The characteristics of the study will be presented by giving reasons for the instruments chosen to answer the research questions. It was recognised in chapter two that a gap existed in the literature and consequently little research had been undertaken into life events and decision making for mature consumers. Therefore it was helpful to begin bridging the research gap by using a methodological approach which also included qualitative data gathering. This ensured research topics were adequately identified before creating operational items in a quantitative questionnaire. The hypothetico-deductive methodological approach has qualitative and quantitative phases with several feedback loops designed to fulfil these requirements.
Qualitative research may be used for preliminary exploratory work before mounting a larger scale or more complex study. It is frequently used as a purely technical first step in the design of structured interview surveys (Hakim, 1992). This kind of strategy is important to studies where the theories underpinning the research are under-explored and little research is available on the topic (Creswell & Plano Clark, 2007). Existing instruments require adaptation for the purpose of this research as this type of procedure is the choice for survey development (Churchill Jr., 1979; Creswell, 2009). Instruments in the area of life events and consumer behaviour such as the questionnaires used by Lee (1995), Lee, Moschis and Mathur (2001), Mathur, Moschis and Lee (2003) and Mathur, Lee and Moschis (2006), were insufficient for this purpose. Whilst they study life events and consumer behaviour they do not delve into the consumer decision making process nor financial services. Their focus is on antecedents for changes in brand and patronage preferences.
The emphasis of the study is on the confirmatory quantitative phase. The qualitative phase was able to elicit in depth responses from participants. However the most prominent findings would not be generalisable without a quantitative confirmatory
phase. Many varieties of research techniques such as experimental and observational were considered and discarded for this research project. They are discussed later in the chapter. For the main data collection a mail survey was considered the optimal method to satisfy the pragmatic constraints and research needs of the project. The hypothetico-deductive methodological approach of this study can be considered a descriptive, pragmatic, sequential-exploratory approach. This method uses qualitative data collection techniques to create a survey to create measurable constructs. The constructs are then confirmed using quantitative data collection that provides information to further theorise and examine relationships (Creswell, 2009; Creswell & Plano Clark, 2007; Teddlie & Tashakkori, 2003).
Descriptive: Descriptive studies are those where the characteristics of a study are known to exist and the researcher wants to better explain the elements under investigation (Sekaran, 2003). The relationship between decision making variables is known (Brennan & Marriott, 1996; McCarthy, et al., 1997; Neal, et al., 2006), as is the relationship between life events and changes in consumer behaviour (Andreasen, 1984;
Lee, 1995; Lee, et al., 2001; Mathur, Lee, et al., 2006; Mathur, et al., 2003; G. Moschis, D. Bellenger, & C. Curasi, 2003). This study describes relationships between life events and consumer behaviour within one framework. It also illustrates how they interrelate for different groups of mature consumers within the population.
Pragmatic: In addition, the study underlying this thesis can be labelled as a pragmatic approach. The broad research methodology is pragmatic because it draws upon any approach necessary to provide a workable solution to the research problems (Creswell, 2009; Creswell & Plano Clark, 2007; Teddlie & Tashakkori, 2003). For example an initial qualitative data collection was undertaken to establish the relevant concepts and themes for financial services decision making. Financial services decision making models are largely lacking in the research (Harrison, et al., 2006). Similarities observed
from the semi-structured in-depth interviews lead to the utilisation of generic decision making models to provide a basis to frame the research. The pragmatic approach is problem centred and draws on both qualitative and quantitative methods to solve research problems (Creswell, 2009). Since research had not been approached in this way before it was important for the study to have a pragmatic approach.
exploratory: This study used a sequential-exploratory approach. Sequential-exploratory studies firstly involve a phase of qualitative data collection. A second phase of survey design and a third phase of quantitative data collection and analysis follow to confirm and build upon the first phase (Creswell, 2009; Creswell & Plano Clark, 2007).
