• Tidak ada hasil yang ditemukan

CHAPTER 6: CONCLUSIONS AND RECOMMENDATIONS

4.4. RESEARCH DESIGN AND METHODOLOGY

4.4.6 Trustworthiness

Qualitative research needs to be trustworthy and reliable in order to represent, as closely as possible,the experiences of the participants.Trustworthiness refers to how data are collected and analysed, and how results are communicated. The researcher needs to understand and accurately represent the situated nature of the participants' interpretations and meanings. Carefully planned data analysis and interpretation would reveal the understanding of participants and a good quality of qualitative data analysis (Ely,1991;Ezzy,2002).

According to Guba's (1981, in Krefting, 1991) model of the trustworthiness of a qualitativestudy,data should be based on four criteria oftrustworthiness: truth value, applicability, consistency and neutrality. These aspects should be used by the researcher to assess the value of the findings of their qualitative studies (Krefting, 1991).

4.4.6.1 Truth value

The 'truth value' of a qualitative study refers to how confident the researcher is with the findings of the information gathered, as well as the research design used, the participants and the context in which the information was gathered. The information gathered should reflect human experiences as they are lived and perceived by the participants (Krefting, 1991; Strydom, 2002). Lincoln and Guba (1985, in Krefting, 1991; De Vos, 2002), termed 'truth value' as 'credibility', when the researcher presents all realities revealed by participants in a study as adequately and accurately as possible. These accurate interpretations of information based on human experience arevalid when they are recognisable to readers who share the same experiencesas the participants (Krefting,1991;De Vos,2002).

Another way of establishing truth value is byusing different methods to produce data.

Multiple research techniques help the researcher gain alternative perceptions about a phenomenon, using different participants in different set-ups. This helps the researcher gain a better understanding ofthe phenomenon (De Vos,2002).Igathered

data by using interviews, observations and a reflective journal (Ely, 1991; Krefting, 1991). I interviewed a few individual participants once and focus groups twice. I also interviewed participants at different times of the year, over a period of six months. I interviewed the focus groups at the end of a school term for the first time, and the second time at the beginning of a new school term. I interviewed some participants during the autumn season, and others during the winter season. I grouped the educators according to experience. I had one group of educatorswho hadworked with learners with autism for a few years, and I had one group of educators who had just started working with learners with autism. I did the same with the individual interviews (Krefting, 1991). This prolonged fieldwork contributed to credibility.

When reporting the findings, direct quotations of the participants were included as a chain of evidence.

4.4.6.2 Applicability

Each situation of the qualitative research is unique and the purpose of a phenomenon or an experience is to study human experience. It is therefore difficult to generalise these experiences to a bigger population group (Krefting, 1991; De Vos, 2002).

Generalisability is not actually relevant to qualitative studies, as a qualitative study is conducted in a natural setting,not a controllable environment. Therefore,the need to generalize the findings to other contexts and settings, or with other groups of participants, is placed on a researcher who wants to replicate the study. The researcher must also keep in mind whether the information and behaviour observed during the interviews are representativeor unusual for these individuals (Krefting,1991).

'Transferability', according to Guba (1981, in Krefting, 1991), is the criterion to be used in qualitative research to value applicability or contextual relevance. This is when the findings of the research study matches similar contexts outsidethe original study Lincoln and Guba (1985, in Krefting, 1991; De Vos, 2002) on the other hand, feel that the researcher should present the reader with enough information to allow comparison to take place. It is the readers' responsibility to compare or transfer the findings to another situation(Krefting, 1991;DeVas, 2002).

4.4.6.3 Consistency

Consistency of data refers to how consistent the information would stay if the research were repeated in a different context, using the same participants (Krefting, 1991; Ezzy, 2002). In a qualitative study, consistency of data is difficult due to unexpected events. The researcher does not try to control the participants or the information gathered, but learns from the various experiences from participants and the situations where the information is gathered.

In a qualitative study, consistency depends on the uniqueness of the participants and the researcher, as humans are not consistent in their emotions throughout the study.

Different perceptions from participants, rather than an average perception, also help with the uniqueness of the study, as well as the good insight of the researcher on a specific topic under research and of the participants (Krefting, 1991).

Dependability, as part of consistency,is where the researcher should be aware of any changes in the research i.e. on the part of the participants, as the researcher is looking for a variety of experiences, and not an average experience. Lincoln and Guba (1985, in Krefting, 1991; De Vos, 2002) are of the opinion that the researcher should be prepared for changing conditions in the phenomenon under study, and in the research design (De Vos,2002).

4.4.6.4 Neutrality

The researcher needs to emphasise neutrality by not being biased in terms of research procedures and results. The findings need to be from information gathered by the participants, and not influenced by values, interest or hopes of the researcher. The researcher should be objective,and any preconceived ideas formed by the researcher should be ignored or bracketed. In qualitative studies the researcher needs to have a good relationship with the participants with prolonged contacts and even observations. However,the researcher needs to keep the data neutral by establishing truth value and applicability (Krefting, 1991; Ezzy, 2002). Ely (1991), refer to it as

being flexible and open to the data and not being cognitively rigid, as this may interfere with the way different aspects of data are seen. According to Lincoln and Guba (1985, in Krefting, 1991; De Vos,2002), if the findings of the research study could be confirmed by another, it is valid. This aspect they call 'confirmability'.

Independent coding is helpful in ensuring the confirmability of the findings.