The major goal of this present study is to investigate how Taiwanese high school students conceptualize textual and visual connection in English learning texts. This chapter details the overall research design, including the participants, theoretical frameworks, instruments for assessing conceptions, and procedures.
Participants
Since it was hypothesized that the variations of conceptions mirrored EFL learners’ performance and competence, the participants were chosen from a language gifted (LG) class and a regular social science (RSS) class on voluntary basis to represent English proficiency differences. The participants from LG class had higher aptitude for language, and had higher language proficiency than those from any other class in the same grade. The major focus of the English class in this school was on developing reading abilities and expanding their English vocabulary. They had five periods of English classes in which they received intensive reading and writing training each week. They are also encouraged to express themselves in various ways such as oral presentations, Power Point Presentations, and dramas. The LG class had more opportunities for oral presentation in class and was given one more English class for training oral expressions. A total of sixteen EFL eleventh graders at a senior high school in central Taiwan were recruited to participate in the present study. Half of the participants were recruited from LG class while the other half of the
participants were from RSS class. The students in the LG class had an average of 65 in their monthly exams, which is considerably higher than those of the RSS class (with an average of 50). The scores of every monthly test of every participant were shown in Table 1. While the average years of English learning in RSS group was less
than 9 years, those of the LG group was 10.5 years. On average, all of the participants in the study had about ten years of English learning experience. They were informed of the research contexts where they needed to present their English reading habits and their views of reading in English textbooks in general during the semi-structured interviews. All of the participants were assigned a number for identification purposes.
Table 1 shows the academic backgrounds and the grouping of the participants.
Table 1. The Academic Backgrounds of the Participants
Group Name
* A sequential identifying number for each student
Note. RSS=the regular social science group; LG= the language gifted group
The present study involved two parts (see Appendix A):
PartⅠ: General questions regarding the participants’ English learning experiences and background
PartⅡ: The main part of the analysis. The participants in the study were asked to describe their English reading habits and experiences.
The guiding interview questions for assessing conceptions of textual and visual connection in English textbooks were modified from Tsai (2004, 2009) and Yang and Tsai (2010), presented in Appendix 1. Prior to the interviews of the research being conducted, many potential interview questions were examined, discussed with a professor, who has conducted research on conceptions, and tried out with five high school students in the pilot. To probe learners’ conceptions of textual and visual relations in English textbooks, the interview started with questions about the
participants’ academic backgrounds and EFL learning experiences in general. Then, the participants were asked to describe more specifically about their reading focus in a lesson of an English textbook, the reading approach they adopted to interpret
multimodal texts in a lesson, and how they utilized textual and visual relations to achieve EFL reading comprehension. They needed to answer questions related to the features they noticed or were attracted to in the English textbook. They were asked to describe how those features in a lesson in English textbooks influenced their reading comprehension. On average, all the interviews lasted for thirty to forty minutes.
In the pilot study, some of the students were inarticulate in their responses to the interview questions. After the pilot study, some of the interviewed questions were revised to cover all relevant aspects and gain better understandings of the participants’
conceptions. For example, “What do you think about the lessons in the textbooks?”
was revised into “How do you describe the lessons that impress you most in English textbooks?” The revised questions were designed to encourage the participants to be engaged in thinking in order to answer the target questions clearly. All the interviews
were audio-recorded and conducted in Mandarin. The recorded contents were transcribed for further analysis.
Phenomenographic Method
Phenomenographic method was employed in the present study for three reasons.
First, the method maps the qualitatively different ways of understanding. It explores various aspects of what people experience and how people conceptualize things.
Second, since categories in hierarchical order often emerge, this method provides a chance to differentiated learners’ views and consequently to refine and change learners’ existing views on things. Third, the approach was used to distinguish differences in people’s experiences of a phenomenon or the ways people understand the world around them (Marton, 1986).
