This chapter presents, discusses and exemplifies the research method used in the study. Research design, hypotheses, research focus, data collection procedures and the analytical framework are presented first. A preliminary empirical test of the validity of a key intuitive construct of this study, the cline of initiative, is then reported on. The final section discussed some problems identified in the application of the analytical framework.
A research design implies a careful plan, which a researcher makes at the beginning of a project to decide on an appropriate approach. In this section, the research design of this study will be described in terms of Seliger and Shohamy’s (1989) following parameters: hypothetico-deductive versus heuristic-inductive purpose; analytic versus synthetic-holistic approach; and qualitative, descriptive and quantitative designs. The present study can be broadly characterised as hypothetico-deductive, analytic and descriptive. These characteristics come into focus in the discussion that follows.
Hypothetico-deductive versus heuristic-inductive purposes
A hypothetico-deductive purpose begins with specific research questions or hypothesis, which narrow the focus of the research and enable the researcher to do a systematic investigation. In Seliger and Shohamy’s (1989) terms, most aspects of a deductive purpose are hypothesis-driven and have some degree of explicitness in data collection procedures typical of heuristic-inductive
approaches, which are more exploratory and may lead to the formulation of hypotheses. The present study can be characterised as hypothetico deductive, analytic and descriptive. It is hypothetico-deductive because it begins with four hypotheses (§ 1.3) that guide the researcher to focus on only certain aspects of the possible data on interaction in first and third-year tutorials.
Analytic versus synthetic approaches
An analytic approach implies that the phenomenon being investigated is analysed into its constituent parts. When this approach is taken, one constituent part or a cluster of the constituent parts may be examined in greater detail to the exclusion of other factors. Also, this approach implies that there is enough information about the constituent parts to be explored in isolation (Seliger and Shohamy 1989:56). A synthetic approach, on the other hand, implies that the researcher is aware of the interdependency of the parts of a phenomenon being investigated and will thus look at the separate parts as a coherent whole. The present study is essentially analytic in its approach as the focus is on investigating a number of specific features of student (and tutor) participation in tutorials: the total number of student discourse acts and turns, discourse act initiative, turn-taking initiative and the possible relationship between certain cohesion features and participation effectiveness, as revealed in the analysis of discourse acts and turns. Ultimately, however, these features are used to define a particular synthesis, namely ‘participation effectiveness’ in a context such as university tutorials.
Qualitative, descriptive and quantitative designs
Qualitative, descriptive and quantitative research designs are presented on a continuum in Seliger and Shohamy (1989), and can also be used in combination to achieve different purposes within a study. Both qualitative and descriptive designs are concerned with describing naturally occurring phenomena, without any experimental intervention, but a qualitative design differs from a descriptive design in that it is heuristic, that is very few decisions are made before the study begins. It is also hypothesis-generating research, while a descriptive design can be either heuristic or hypothetico-deductive (Seliger and Shohamy 1989:11). In addition, descriptive research can be either synthetic or analytic in its approach and it does not manipulate naturally occurring phenomena.
Contribution of the study
This section discusses the contribution of my study in terms of the theoretical-methodological, descriptive and applicational levels.
In this study with its central construct, ‘ participation effectiveness’, I sought to make a contribution to the field of discourse analysis at a theoretical-methodological level by developing an analytical framework based on ideas drawn from Crombie (1985a and 1985b), Hubbard (1998) and Van Lier (1988). This analytical framework combined six discourse acts (§ 3.5.2) and four turn-taking categories that measured students’ discourse acts and turn participation and their initiative at discourse act and turn taking levels. Apart from combining the two frameworks, this study also makes a contribution by improving on problematic original definitions of both discourse act and turn categories and turning them into more viable operational definitions, so ensuring that their application in analysis, though not unproblematic, is less of a high-inference procedure than is often the case in discourse and pragmatic studies (§ 3.5.2). This study also expands Crombie’s (1985a) elicit category into three question types, namely closed display questions, closed referential questions and open referential questions in order to explore effects of tutor discourse behaviour on student participation more closely. in discourse act participation. The female students also had a higher number of turns than the males and fewer non-initiative-bearing turns and the statistical test indicated a strong tendency towards a significant difference in favour of females. This finding contradicts earlier research which found that female students talked less both in frequency and duration than the male students (Brooks 1982; Coates and Cameroon 1988). In these studies, however, neither the gender parity issue nor performance per individual student were considered. This was also the case in De Klerk (1994, 1995a & 1995b) and Sternglanz and Lyberger-Ficek (1977), where the males participated more than the females. In Ricks and Pyke (1973 in Smith 1991:40), where males and females were equal in number, equal achievement and similar interaction patterns were reported. This suggests that unequal numbers can affect participation, supporting the findings of my study, where the importance of this factor has been pointed up through the provision of means-per person as well as group statistics. The findings with regard to the effects of tutor gender on students of different genders indicated that the females’ mean values for discourse acts in the female-led tutorials were four times higher than those of the females in the male-led tutorials. The males’ mean values also for discourse acts in the male-led tutorials, on the other hand, were higher than those of the males in the female-led tutorials. Also, in terms of turn participation, the male turns per student were higher than those of the females in the male-led tutorials, while in the female-led tutorials, the female turns per student were higher than those of the male students. This finding contradicts earlier findings by Boersma et al.
Discourse acts and functional units
The main focus of this study is the quantity and quality of students’ spoken discourse. Both aspects are related to the notion of ‘effectiveness’ in participation, which is operationalised in terms of the amount of participation generated by the students at discourse act and turn taking level, and the degree of initiative also at discourse act and turn taking level. In order to quantify and compare students’ discourse in and within turns, Crombie‘s (1985b:45) work on interactive semantic relations and general semantic relations forms a background to the analysis of relations within and between turns . The discussion below begins with the identification of textual-units, followed by the discourse acts and then the sub division of elicits. Crombie makes a distinction between interactive semantic relations, as in  and  (i.e. elicit, and reply-inform) and general semantic relations or discourse value relations (i.e. Reason-Result) between and also within turns. The discourse values are also called binary values because they require two linked components (Hubbard 1989:125), such as Reason-Result, as inT111  and ; and Condition-Consequence, as in T311 .
CHAPTER ONE: INTRODUCTION AND BACKGROUND
1.1 The research problem
1.4 Research design
1.5 Structure of the study
CHAPTER TWO: INTERACTION, INITIATIVE AND ACQUISITION
2.1 Interaction, input and output
2.2 Interaction in small group discussions and tutorials
2.3 Student and tutor gender as variables in discourse interaction
2.4 Tutor discourse behaviour
2.5 Conjunctive cohesion
CHAPTER THREE: RESEARCH DESIGN
3.1 Research design
3.3 Research focus
3.4 Data collection procedures
3.5 The analytical framework
3.6 The cline of initiative study
3.7 Comments on the analytical framework
CHAPTER FOUR: RESEARCH FINDINGS
4.1 Hypothesis 1: Year of Study Hypothesis
4.2 Hypothesis 2 : Student Gender Hypothesis
4.3 Hypohtesis 3: Tutor Genter Hypothesis
4.4 Hypothesis 4 : Tutor Discourse Behaviour Hypothesis
4.5 Cohesion analysis
4.6 Chapter review
CHAPTER FIVE: CONCLUSION
5.2 Synoptic review
5.3 Limitations of the study and suggestions forfurther research
GET THE COMPLETE PROJECT
AN APPLIED LINGUISTIC INVESTIGATION OF PATTERNS OF INTERACTION IN UNIVERSITY TUTORIALS