• No results found

Chapter 4 Data Analysis & Results

4.3 Qualitative Data Analysis

4.3.2 Phase One Qualitative Data Analysis

4.3.2.1 Phase One Challenges and Decisions

While doing the analysis, I faced many challenges which required me to take informed decisions as to how some of the responses may be classified. The boundaries in some of the students‟ responses were not clear enough, therefore, I had to decide on how to classify such units and justify each decision. Hence, each time I faced a response that did not lend itself easily to the coding schemes, or had any obscurity in it, I had to deal with it on the spot. The decision taken was written

on a post-it-note and put on an A4 blank sheet to keep all decisions visible in front of me for further reference as an attempt to achieve consistency.

Similar to Goh (1998), coding the data on metacognitive knowledge was one of the most challenging stages in my research. There were instances where items that had been placed within a particular group seemed to easily fit into another group. The criteria I used when taking decisions may appear to be subjective at times, particularly when the distinction between two sub-categories depended on my own interpretation of the participants‟ responses. However, as Goh (1998) explains, by applying the same criteria consistently throughout the analysis, I attempted to ensure that there was some uniformity in the way my interpretation of the listening diaries had been categorised (p. 167). Goh states that “categorisation of qualitative data is by nature a subjective process and researchers do not make claims for the objectivity and completeness generally associated with scientific inquiry” (ibid: 170). I will now present the challenges I faced and the decisions taken with some extracts from students‟ responses.

1. Linking two idea units in a cause and effect form, through the use of words such as so, that, which, etc. I decided that, though the response had two idea units, I would treat them as one. The reason for dealing with such a challenge this way is that since one happens as a result of another, and that the two are directly related, it would be more logical not to separate them. All instances of cause and effect responses were indications of person knowledge, because the students related the demands or nature of the task to themselves as learners. This can be seen in the following extracts:

Because the speaker was talking slowly so we had time to understand what he was saying

Also, he was talking slowly which gave me enough time to take notes

Informing us of the divisions of the lecture before giving the details helps us to concentrate more

The speaker was giving way too [many] information that I had a bit of difficulty [of] writing down my notes

2. The three sub-categories in Goh‟s coding scheme, problems during listening and obstacles to listening comprehension, both under person knowledge, and factors that affect listening comprehension, under task knowledge, all shared one common feature, that of “what the subjects thought were hindrances to their listening comprehension” (Goh, 1998, p. 166). Goh provides the following definitions for the two subcategories of problems and factors. Problems during listening, she says, “would refer to any difficulties the subjects had experienced and that related directly to one of the three cognitive phases of comprehension, namely perceptual processing, parsing and utilisation”. Factors that affect listening comprehension, on the other hand, “would refer to anything that the subjects perceived could either impede or enhance their comprehension” (ibid: 167). I found it inconsistent that Goh defines factors that affect listening comprehension to include those that could “impede” or “enhance” listening comprehension and then mentions only “unfamiliar vocabulary” to the exclusion of familiar vocabulary, which I believe is also a factor playing a significant role in enhancing listening comprehension. The challenge was that participants would many times refer to the same idea but use words like easy, difficult, familiar, new, clear, rather than unfamiliar vocabulary. Therefore, since all of these comments on vocabulary are, in my point of view, among the factors that affect listening comprehension, and are a comment on the familiarity, or not, of the vocabulary to the learners, hence I decided to treat them all under the same sub-category: (un)familiar vocabulary. The following are examples from students‟ responses:

Because it did not include any new words Clear vocabulary

There was not any hard words

Words used in the text are clear and understandable Without difficult vocabulary

3. The participants used many ways to comment on types of input, such as the

use of examples, being an easy topic, using an easy/ clear style, the lecture being well-organized, etc. Since these, in my point of view, all indicate the

same idea, I decided to treat them all as a comment on types of input. Examples from participants‟ responses are:

the topic was easy

the lecturer’s style was easy

the lecturer gave some examples to aid understanding the lecture was organized

4. I had a problem in classifying the response because the talk was very clear, since it did not seem to lend itself clearly to any of the sub-categories. I decided finally to group it under speech rate, as the sentence seemed to suggest reference to the speed of the speaker.

5. There was another comment that was not easily classified, which is because

of the speaker’s clear pronunciation, as there was no clear reference to

pronunciation in the coding scheme. However, since pronunciation falls under accent, I decided to group this response under the code: different varieties and local accents.

6. Another unclear response was it was clear, which occurred as an answer to why did you find the text easy? I found this somehow problematic, since I was unsure whether clear here means clear ideas or clear voice. However, since the sentence „the text was clear‟ is taken to mean clear ideas, rather than voice, hence I decided to group this response under types of input.

7. Some responses stated clarity of voice and tone as one of the reasons why the text was found easy, or difficult. This appeared at first to be one unit, but then I came to realize that clear voice was used by participants to refer to the quality of the recording. Hence, I decided to break this response down into two units, clear voice fell under physical factors, whereas clear tone under prosodic features. Both, however, fall under the category „factors that affect listening comprehension‟, which comes under task knowledge.

8. Another issue I faced was that there was no category in Goh‟s coding scheme that related the listening text to the required task, although this is a common factor which affects listening comprehension; difficulty of the task. In other words, many times have participants indicated that one of the things that has helped them to understand or to find the text easy was the simplicity of the

task or simple questions. This did not seem to belong anywhere in the coding

scheme! However, since task knowledge refers to the demands of a task, I decided to include it under task knowledge. Therefore, I decided to create a new category under factors that affect listening comprehension, with the code „difficulty of the task‟.

9. Comments on familiar vocabulary caused me some confusion as well since this element appeared under task knowledge and person knowledge in Goh‟s coding scheme. Hence, I needed to make the boundaries more vivid between these two codes. The decision was that if the participant attributed it to herself, e.g. saying some of the words I know, then it would be considered under person knowledge. On the other hand, if it was written as a general comment on vocabulary, then I considered it as task knowledge.

10. I had a major issue with classifying responses related to previous knowledge and having some background knowledge related to the topic of the listening text. Goh‟s coding scheme classifies it under task knowledge, as one of the factors that affect listening comprehension. Yet, from my point of view, it may be better grouped under person knowledge because it is a characteristic of the person, having information related to the text rather than the task itself. This leads to another relevant issue, which is distinguishing between common topics and having background knowledge related to the topic. Not every common topic do students have some information about, hence I decided to treat them as two separate units. The latter falls under person knowledge whereas the former belongs under task knowledge. Goh in fact in the original study (Goh, 1998) identifies “insufficient prior knowledge” as one cause of problems during listening, as part of her discussion on person knowledge of

her participants (p. 346). This again indicates how fuzzy the boundaries are between the different types of metacognitive knowledge.

These were the main challenges which occurred when analysing phase one diaries and decisions in regards to them had to be established before presenting the results of the analysis. The following part will present the results of the analysis and will be divided into three major sections according to the three probes in the guided listening diary: probes 2, 3, and 4. Since the first probe in the diaries was an either/or question: „did you find the task easy, difficult or neither of the two?‟, and this probe was directly related to the second probe, I decided to merge these two questions into one section and do the analysis for them simultaneously. The types of metacognitive knowledge elicited by each diary probe will be presented under each major section. This will first be summarized in table format and then major themes will be presented separately, accompanied by illustrative examples from the participants‟ diaries.