• No results found

Chapter 3: Methodology

3.4 Research Design

3.5.1 Pre-test, post-test and delayed post-tests

A non-equivalent groups design was used, with a pre-test (week 1), post-test (week 10) and delayed post-test (week 13) structure. Non-equivalent group designs are structured like a pre-test, post-test experiment but lack the key feature of random assignment from the general population as a whole. Instead two groups are used which are thought to be comparable. In this case, as outlined in section 3.3, the subjects were students from two year nine classes at a Bruneian girls’ secondary school who were randomly assigned to either a group receiving focused written CF, or unfocused written CF.

I designed these three tests, each of which consisted of a series of six pictures and key words. Three tests were used in order to investigate any possible differential effects following either written focused or unfocused CF over an extended period of time. The pre-test enabled base-line data to be collected, while the post-test allowed for a

comparison of written accuracy scores immediately following the eight weeks of written CF. A delayed post-test was used in order to examine any delayed effects of the written CF. Each of these tests used a theme related to a common event in the students’ lives (Appendix 1). The pre-test showed pictures of a student getting ready for the first day of school, the post-test showed pictures of an injury and a doctor’s office, and the delayed post-test included pictures of going on a vacation. These themes, which were familiar to students, were selected in order to increase construct validity. As the classroom teacher

51 of the participants previously, I had observed these themes being repeated in student textbooks. Therefore, if an individual student had no personal experience of travelling away for a vacation, they did have experience of reading and writing about vacations. Each test was structured and organised in the same way, and the tasks were the same. This meant that I was able to establish some level of comparability across the three tests.

Pictures were used as story guides with only a few key words so that all students regardless of their individual English language ability could access and understand the theme. The six pictures were collected from a clip art site and were selected for

relevance to the theme of each test. Students were also told that they were to use the picture stories as a guide, and that they could write an imaginative narrative. This enabled all students to write a short composition even if they lacked the vocabulary knowledge to write about the specific event depicted in the picture story. For example, if students did not know the vocabulary associated with playing basketball as is depicted in the post-test, they were still able to write a narrative about the same series of events for a sport which they were more familiar with. The tests were not designed to assess

students’ particular thematic vocabulary. Rather, the objective of the tests was to

generate narrative writing in the past tense which enabled an examination of how the

students used the five targeted grammatical features (copula ‘be, regular past tense,

irregular past tense, articles and prepositions). This allowed for a degree of

comparability between the three tests because students were able to write a narrative composition of about half a page even if they did not have the vocabulary knowledge related to a particular theme of a test. The format, instructions and conditions of each

52 test were the same, and the desired outcome of each was the production of a past tense narrative composition. The use of regular and irregular past tense forms is an area which I have observed Bruneian students having difficulty with previously, and it is also an area which Sheen et al (2009) investigated with regards to the effect of written CF. In order to build on knowledge from that study, and investigate the effect of written CF on particular linguistic features which Bruneian students have difficulty with, the writing tasks used in the current study were designed to generate writing in the past tense. Moreover, students are used to similar activities in tests, so the process of writing a past tense narrative story based on a series of pictures is familiar to them and did not require teaching.

For each of the three tests the students were given the pictures, and were encouraged to discuss them with each other in pairs for five minutes. They were also asked by the researcher if they had any questions about the pictures, and if required clarification was given. A few students used this time to ask for a particular vocabulary word, but all were familiar with the process of writing a narrative based on pictures from school examinations and the task itself did not require further clarification. As the tests were not concerned with students’ vocabulary knowledge, if a student asked for a particular word prior to the test they were provided with it. Students were then told they had 30 minutes to write their own narrative story of approximately two paragraphs.