Data Analysis
The interviews were audio-recorded and analyzed through phenomenographic approach (Marton, 1986; Richardson, 1999). The approach is iterative in nature and can be used to describe the phenomenon involving a small sample of subjects (Richardson, 1999). In phenomenography analysis, researchers observe the most distinctive patterns found in data that distinguish conceptions qualitatively (Benson &
Lor, 1999). Patterns that strike structurally significant differences are useful to understand how people define and construct a certain part of the world (Benson &
Lor, 1999; Marton, 1986). Phenomenographic analysis focuses on how a subject matter appears to people and involves sorting and categorization of similar ideas into hierarchically ordered categories of conceptions and reflects qualitatively different ways of understanding a phenomenon (Brew, 2001; Richardson, 1999). Because the intention of phenomenography is to describe different ways a phenomenon is viewed,
it is particularly suited to an exploration of the variations in the ways high school students conceptualize the visual-textual connection in textbooks. The results gathered through phenomenographic method could demonstrate a hierarchical relationship in which some categories are less complex or sophisticated than others (Brew, 2001).
The participants’ conceptions of textual and visual relations in English learning textbooks were assessed and investigated on the basis of phenomenography analysis.
The verbatim transcripts are the major data of analyzing the participants’
conceptions. First, the researcher sought for relevant answers to the interview questions including words, sentences and turns of conversation sections. The researcher distinguished the relevant conceptions from the irrelevant ones by
bracketing the responses not related to the interview questions, as shown as follows:
T: What draws your attention when you open the English textbook?
Participant 1: I saw pictures.
T: Why?
Participant 1: Because I thought I don’t understand the words. The pictures are more appealing.
T: The pictures draw your attention. What does this have to do with your reading comprehension?
Participant 1: Pictures makes me understand more about connections between texts and pictures.
T: What kind of connections are you talking about?
Participant 1: Take this paragraph, for example, you may not make meanings from the textual descriptions, but the picture tells you the percentage of the things.
T: So you can understand the concept of percentage from the picture. How do you usually read a lesson?
[Participant 1: I don’t understand meanings of a few words.
T: What words?
Participant 1: The words I haven’t learned. ]
Second, the explanations were then labeled as ideas. An idea extracted from the responses was counted in terms of consistent and structural meanings (Eklund-Myrskog, 1998). An idea is counted when it is relevant to the interview questions, unique, and non-repetitive. In other words, an idea could be a phrase, a sentence, or the main idea of several sentences. Each idea showed a unique way of viewing the phenomenon being studied. The number of relevant explanations to the interview questions was calculated. At this stage, the data yielded 86 ideas in total. The following instance demonstrates how an idea is identified:
T: What does the picture tell you?
Participant 2: Playing.
T: What do you mean by playing?
Participant 2: I think it is about traveling.
T: And is the lesson about traveling?
Participant 2: Well. Not exactly. The picture tells you part of the story. The main idea of the lesson is jobs. Idea 1
Then, the researcher extracted and coded the most salient ideas related to textual and visual connection. The coded ideas exemplified the participants’ views of visual-textual connection in English textbooks. Next, the researcher assigned one label that characterized the participants’ views of the present topic and sorted them into groups according to their qualitatively different representations of conceptions. The following shows how a code is assigned to an idea:
T: How do you make sure you have learned something from a lesson?
Participant 2: Whenever I pass the test.
T: Which lesson impresses you most?
Participant 2: This one. Twilight. Because of the movie. Also, the pictures in the lessons are interesting. Idea 1: engaging
T: How do you take advantage of the pictures?
Participant 2: I have no ideas.
Third, similar ideas were clustered and assigned to a more representative
category and then were ordered hierarchically. Through comparing the similarities and differences among data, a more representative category of conception emerged.
