effects of QtA lessons on recall, inference generation, three types of comprehension questions and students’ perception toward the training. The answers to the research questions will be presented and elaborated.
The answer to the first research question “Do QtA lessons facilitate the recall of texts by junior high school students?” is from two types of data, recall and the students’ answer to the factual questions, the first level of comprehension questions.
QtA training does not have a facilitating effect on the memory of the story in the recall, nor does it make a difference in answering factual questions between Group E and Group C. Thus from these two results, it could be concluded that QtA training did not enhance students’ ability in remembering the text or retrieving information from the text.
The answers to the second research question “Do QtA lessons produce a facilitating effect on three levels of comprehension for junior high school students?”
is that training makes a difference in reader-based inference and answer to the interpretation questions, which is comparable to text-based inference, but not in
text-based inference and answer to responsive questions, which is similar to reader-based inference. These results are also based on two types of data: two levels of inference made in the recall, text-based and reader-based inference, and answers to two levels of comprehension questions, interpretation questions and responsive questions. The two types of data differ in that the former is generated by students autonomously while the latter is cued by the teacher. Nevertheless, the nature of the knowledge source makes the two comparable – text-based inference and answer to the interpretation questions are similar while reader-based inference and answer to the responsive questions are approximate.
Hence, it may be well said that the training encouraged more activation of reader knowledge to make sense of text, especially when readers are free to resort to their knowledge source without constraint from teacher cues. Therefore, after training, students in Group E improved more reader-based inference in their free recall than students in Group C. However, when the reader-text interaction is subject to the external constraint, as in answering comprehension questions, readers’
knowledge-text interaction is confined and the activation of knowledge has to lower to the text level, making the effect of training significant with interpretive question, rather than responsive question. Hence, unlike with inference generated voluntarily in recall, students in Group E increased scores with interpretive questions, but not responsive questions, after training.
The answer to the third research question “What is students’ perception toward QtA lessons?” is that in general, participants thought they made some progress in reading comprehension. Raising interest and understanding questions and the content about the discussion were difficult for them. A majority of students in Group E favored more QtA training in the future, even though some of them were not aware of the power of QtA training.
Discussion
The present study shows the effects of QtA training on two measures, reader-based inferences and interpretive questions. However, measures of recall, text-based inferences, factual questions and responsive questions did not show an effect. The discussion will first present the reader-based inferences. Then the teacher-cued interpretive questions will be discussed later.
Reader-based Inferences
From the written recall, the result shows that readers with QtA training made more reader-based inferences. This finding presents the evidence that QtA training might enhance readers’ ability to make inferences based on personal knowledge. In doing the recalls, participants in Group E did not write down what they remembered in the text, but provided evidence of understanding and thinking, the integration of knowledge received and experience perceived. However, Group C participated in the traditional teaching method and did not generate this type of inference as much.
Although the number of participants demonstrated the difference was only few in number, the significant difference did show the potential of QtA in promoting reader-based inferences.
The efficacy of QtA training on reader-based inferences might be attributed to queries. Group E was trained to think and understand under a fixed set of Initiating Queries and a flexible variation of Follow-up Queries. These queries pushed Group E to read the lines, between the lines, and beyond the lines (Gray, 1960). The intensive training might help Group E form a habit to read and think. The finding was in line with Applegate et al. (2002) that readers should not be bound in low-level
thinking, but should try to pursue high-level thinking, like generating inferences that might draw a logical conclusion or responding to the text.
The result of this study is also in line with transactional model. As reviewed in Chapter Three, transactional model proposes readers take either efferent reading or aesthetic reading (Rosenblatt, 1993) to transact with the text encountered. It is speculated that QtA might promote linking of their personal experience to the text and aesthetic reading which resorted to individual understanding (Applegate et al., 2002).
Unlike Group E, Group C under instruction focusing on language might predominantly focus on efferent reading that they sought textual information rather than associated coming messages with endowed knowledge. It could be concluded that Group E were more likely to involve themselves in the reading.
