3. Research Design and Methodology
3.4 Further methodological issues to be addressed
The following key terms are explained: Reliability, Generalizability,
Transferability and Validity as well as ways to address methodological issues.
Reliability
The term reliability refers to results which are considered consistent over a period of time and are an accurate representation of the total population under study, as well as results being produced under a similar methodology (Golafshani 2003). If there is consistency with answers provided by the qualitative and quantitative methods at two different times, then the research methodology and its methods are considered to be reliable. It is as such determining whether the results are replicable.
Generalizability
Generalizability refers to the ability to generalise the findings from the sample population to the overall population to which it was drawn. Such generalizability may not be absolute, but is considered statistically probable. It requires obtaining data from large populations which is more associated with quantitative research. The larger the sample, the more generalizable the results are (CSU 2013).
Transferability
Transferability is similar to external validity in that it refers to the extent of which findings can be generalised (UniSA Institutional Repository 2013). As
95 such it refers to whether findings and relationships found with a specific population, at a specific time and place would be observed with other populations, at other times and at other places (Maxfield and Babbie 2001).
This includes various sub groups, the current time era, settings and not the overall population (in which generalizability refers). Such sub groups with this research context include disability groups which in the context of this thesis is a broad term defined as those who consider themselves with an inability or limited capacity to use technology. Both Lincoln and Guba (1985) emphasise the need for providing the appropriate data to allow for such transferability to take place rather than an index for it.
Validity
The term validity refers to whether the research measures what it was intended to measure and whether such results are accurate. It considers the integrity and therefore validity of the findings generated from the research. In other words it is assessing the accuracy of the results provided. As such the term validity can be described by a number of terms including trustworthiness, worthy, relevant, plausible and confirmable, credible or representative.
Further, the term validity should also consider a ‘realist’ approach (Denzin and Lincoln 1998). This is whereby validity is determined represents accurately the phenomena that are intended to describe, explain or theorise (Hammersley 1997). Validity is of considerable importance in producing findings of integrity and making the research effective and contributing to knowledge.
96 There is no “state” of validity as such, but validity should be considered as a matter of degree instead (Gronlund 1981). While methodological strengths and weaknesses have been evaluated pertaining to the validity of the different data collection methods used in this research, the concepts and types of validity of the entire methodological approach to the research are also of importance.
Issues that pertain to validity (and how they were a potential issue with the research) included:
Ensuring that resources were adequate
Choosing an appropriate methodology
Selecting appropriate instrumentation
Using an appropriate population sample
Devising an appropriate instrument
Minimising bias (Cohen et al 2005)
Each of these issues was addressed at the design stage of the research and the ways in which these were met via solutions are described.
Adequate resources
Adequate resources were made available for the quantitative surveys by providing appropriate instructions, explanations, the actual questionnaires in print format, a SAE to return the questionnaires while also allowing an appropriate amount of time to complete the questionnaires (e.g. in the lunch clubs). Each Age UK lunch club was contacted via telephone and consent was obtained prior to any questionnaires being sent out, with the research itself
97 being fully explained. With the interviews, focus groups and cultural probe workshops, papers and pens were made available, should those involved prefer to write down notes or observations based on their experiences and discussions. The interviews were kept within a reasonable timeframe (the maximum duration of 10 minutes) while still eliciting all the necessary data so that it could be turned into information. Any further discussion or explanation about specific aspects of the research at the participant’s request was provided by the researcher at his discretion. Any clarification or explanation about a question asked was again provided by the interviewer as it was absolutely paramount that, in order to maximise validity, the question should best be understood.
Appropriate Methodology
The methodology was chosen due to its problem solving nature approach by combining both qualitative and quantitative methods to answer the research questions with appropriate validity via triangulation. A substantial part of it also involved working interactively and collaboratively with older adults in an ethical manner, while still being an excellent problem solving approach. Each data collection method or stage was designed to fulfil the individual objectives and aim, while also incorporating triangulation (as discussed in the next section) to answer the aim and objectives while maximising validity. A key component of the methodology allowed and encouraged the participants to self-reflect their learning methods and consider new ones or aspects of methods that influenced learning as a spirit of reciprocity. Finally, putting users
98 at the forefront of the research was considered imperative to elicit data to understand the ways in which older adults learn to use or could use technologies with collaborative learning and not necessarily in experimental based settings.
