In the present study interviewing was chosen as the means of collecting data. Two kinds of techniques were employed. Semi-structured interview was used to elicit the student teachers’ perceptions, while stimulated recall interview was employed in order to find out what kind of motives there are behind the actual language choices. The process of data collection was set to motion by recording lessons held by the
participants. Approximately a week after the lessons of the given participant had been recorded, an interview session was arranged. The session started off with structured interview and proceeded to stimulated recall interview. The nature of semi-structured interview will be described and employing this interview technique for the purposes of the present study will be discussed in what follows. After this a similar description of stimulated recall interview will be provided.
Semi-structured interview lies in the middle of the structured-unstructured continuum of interview techniques (Borg 2006:189-190). As Fontana and Frey (2000:647) point out, in a structured interview all participants are asked the same questions in the same order. In addition, the answer alternatives are said to be predetermined. According to Borg (2006:203), semi-structured interview differs from structured interview in that it is based on “a set of topics or a loosely defined series of questions”. Borg adds that in a semi-structured interview the participants have the opportunity to talk freely of the topic at hand. In other words, when compared to structured interview, semi-structured interview is less strict in nature and leaves room for variation.
In the interviews made for the present study the participants could answer how they saw fit and their responses were not much controlled by the researcher. The unrestricted nature of the interview could be seen also in other points. The order of interview questions could be altered according to the order in which the participants chose to address the issues to be discussed. Clarifying questions were asked when needed. The questions used in the semi-structured interviews concerned the ideal distribution of the two languages and the uses, benefits and disadvantages related to Finnish (the L1) and English. The participants were also asked to describe an unsuccessful and a successful language choice. In addition, they were asked whether or not they plan their language choices and if yes, on what basis they decide on the language in the planning stage. Furthermore, the participants were asked what kind of instructions, if any, they had received from the teacher trainers regarding the language choice. The interview questions used in the semi-structured interviews can be found as an appendix in the end of the study.
While the first part of the interview was semi-structured and focused mainly on the participants’ perceptions, the latter part of the interview consisted of stimulated recall. This interview type was chosen in order to find out what kind of motives the participants find for their own language choices. The reason for employing stimulated recall was that it could elicit information about the participants’ language use which
could not have been attained in any other way. Some motives behind language choices cannot be identified through observation; instead it may be that sometimes only the participant can tell what exactly caused or affected a certain language choice and what its purpose was. For instance, Macaro (2000) and de la Campa and Nassaji (2009) chose stimulated recall for studying the perceptions and attitudes of their participants. In using stimulated recall technique there is the underlying idea that teachers’ actions are influenced by the way they perceive language and language learning (Nunan 1989:66).
In a stimulated recall session a stimulus is used “to elicit verbal commentaries about the cognitions (typically thoughts or decision-making) ocurring during previously performed behaviours” (Borg 2010:209). In practice this means that the participant watches video recordings or listens to audio recordings of his/her teaching and describes his/her actions giving reasons for them (Nunan 1989:69).
In the present study video-recorded lessons were used as the stimulus and the participants were to comment on their language choices. In order to obtain the material to be used as the stimulus two 45-minute lessons were recorded per each participant. It was important that the recordings would exhibit natural behavior. For this reason the influence which the researcher might have on the student teachers’ actions was kept to minimum. Prior to the recordings only the necessary information about the study was given to the participants. Thus the language use of the participants was not affected by the objectives of the present study.
After recording the lessons, the researcher watched the recordings selecting clips to be shown to the participants. The criterion for choosing a clip was that it had to exhibit a language choice of some kind. This included switching from one language to another within or between speech situations as well as choosing the base language for a certain speech situation (a case in which no switch occurs). Language choices in cases where the teacher read aloud written linguistic examples or came up with an example sentence for the sake of exemplifying the targeted language item were not viewed as relevant. This is because no actual choosing was involved in the forming of the utterance: linguistic examples are bound to be in the target language. Neither was translating linguistic examples from one language to another taken into consideration.
However, choosing the language for discussing the linguistic examples was seen as an actual language choice. Morover, translating one’s own speech to the pupils was seen as a language choice. An example of such a language choice would be giving instructions in one language and repeating them in the other. The point in selecting the video clips was to locate all language choices made by the participants. In some cases this was
problematic since the language choices occurred with such short intervals that it was impossible to capture only one language choice into one video clip. In such instances the researcher selected a clip containing several language choices, which the participant could then comment on as a whole.
In the stimulate recall session the researcher showed the video clips to the participant. After each video clip the researcher asked why the given language choice had been made. In addition, it was suggested that the participants would evaluate the successfulness of their choices. However, the evaluative comments were not included in the final analysis since the researcher decided to concentrate solely on the motives found for the language choices.
Each student teacher participated in one interview session, during which both the semi-structured and the stimulated recall interview were conducted. The duration of the sessions varied to some extent. The duration depended on how lengthy or concise replies the participant produced. Another factor affecting the duration was the number of clips shown in the stimulated recall interview. This, on the other hand, depended on the number of distinct language choices made by the participant during the recorded lessons. After the interviews had been conducted, they were transcribed focusing mainly on the content and wording. Details of interaction such as laughs, pauses or intonation were generally not taken into account since it was not necessary for the purposes of the present study. One exception to this was made in a case where the participant’s tone of voice was crucial in the interpretation of the utterance. Transcribing the interviews was followed by the analysis of the gathered data. The analysis method chosen for the present study and the steps taken in the analysis process are introduced in the following section.