Teacher-cued Interpretive Questions
The result of this study indicates that Group E was able to gain higher scores than Group C in answering interpretive questions after treatment. As mentioned in Chapter Three, Group E in this study read the text and discussed with three types of queries: the Initiating, Follow-up, and Narrative Queries. These queries prompted by the teacher-researcher led participants in the study to pay special attention to main idea, characterization or judgment of the story (Beck et al., 1996, 1997). Constant practice of answering these teacher-guided queries might enhance readers’ ability to construct meaning of the text information and interpret the text. The result of this study responds to Sandora et al.’s (1999) conclusion that distributed discussions focusing on meaning done collaboratively and guided by a teacher seemed more effective.
The finding of the current study supports Beck et al.’s proposal (1996, 1997) that QtA training might promote readers to construct their own meaning and present
their own ideas. In this study, the teacher-researcher used discussion moves to help students build meaning of the text and perform deep reading. For example,
“Marking” might successfully aid Group E in the study to notice key information;
“turning back to students” might completely invite them to present their understanding or to what extent they comprehend; “recapping” might effectively assist most participants in this study to grasp the gist of the specific segment. In the process of the discussion, Group E was trained to read globally for meaning, accompanying with “annotating” move to deal with problems in the passage. Hence, it could be speculated that the use of discussion moves has the effect of pushing students in this study to read in-depth. In this study, this type of discussion, like a shared think aloud (Kucan & Beck, 1997) might activate readers’ prior knowledge and present their unique interpretation, similar to generate reader-based inference.
Unlike in the traditional method, participants in this study were not provided with training regarding language aspects. It was likely that these students became aware of the need to free from the text elements, so that a meaning-focus that QtA training highlights drove Group E participants to capture the macrostructure of the text and create a situation model from it (Hill & Parry, 1994; Kintsch, 1998).
However, Group C in the study received traditional grammar-translation training did not emit power in answering interpretive questions. Focusing on the explanation of the text might not generate the thinking ability underneath the reader’s mind, nor did the participants in this group know how to apply their background knowledge with the text they encountered.
In sum, Group E, who received QtA training, did better than Group C, who obtained traditional training, in terms of interpretive questions. QtA training, therefore, might be taken as one effective strategy to generate more engaged readers to learn from text (Kintsch, 1998, p. 329).
Other Measures that Were not Affected by QtA Training
The following will explain possible reasons why QtA training did not manifest its power on these measures.
Text-based Inferences
Student-generated text-based inferences did not yield a positive effect from Group E in the present study. Based on van den Broek, Lorch Jr., Linderholm &
Gustafson (2001), text-based inferences are “repeated or paraphrased text statements”
(p. 1085). Readers generating this type of inference proceeded the text at a surface level, failing to read globally. Group E, unlike text-bound readers, unconsciously read above the surface level. The finding echoes van den Broek et al. (2001) that readers with an entertainment goal paid more attention to the link between text information and their individual perception in contrast to those with a study purpose focusing on coherent texts.
Another postulation on the result lies in that generating reader-based inferences demands more cognitive power. Trabasso and Suh (1993, as cited in Barry &
Lazarte, 1998) pointed out that readers’ recall of text decreased, for they exerted themselves to increase ideas “that moved away from the text” (p. 191). Thus, it might be because Group E’s cognitive power was used in generating reader-based inferences.
Responsive Questions
In general, the finding of this study accords with the premise that QtA training might ignite readers’ zest for reading (Rosenblatt, 1983), for their unique interpretation was valued in an equal manner. However, responsive questions
(teacher-cued reader-based inference) did not show an effect.
For responsive questions, the possible reason might be that inference generation is a private mental activity. Participants in the current study with QtA training would infer whenever they thought necessary. They might not manifest the ability in a presupposed direction of inference imposed by the teacher.
Factual Questions
In terms of the effect on factual questions, the finding accords with the above-mentioned result that the training did not increase the attention to details.
