Data Collection

The chosen data collection methods for the present study were video observation and an interview. Two kinds of methods were used to collect the

data to make the study valid. In other words, the choice to include two different means to gather information ensured that the study “does what it is intended to do” (Carmise and Zeller, 1979: 12). Furthermore, it can be generalized that when studying people’s intentions to behave in a certain way the best data collection method is to ask questions, but when the focus is on interaction behavior data collection should be done through observations (Jyrink 1977, as quoted in Tuomi and Sarajärvi 2009: 71). As the present study explores both of the above-mentioned aspects it was necessary to combine the two methods which also help to validate the analysis of the present study.

Four 45-minute lessons with the same class and teacher were observed and video recorded and while observing the teacher’s behavior and language use field notes were taken, reporting on instances of code-switching, activity types, general language use and behavior. The data was collected from EFL lessons which followed the English subject syllabus and the National Curriculum for Finnish General Education. The observed lessons took place among fourth graders who had been studying English for three semesters prior to the start of the data collection. The lesson observations were divided into two separate occasions, the beginning of the semester and the end of the semester. I observed two 45-minute lessons in January 2013 and after four months, in May 2013, another two 45-minute lessons were observed. The lessons were recorded with two video cameras in order to secure that all talk would be audible when returning to the data. Video observation was chosen as a data collection method due to its many advantages, such as, easy access to natural environment and interaction, access to rapidly changing situations, re-accessibility, accurateness, ability to observe non-verbal characteristics and material activity, and ease to identify speakers (Jordan and Henderson 1995: 50-53).

In addition to video observation, the teacher was interviewed after the first two lessons in order to get an understanding of her own perception of her language use. That is, the interview enabled the teacher to explain her language choices in detail and talk about her pedagogical views. The interview was conducted on

the same day with the second lesson observation at the school after the teacher’s working hours. The interview lasted approximately 45 minutes.

A semi-structured interview was chosen in order to get multifaceted data, since it allows flexibility during the situation (see Appendix 1 for the interview outline). In addition, the decision to conduct a semi-structured interview made it possible to ask clarifying questions or change the order of the questions during the interview (Tuomi and Sarajärvi 2009: 73). Furthermore, a semi-structured interview made it possible to correct misunderstandings. Finnish was used as a language of communication during the interview in order to get a maximal amount of information and create a conversational atmosphere, since Finnish was the mother tongue of both the teacher and the interviewer. The interview was also videotaped to secure that all the significant information would be easily accessed when returning to the data.

The teacher did not acquaint herself with the interview questions prior to the interview, although that could have increased the depth of her answers (Tuomi and Sarajärvi 2009: 73). I decided to keep the actual questions as a secret before the interview because otherwise they could have affected the teacher’s language use during the lessons. The teacher was, however, informed that the study focuses on her language use, but it was not specified in any other way before the first lesson observations. In other words, the teacher was aware that she was the subject of the study and was informed about the theme of the interview. As Tuomi and Sarajärvi (2009: 73) point out it is ethically justifiable and necessary to inform the interviewee about the topic which will be discussed.

As the study is longitudinal, I decided to conduct the interview after the first two observed lessons instead of after the second observation period. The decision made it possible to start the analysis right after the first part of the data collection, and thus, expedited the analytic process by four months. It is, however, possible that the fact that the teacher was more aware of the precise research topic had an influence on her language use and behavior during the

latter lesson observations. Thus, this was taken into account when analyzing the data.

Choosing the combination of observation and an interview as data collection method was thoroughly considered. Neither of the methods alone would have been sufficient in terms of the interests of the present study. On the one hand, observation is an effective means to gather information about people because they do not always behave as they might say (Johnson and Christensen 2008:

211). According to Johnson and Christensen (2008: 211) “an advantage of observation over self-report methods [e.g. interviews] is the researcher’s ability to record actual behavior rather than obtain reports of preferences or intended behavior”. On the other hand, it is impossible to know what a person is thinking, and thus, to determine why he or she behaves in a certain way. In the present study the observation was naturalistic. That is, the situation was not manipulated or interfered, apart from the presence of the recording devices and the researcher, in any way and the participants’ behavior seemed genuine and authentic. However, it has to be pointed out that the presence of the recording devices may affect the participants’ behavior, as Edwards and Westgate (1994:

77) explain the phenomenon called “observer’s paradox” takes place in all naturalistic observation. Indeed, during the lessons both the teacher and the pupils made a few comments about the presence of the video cameras. In order to find the reasons for the teacher’s behavior and language choice it was important to include an interview that was qualitative in nature. Patton (1987, as quoted by Johnson and Christensen 2008: 202) explains the benefits of qualitative interview data and supports the view that more understanding of people’s behavior can be found through interviews: “Qualitative interviewing allows a researcher to enter into the inner world of another person and to gain an understanding of that person’s perspective”. Thus, together the two complementary methods enable a more comprehensive reflection.