Instrumentation
The methodology incorporated a number of strategically chosen instruments (also known as methods) to collect, analyse and interpret the data, exploring both related and unrelated topics to answer the objectives and aim of the research while trying to reach consensus. They were mainly qualitative methods complemented by a quantitative method. They allowed the participants to project their views, document their experiences (with the researcher assisting such documentation) as well as answering the self-completion questionnaires. To re-iterate the data collection methods consisted of semi-structured interviews, focus groups, and cultural probe workshops.
Both coding and framework approaches were utilised to analyse and interpret the data. A platform was formulated to interpret and synthesise the data into producing learning strategies within a learning framework consisting of different categories.
Population sample
An appropriate population based on a convenience sample approach was chosen to participate in each of the 10 interviews, 3 focus groups, 11 research diaries and the surveys. The population consisted of any person over the age of
99 50. This population was sub-divided into further categories which were differentiated by a) gender and b) age categories (e.g. 50 to 59, 60 to 69 and 70 and over) and c) their previous experience of technologies.
The convenience sample strategy
The convenience sample strategy was used because data can be gathered efficiently due to wide availability of the population sample (those over the age of 50) for both qualitative (e.g. interviews and focus groups) and quantitative (e.g. surveys) methods. It involves the selection of the most accessible participants (Marshall 1996). This is because most, if not all, people who are over 50 will be situated at the Age UK lunch clubs and Nottingham Elders Forum. As such it is readily available and cost effective, as any member of the over 50 age population can participate in the research.
Devising a platform for producing the framework
The formulation of a platform was important in producing a framework of effective learning strategies. It involved taking the analyses from each question in each data collection method. So question 1 for example in the semi-structured interview was answered 10 times by 10 different participants. The answers were analysed via coding and framework analysis. The findings were compared in relation to the question. The key to the platform was its simplicity, in processing the outputs from the framework and coding approach, into producing meaningful information in a structured and framework consisting of
100 different categories as a document. The platform was required for the aim yet objectives 1 and 2 were pre-requisites.
Minimising bias
As discussed previously, it was important to maintain a neutral and supportive stance when engaging with all participants, yet not to influence any answers provided by the participants. This was especially important when conducting the interviews, as well as interpreting the data. Bias can result in both the qualitative (e.g. interviews, focus groups and cultural probe workshops) and quantitative (e.g. surveys) methods. Key causes of bias in interviews and focus groups include the characteristics of the interviewer, respondent and nature of the questions (Cohen 2005). In particular it was important to be both neutral yet supportive.
To minimise bias with the interviews and focus groups (that is the data collection through to the interpretation) the following was applied: a) a neutral and opinion-free approach b) no answers were sought that supported any preconceived notions and c) any potential misunderstandings were addressed, both with the questions that were being asked by the interviewer and the answers provided by the respondent. With the quantitative surveys the main drawback was the sample size. Both questionnaires received a total of 66 responses. While this was not necessarily generalizable to the older adult population, the findings were used as an indicator and as such for informing, as well as testing or reinforcing previous findings. There was also some bias in the cultural probe workshops with observations and the recordings obtained. This
101 is due to interpretations being present. The presence of the researcher may also have influenced the learning of technologies by the older adult participants and therefore the findings.
Statistical significance (of the quantitative data and findings) which is about the estimation of confidence that the results can be generalised to the overall population is also of importance. In the context of this research it determines whether there is a significant relationship between at least two variables to explain phenomena. It involves setting up a null hypothesis and establishing a level of ‘acceptance’.
The next section describes some philosophical epistemological considerations.
Such considerations underpin the research design and strategy in providing knowledge and understanding with appropriate validity and integrity.