Group E in the present study did not benefit at retrieving the information from the passage. The queries teachers asked in QtA training might not promote answering factual questions. However, under traditional training, Group C who was supposed to have better control over details in the text did not show advantage, either.
Written Recall
Besides responsive questions, student-generated text-based inferences, and factual questions, QtA training did not enhance readers’ memory of the text or the ability to retrieve information from the text. Comprehension contains several facets, such as word-decoding, sentence-level, and situation level (Graessar et al., 2001).
As discussed in the previous section, this finding might broaden our vision to assume reading comprehension lies not in memory only. In other words, students that could tell every detail of the passage might not mean full understanding of the text. In this study, Group E did not produce more idea units that matched the texts, which differ from Sandora et al.’s (1999) findings that QtA training pushed students to make better and longer recalls. The contradictory result might be due to two reasons. First of all, Sandora et al. asked the students to do an oral recall while the present study used a
written recall. When conducting an oral recall, students tended to do less thinking and produced more text details. Writing Chinese characters is a more troublesome task, so participants in the present study might not be able to produce as many pausal units as oral ones, especially when there were interviewers in Sandora et al’s study who tried to elicit more response from students (1999).
Second, practicing QtA strategy, participants might require more time to process thinking. Time to write a recall in Chinese words might be constraining.
This could explain fewer amounts of propositions that Group E in this study produced after training because they needed more time to question the author than students in Group C.
To sum up, interpretive questions and reader-based inferences present positive effects that QtA training enhances power for linking of text and knowledge.
Therefore, QtA readers grow to be active readers with critical thinking rather than passive followers of others’ ideas.
Participants’ Perception of QtA Lessons
From the questionnaire, some phenomena have been found. First, generally speaking, participants had a positive perception toward QtA lessons even though some of them were not conscious of it. A majority of them stated that reading more stories in QtA training would enhance reading comprehension which supports Beck et al.’s proposal that QtA promotes reading comprehension (1996,1997). Secondly, the two major difficulties in the process are “raising interest in stories” and “understanding the questions and content for discussion.” For some Group E participants, like junior high students in Taiwan in general, English is still a demanding subject, no matter how easy the task or the content may be. Some of the participants stated that they disliked English learning and were not confident of themselves in English lessons.
The feeling of low self-esteem might cause these participants to be uninterested in the stories and to avoid the content and questions in the discussion.
For future implementation, a majority of Group E favored QtA lessons, for they could interact and discuss in interesting and relaxing classrooms. This shows that building a lively classroom might facilitate students’ learning. In fact, participants in the present study thought that they could read more articles, understand the content through discussing with classmates and have opportunities to answer the questions.
Also, participants found they spoke more in QtA lessons, which accorded with Beck et al. (1996, 1997, 2004) that QtA lessons elicited more student talk. Hence, it might be concluded that QtA training is a feasible teaching strategy in an EFL context, like Taiwan.
Pedagogical Implications
Several implications are drawn based on the findings of the present study.
First of all, the results of the study indicate that using a different mode, Questioning the Author, to manage a reading class is a good alternative. In traditional English classrooms in Taiwan, teachers dominate most of the talking owing to a tight schedule.
The QtA lesson, a more interaction-oriented approach, can release the domination of the teacher to more student talks. It is suggested that English teachers in Taiwan can create a dynamic classroom via questioning and discussing. EFL readers in Taiwan can be offered room to experience the product of their transaction with writers (Rosenblatt, 1993), create a situation mo del for text (Kintsch, 1998), do critical thinking and expand the schema. In this way, stressed-out English teachers in Taiwan may refresh themselves with their students’ interesting ideas while worn-out EFL students in Taiwan may voice out their thoughts instead of listening to monotonous translation and lectures in reading lessons.
Second, the finding of this study suggests that reading lessons in Taiwan should include different levels of comprehension. As the result reveals, combining the content knowledge with students’ personal experiences may facilitate junior high readers not only to read the text but also to construct their own text and begin to make reader-based inference. They may learn to transact with the author (Rosenblatt, 1993) and come to their own conclusion. Thus, it might be expected that junior high students transform from text readers to independent comprehenders.