The categories were examined again to make sure they adequately represented the depth and breadth of the participants’ views. Two raters of conceptions were involved in the present study. The first rater is the researcher of the present study. The second rater is the adviser of the present study researcher. All of the research data were sorted through rounds of discussions before any attempt to delineate categories of
conceptions was made. The two raters independently conducted segmenting and coding of the sample transcripts from one fifth of the total participants based on the agreed criteria. Inter-rater reliability, the level of agreement between raters, was counted and the agreement between two raters is 86%. Reliability was calculated using the formula: 2M/(N1+N2), (Holsti, 1969) and M represents the number of agreed coding by the two raters while N1 and N2 refer to the numbers of coding done by the two raters respectively. Disagreement and differences of coding and grouping were discussed and reconciled through further discussion.
Then, the rest of the ideas was coded. The results of the coding were finalized after a few rounds of discussions with the second rater when necessary in order to stabilize the coding and categorization phase. Broad clusters of similar ideas were
combined and were assigned a label with broader, more representative meaning until the differences between categories were clear. By doing so, the emerging conceptions became as qualitatively separate as possible. The sorting and grouping continued iteratively for several rounds until the stabilized categorization was reached. The categories describing the participants’ conceptions were recursively checked and refined.
Disqualified Ideas
Apart from the categorized conceptions above, 13 out of 86 ideas were excluded from the five categories. The excluded ideas lacked detailed explanations or contexts and thus were unable to be categorized. The participants plainly described their ways of reading without giving needed explanations or contexts. Take the idea of
participant 13 for example,
#13: I usually ignore the pictures.
She stated that she usually ignored the pictures but did not provide follow-up explanations, which made it difficult to be coded. Another example of disqualified ideas is listed as follows:
#1: I learn the lesson without comprehending the contents. I simply memorize it.
She stated that she learned something from the lesson without processing the reading and that she mechanically memorized the reading. The idea has little
relevance to textual and visual relations and thus cannot be coded. The previous two ideas are examples of disqualified ideas. After excluding the disqualified ideas, the present study identified 73 out of 86 textual and visual related ideas for further calculation.
Calculation of Ideas
Finally, after categorization was done and disqualified ideas were eliminated, the frequencies of conceptions were listed by calculating the respective number of
conceptions in each category. Since the present study investigated the students’
conceptions of textual and visual connection in English textbooks instead of classifying each student, the students were allowed to have one or more than one conceptions belonging to different categories. When one participant held more than one view that belonged to different subcategories of the same conception, the views were also counted based on the subcategories their views fell into. Moreover, two groups’ distribution of conceptions was also illustrated by calculating their respective proportions of conceptions. In other words, when a student’s ideas fell into three subcategories of the category, the three conceptions were all counted. For example, the subcategories in “Read bi-modally to comprehend” include “Understanding the meanings”, “Guessing the meanings”, and “Anticipating contents”. The three
subcategories share high degree of resemblance. Student number 10 contributed three ideas listed below:
#10: I don't like reading words. Reading the pictures make me understand faster. (Understanding the meanings)
#10: After finishing reading the pictures, I will take a look at the content if I still find it difficult to take a guess from the pictures. (Guessing the
meanings)
#10: I usually look at the pictures before I read…The pictures may have something to do with the title. (Anticipating contents)
The first idea belongs to ‘Understanding the meanings’, the second one belongs to ‘Guessing the meanings’ and the last one falls into ‘Anticipating contents’
respectively. Thus, three of the ideas were all counted for further analysis.
However, the same student’s ideas belonging to the same subcategories were counted as one idea. For instance, student number 1 stated:
#1: I know the meanings of the words if I know the words before…pictures make you understand the meanings of words in the very beginning.
(Understanding the meanings)
#1: The pictures are helpful in understanding the content of the lesson. ....In the beginning, pictures make you understand the meanings of words.
(Understanding the meanings)
These two ideas belong to the same subcategory, and thus were counted as only one idea to avoid repeated counting.
After excluding the overlapped ideas in the same subcategories, 57 out of 73 ideas were left for calculating frequencies and plotting distribution of conceptions.
CHAPTER FOUR