The search for answers to the research problem questions driving the research is the key to scientific rigour. This search for explanation beyond description of a situation or phenomenon produces research of more substantial adequacy and interest. However, understanding explanations relies on the provision of
preliminary accurate and reliable description (6 & Bellamy, 2012). 6 & Bellamy (2012) outline that principles of scientific rigour should be demonstrated throughout all phases of a research process and in the reporting of a research project; adequacy and reliability should be evident, for example, in the research questions, the design of the research and the methods of data collection.
Qualitative research investigating user’s opinions and experiences lends the data to be subjective in nature. Reliability and validity have important implications for the success of subjective measures (Noyes & Baber, 1999). Reliability refers to the internal consistency of a measure, meaning the ability to repeat a research process and obtain the same results. Validity refers to whether the method of research measures what it was intended to (Noyes & Baber, 1999). These measures exist independently of each other, but increased instance of both increases the rigour of the research. Golafshani (2003) outlines that reliability and validity are treated separately in quantitative research but are not viewed separately in qualitative research. In fact, the terminology reliability and validity tends not to be used in qualitative research (Golafshani, 2003; Robson, 2011).
Researchers adopting qualitative, flexible research designs avoid the terms reliability or validity, or even propose they are not relevant in such research (Guest, 2012; Robson, 2011). Morse (1999) argues that to state these terms are not applicable in qualitative research is to infer that qualitative research cannot be reliable or valid (in: Robson, 2011 pg 155). By the definition that scientifically rigorous research must be reliable and valid, if qualitative research is not reliable and valid and it cannot be rigorous science. Robson (2011) points out that, whilst an extreme interpretation, this argument poses an important consideration for presenting scientifically rigorous qualitative research. Straus and Corbin (1990) discuss this in their earlier writing by
97 proposing that the “usual canons of good science” require redefinition in order to fit the realities of qualitative research.
This implies recognition that the aspects of reliability and validity in research should not be dismissed but rather redefined to be more relevant to qualitative research design. Researchers taking a qualitative, flexible approach to their study have replaced the terms of reliability and validity with other terms such as credibility, dependability, confirmability, trustworthiness (Guest, 2012; Robson, 2011). These terms place more emphasis on the confidence of the truth and accuracy of the research findings and whether the research process is consistent and carried out with attention to the conventions of qualitative research (Guest, 2012).Whilst reviewing and recognising other researcher’s application of terminology and opinion in this research arena, Guest et al (2012) argue that the terminology reliability and validity should be kept in qualitative research, whilst recognising their limitations.
Researchers have discussed the implications or need for reliability and validity in particular regard to case-based research. The issues raised are very much in-line with those concerning qualitative research. For example, 6 & Bellamy (2013) state that case studies must meet certain standards of reliability and validity but that it is unclear what these standards are in order to demonstrate that they have been met. Denzin & Lincoln (1998) use alternative terms such as dependability and
confirmability to refer to the concept of reliability, in the same sense as other qualitative researchers who redefine the term using similar terminology to more appropriately fit the nature of the research.
A limitation of case-based research is that it does not involve a large participant sample or volume of data. It is, therefore, susceptible to producing insights particular to that case and the relevance of this in other contexts is unclear (6 & Bellamy, 2012). This, as with qualitative research explained above, has implications for the replication of research if applied to other cases. In this regard, the term reliability as applied in quantitative research is not suitable terminology. Within and across social contexts great variance can be observed, so empirical research is difficult to replicate. However, if regarding the rigour of research in terms of
98 credibility and trustworthiness, it is important to demonstrate that data were obtained and analysed in ways that were pertinent to the research design and data were accurately used to produce research findings (6 & Bellamy, 2012)
Small samples and a lack of being able to replicate the research or generalise findings is not to dismiss the value of a case-based research, however. 6 & Bellamy (2012) purport that researchers may overstate the particular nature of case-based research in such a way that it sacrifices the possibility of contributing to theory. However, as Gillham (2000) outlines and as described in section 3.3.4 reviewing case-based research, a single case holds value in contributing to theory through falsification of a phenomenon or highlighting the existence of a phenomenon previously not recognised but which cannot be ignored.
Darke et al (1998) highlight that the reporting of case-based research can be difficult and has often been considered to be lacking in rigour. In order to increase validity within case-based research, 6 & Bellamy (2012) outline a strategy applicable to interpretive, qualitative research in general. They describe checking back with participants to ensure that the data collected accurately captured the subjective meaning and that the inferences made by the researcher from the empirical study make sense in the context of their experiences. This was achieved in this research by the format of the research methods. An empirical study of contextual interviews followed by further exploration with participants in the same case through a focus group and, subsequently, more contextual interviews enabled a validation of insights and greater contextual understanding.
An audio or video recording of data collection during these studies ensures that all information and a greater level of detail are captured and retained for the duration of a research project. This not only adds value to the research process by enabling flexibility in carrying out conversational research methods, but provides support to substantiate the validity of the research. Inaccuracy or incompleteness of data reduces the validity of what the researcher reports. Therefore, a recording of the data with all details captured ensures the researcher has all the information available
99 to work with rather than just their own note-taking or memory, which could result in data being omitted if not registered at the time.
3.5.1 Pilot studies
A pilot study is a small-scale version of the intended process of carrying out a major research study. It can also be known as a feasibility study because the purpose of the small-scale study is to assess the feasibility of the planned protocol (Robson, 2011). It is an opportunity for the researcher to practise and prepare for
conducting an empirical study. It also allows for a pre-test of particular research measures, such as an interview guide or questionnaire (van Teijlingen & Hundley, 2001). One of the key reasons conducting a pilot study is important is that this pre- testing may give prior indication of difficulties or the research: where an instrument is too complicated or inappropriate, where protocols may not be suitable or where failures may occur (van Teijlingen & Hundley, 2001).
Conducting a pilot study cannot guarantee the success of a major study. A pilot study being conducted successfully cannot create an assumption that the main study will also be as successful. Although it can give an indication of the likely success or failures, they are based on small samples and cannot be extrapolated to the main study with accuracy. It can, however, reduce the risk of failures by identifying any potential areas early on. By pre-testing the feasibility and practicality of a research study process, any issues may be identified enabling them to be addressed and modified where necessary prior to commencing the main study.
There are aspects of case study research that can make piloting more difficult to do (Robson, 2011). Robson (2011) suggests that for case-based research piloting is less crucially important. If only one case is to be considered or there are such particular features of the selected case that an appropriate exemplary equivalent could not be referred to, then a pilot study would not be suitable (Robson, 2011). In this
research, a pilot study was conducted where practicable but not in all parts of the research. This is due to such particular parameters of the research making the sample, albeit important, relatively small and equivalent representative samples
100 difficult to obtain without detracting from the main study, which was of greater importance to the research process.