Thirdly, in terms of the effect of QtA lesson on students’ improvement in answering interpretive questions, the idea which English classes need to include a meaning-based lesson is suggested. It is shown that language based training only promote less of ability to interpret the text. Hence, the value of such approach might be limited. To promote reader interpretation in the reading process, more innovative approaches are called for.
Moreover, the result of this study indicates that QtA training might be a feasible approach to be integrated either as an activity in a regular class or an extracurricular activity in our local situation. In regular class time, it is suggested that high school English teachers as well as elementary school teachers adopt QtA strategy in reading lessons. High school teachers might incorporate the QtA approach with six emerging topics issued in Ni ne-year Integrated Curriculum. The teachers can introduce passages whose topics include gender, the environment, information, home economics, human right, and lifelong development (MOE, 2003) and conduct a discussion that guides the teenagers to understand the social issues and make their own judgment. These students might be trained to reason from different perspectives. In addition to practice in a regular classroom, high school English teachers can apply the QtA approach in extracurricular activities, such as establishing an English club. The club may recruit EFL students with high motivation and
proficiency, so that they can read more difficult texts to question not only the authors but also the members.
As for elementary school context, the English teachers may elicit the students’
spontaneous response in the discussions and encourage active participation. In this study, over seventy percent of Group E students favored QtA training even though some of them did not exhibit huge interest in English learning or perform well on the recall/questions. Their affirmative attitude supports this approach regardless of their poor performance. If younger students can also take this lesson and foster interest in English learning, the issue of the polarization of junior high school students’ English proficiency might be improved.
Furthermore, in terms of teacher, in this study, the teacher released more responsibility to the EFL readers gradually as these readers initiate more talk in QtA lessons. The discussion context in the QtA lesson may help redefine a teacher’s role as Pearson and Fielding (1996) has suggested. The teacher in the present study made efforts to plan the lessons in advance to make sure the segments appropriate, the queries sufficient and the discussion moves dexterous. However, it is not easy.
Therefore, it is suggested that teachers collaborate to work on the lesson plans. The redistribution of EFL teachers’ workload might make teaching flexible and leave room for the teachers to pursue professional development (McKeown & Beck, 2004).
As this study reveals, when English teachers in Taiwan bear less responsibility for students’ learning in QtA lessons, they may conserve more energy to design creative lessons or do action research.
Last, in terms of student, based on the finding of this study, what resides in the students’ heads can be activated with a strategic key. The QtA lesson may be the key for students in Taiwan to explore possible interpretations when reading, to construct their own meaning of the text and to become independent thinkers. The finding of
this study can help fulfill one of the ten goals highlighted in Nine-Year Integrated Curriculum Guidelines, “independent thinking and problem solving” (MOE, 2003, p.
130). Nevertheless, the training effect can not be achieved in a day. Some of the low achievers in this study had hard time catching the essence of the discussions or failed to participate in the collaboration of meaning construction. But they still expressed that they started thinking and they found English reading more interesting.
The QtA lesson seems to serve as a pressure-proof context where EFL students in Taiwan, especially those with below-average ability, can regain their confidence in English learning.
Implications for Future Study
This study offers some positive evidences on the efficacy of Questioning the
Author training. However, the results should be elucidated cautiously as some limitations can not be neglected and need to be aware for future research.
On Procedure
First, due to the constraint of time, the effect of the training in this study was examined in the immediate posttest. We do not know whether the effect of reader-based inference and teacher-cued interpretive questions would sustain over time in a delay test. A second or a third posttest is suggested two weeks later or a month later in order to test how much the participants in the study still remember regarding the posttest passage. As for short-answer questions, the procedure can be the same as the posttest process. The results might be compared to find if there are differences between their responses to the same questions in the first posttest and those in the delayed posttest.
