CHAPTER FOUR: RESEARCH DESIGN
4.14 Research Reliability and Validity
Several strategies were used in this research study to enhance the validity and the following section addresses the specific concerns with regard to validity of both the quantitative and qualitative methods. Firstly, the validity of postal questionnaire can be seen in the issue of non-response (Cohen et al., 2007) In other words a question I considered was, would the participants who failed to return their questionnaire have given the same distribution of answers as those who did return the questionnaires. In order to reduce volunteer bias and to ensure a maximum response rate, a comprehensive follow-up strategy to the initial distribution of questionnaires was implemented. This strategy has been outlined earlier in this chapter. A further issue in considering the reliability and validity of the questionnaire is that of sampling. The steps taken to ensure a representative sample have been also been outlined earlier in this chapter.
The trustworthiness or otherwise of findings from qualitative studies is the subject of much debate (e.g. Robson, 2002). With regard to the qualitative methods used in the study both issues of internal and external validity were considered. Some researchers prefer the terms credibility as a descriptor of internal validity of research and for external validity, transferability has become widely accepted in qualitative research (Mertens, 2005). Internal validity refers to how research findings match reality. From a constructivist perspective which is the basis of this study, triangulation remains a principal strategy to ensure for validity. Information from a variety of sources including teachers, parents and children was obtained in order to make the research findings as robust as possible (Maxwell, 2005). It was felt that the use of triangulation of both quantitative and qualitative research methods and multiple data sources also strengthened objectivity.
Because the researcher is the primary instrument of data collection, reflexivity or the process of reflecting critically on the self as researcher was a key element incorporated throughout the research process. From the beginning of the research and on a continual basis throughout the research process, time was spent in critical self- reflection regarding assumptions, worldview biases and theoretical orientations. I
carefully scrutinized the data for extracts that conformed to and contradicted analytic themes (Miles and Huberman, 1994) while a ‘constant comparisons’ approach (Silverman, 2000) was applied to further refine research findings. Merriman (2009) suggests that qualitative researchers can never truly capture an objective reality, however, the process of constantly looking for alternative explanations in the data has added to the rigour of the research.
A common strategy for ensuring internal validity is respondent validation (Merriman, 2009). Respondent validation facilitates continual questioning of the data through engagement with the participants and their responses to emergent findings. Lincoln, Lynham and Guba, (2011) promote the use of respondent validation in case study research in order to allow respondents’ perspectives on the emerging nature of the findings to be included within a grounded theory data analysis framework. The inclusion of respondent validation also adds further rigour and credibility to the research findings (Merriman, 2009). However, Lacey and Luff (2001) also caution that competing or contesting interpretations by respondents, dynamics of trust between respondents and researcher and confidentiality issues can limit or call into question the effectiveness of such respondent validation. Informal respondent validation was carried out with the participant teachers, parents and children through individual discussion of their experiences and through their ongoing communication and contact with me throughout the research phase. This resulted in a deeper exploration of varying emergent patterns in the data analysis. Although presented as a linear, step by step process, the research was in fact an iterative and reflexive process and my engagement with the data through a process of constant comparison and analysis combined with my ongoing reflection and reflexivity served to question and challenge emergent themes of this study. Finally, I also believe that good levels of credibility were established by conducting the survey, observations and interviews over a prolonged period of approximately a year.
External validity is concerned with the degree to which the findings of one study could be applied in other situations, in other words the transferability of the study. One of the most commonly mentioned strategies to enable transferability is the use of thick descriptions (Merriman, 2009). Thick descriptions of events were provided so as to reflect the complexity of situations and to strengthen the transferability of the research findings. A further strategy for ensuring the external validity of the study was the use of maximum variation in the case study sample.
Although all schools in the case study were small schools with up to 56 pupils, a variation of cases was selected in the Munster region which included urban and rural schools, DEIS and non-DEIS schools and schools where class size varied. Finally, an extensive audit trail of research activities was kept in order to ensure dependability in this study. The audit trail included records of raw data, field notes, research journal and details of coding and analysis used.
4.15 Summary
Using a concurrent triangulation mixed-methods design (Creswell et al., 2003), this study aimed to present a comprehensive picture of early childhood pedagogical practices in Junior Infant classes of multigrade schools. The study which was conducted in eight case study multigrade schools, explored the perspectives of the teachers, children and their parents. An eclectic methodology was used to gather data from a broad range of sources which included observational data, interviews and focus groups to gather information about the practices as well as a content analysis of WSE and WSE-MLL Inspectors’ reports. All of this information was supplemented by data gathered through the questionnaire survey which documented pedagogical practices and attitudes of teachers in a way that more generally represented a nation-wide sample. Grounded theory provided a framework for analysis based on the rich data sources researched within the interactions of the participants in the situated social context of multigrade schools and classrooms. The broad range of data sources; the prolonged nature of the research period; the on-going analysis and review of findings and multiple triangulation facilitated the emergence and validation of the key findings which are presented in Chapters 5 and 6.