The sequential-exploratory process was preferred to the simultaneous data collection as presented by other approaches such as Concurrent Triangulation Strategy (Creswell, 2009). The research is in its infancy and thus the instruments are under developed to test the research questions. In this study an exploratory qualitative phase was required to inform the development of the quantitative survey. Without this qualitative phase the quantitative survey could not be developed and as such had to be completed in sequence. For this purpose the Sequential Exploratory Strategy was ideal for the research areas stage of development and the goals of the study. However, generating an entirely new survey that is able to measure what is intended is a difficult and complicated task. It requires expert advice and guidance at each step of the process.
Therefore relevant industry experts were consulted throughout the process.
3.1.2 Processes
This section provides an overview of the procedures undertaken by the researcher to obtain the results of the study. The study consisted of three phases as depicted in Figure 3-2. They can be broadly described as a qualitative phase, the development of the survey and a quantitative phase.
Figure 3-2: Research Phases
The tasks undertaken for the completion of the above phases can be seen in Figure 3-2.
Some steps of the study enriched and informed others and therefore happened concurrently which is indicated by the arrows in Figure 3-2. Phase one began with an ethics approval of the project. This focussed on the sensitivity of the information to be collected for example finances and personal life events. Whilst it was important not to offend participants it was also important to obtain accurate information. Several decisions were made to ensure accurate reflections were possible. Participants agreed to an interview where all information was to be kept anonymous and not reported outside of the cumulative use to construct a questionnaire. A Convergent Interviewing process followed. The results of the data collection were analysed concurrently until the results converged to produce no new information. These results then formed the basis of the survey developed in phase two where ethics approval was again sought and granted.
The survey went through a thorough refining process involving a pre-test phase which
Qualitative
further informed the questionnaire development before being released for the formal quantitative data collection of phase three.
Phase one: The initial phases of the qualitative section were conducted concurrently with a review of the extant literature. The interviews provided the necessary depth of insight into the themes and issues lacking in the existing literature giving a more comprehensive understanding of the concepts involved.
Phase two: The data from the qualitative interviews informed the development of a substantial quantitative survey. The survey was pre-tested on 25 participants to ensure it was valid and fit for its purpose. The resultant survey became slightly smaller due to the removal of several obsolete sections as well as the reduction in outlier questions. Some rewording of questions was undertaken to more accurately measure participants’
responses. These changes are discussed in more depth later in the chapter.
Phase three: In phase three, the quantitative part of the study, a pre-participation request letter was distributed to 10,000 customers of the sponsoring body, who provided both financial and in-kind support. This satisfied ethical considerations as well as university policies and compliance with the Privacy Act. Positive responses were received from 1,000 potential participants and surveys were sent to each of the respondents. Seven hundred and seventy six completed surveys were returned, A response rate of 77.6%
from those who received surveys was achieved and considered acceptable. Surveys with low overall response rates, 5 percent to 54 percent, have been found to be only minimally less accurate with regard to representativeness (Holbrook, Krosnick, & Pfent, 2007). Missing value analysis was also conducted and missing data was missing completely at random.
The components of the three phases are outlined in Table 3-1. Each phase required certain steps, procedures and products which are also outlined in Table 3-1.
Table 3-1: Detailed Overview of Research Phases
Phases Steps Procedures Products
Simultaneous Data
Pre-testing Snowball sampling used to test the robustness and readability of the survey
Edited and final survey instrument
Data collection • 10,000 initial letters for opt in to survey
• These surveys were entered into a data base electronically
• Data cleaning & decoding process
Data set ready for analysis
Data Analysis SPSS used to undertake missing data analysis, tests of normality,
The research was undertaken in three main phases which are outlined in Table 3-1 and Figure 3-2. These phases are described in more detail in the following sections.
The results of the qualitative phase informed the design of the survey questionnaire. It was quickly established that there were many variables in all areas of decision making
which life events could influence. As a result the research became exploratory in nature.
The focus turned to establishing the relevant variables and determining their influence as a result of an impending life event. It was considered more appropriate to use broader research questions in the study rather than test specific hypotheses.