Second, the think aloud protocol, a widely-used procedure to investigate the
readers’ mental functioning, can be used for selected students to see the effect. How participants from Group E and Group C manipulate the flow of their thoughts when encountering reading might still be a puzzle. It is suggested that an application of think aloud protocol (Kucan & Beck, 1997) of some student representatives from both groups can be made to trace the different mental processes between the QtA training and a traditional training. Thus, the effect of individuals applying QtA can, therefore, be revealed.
On Participants Selection
The findings of this study were drawn from the researcher’s students.
Although the teacher-researcher did not inform these students of the experiment, the subjective bias can not be ignored. As a result, classroom observation (Beck et al., 1996,1997; McKeown & Beck, 2004) of QtA training and traditional training done by some objective observers is suggested. The observers can categorize classroom discourse, such as analysis of students’ responses or turn-taking between the teacher and the students, to extract rich data and provide insightful suggestions for teacher-researchers to improve education environment.
Participants in this study included only two classes of junior high school students in Taipei City. The result provided by a small number of participants might not be generalized to students of different levels in Taiwan. Also, the fact that the ninth graders in the study were facing the coming BC test might not yield their best performance. If future research can recruit students of other grades, the number of inattentive participants might decrease. Thus, more students may benefit from the QtA approach. What is more, we do not know whether the effect of QtA training might be different if participants are compared based on gender or proficiency level.
Future research might incorporate the gender or proficiency factor to explore the
effect of the QtA training.
On Materials
Since proficiency level is not the main concern in this study, all participants read the same materials in the pretest and the posttest. In future study, it is suggested that two or three sets of test materials may be given for high achievers and low achievers or mid achievers respectively. Thus, different findings might be yielded.
The present study adopts fables as the main genre. One or two Group E participants reported in the questionnaire that they did read the treatment stories in Chinese language before. In order to suit students of different needs, an inclusion of materials from various sources, especially expository texts (Beck et al., 1997), is recommended. If narrative and expository texts are involved, whether participants take the efferent reading (Rosenblatt, 1983) or whether they read for a study purpose (van den Broek et al., 2001) may yield different results to look into. In addition to types of texts, passages from diverse authors (Sandora et al., 1999) or origins may be considered as treatment or test materials. Bringing works of different authors from mixed origins might help EFL readers in Taiwan develop multicultural perspectives.
Duration of the Study
Insufficient time for QtA training might be another puzzle that QtA failed to show the obvious result. In Sandora et al. (1999), the period of treatment lasted for four weeks while the intervention in the present study endured five weeks. One more week in the present study did not guarantee the improvement of the recall in Group E. As there are more complicated factors interweaving in the study of second language, like memory problems, or participants’ inattentive behavior (Frantzen, 2003), the increase might not be so salient within such a short time training. Longer
treatment time might yield more promising results. This speculation echoes Huang’s study (2004) that most participants were used to traditional English classroom that they had for three and a half years at least. It would be hard to declare huge change in their performance within five weeks, comparing the learning habit kept for such a long time. Chang (2002) even mentions a conventional belief that “acquisition of new strategies requires persistent practice over time ” (p. 9). For this reason, QtA training might be practiced longer to put out the full powers over the contribution to greater improvement in reading comprehension.
Conclusion
This study provides some evidence that this instructional approach, QtA, may promote higher level reading comprehension. The QtA lesson, though failing to enhance readers’ memory of the text and the performance of answering factual questions, successfully promotes EFL readers to be more independent thinkers or even to ask questions themselves. They can infer based on personal knowledge or interpret the text and ma ke critical judgment. Instead of accepting whatever they read, they learn to integrate and reform their thinking.
The positive effects that QtA training achieves in this study might help consolidate its future implementation in Taiwan. In this preliminary study with an attempt to reveal the relationship between QtA training and inference generation, the conclusive findings can be a touchstone with hope that more researchers with interest will join the field. Therefore, guidelines for enhancing reading comprehension regarding Questioning the Author have yet been provided and waited to be explored.
