Merger and Change
5.10 The Rationale for Case Study
By their very nature, narrative research and phenomenology are not
considered appropriate approaches for this study. Grounded theory seems to have greater potential because it allows for emergence from data
collected that relate to a particular phenomenon, but it is limited for this particular study by its focus on theory creation. Ethnography, again includes appropriate elements, particularly for the initial research setting where immersion in the field for an extended period is planned. However, the examination of culture, fundamental to ethnography, and the possibility within this research of including further settings limits this approach.
Case study provides the most appropriate approach for this study as it has the flexibility to be extended to more than one case and the potential to allow for comparison between two or more cases. It also offers the opportunity to examine in depth the circumstances within a particular bounded system. The ‘case’ in this study is a primary school facing challenging circumstances.
Having established an appropriate approach for this research, associated elements are now considered. These include ethics, sampling, data collection methods and analysis of data.
5.11 Ethics
Ethical issues require consideration for different phases of the research process. Creswell (2013) suggests these to be ‘prior to conducting the study, at the beginning of the study, during data collection, in data analysis, in reporting the data and in publishing a study’ (p57). Underpinning all of these is the researcher’s personal code of ethical practice, ‘a set of rational principles appropriate to their own circumstances and based on personal, professional and societal values’ (Cohen et al., 2011, p102). Busher (2002) suggests that an ethical framework for the researcher ‘must inform their moral judgements in carrying out their work’ (p73). Qualitative research methods taught sessions and research opportunities, as part of a Masters Degree, provided an initial personal framework, augmented by further review of existing literature relevant to the methodology and methods selected for this study.
Prior to conducting a study, approval from the ethics committee or review board of the supporting academic institution should be sought. This, and the consents required in order to gain access to the settings and individuals involved requires an outline of the proposed research as well as a
consideration of the ethical issues that might arise. Codes of practice for ethical conduct of research can be useful in guiding a researcher’s approach and are produced by professional organisations such as the American Psychological Association, the International Communication Association, the Canadian Evaluation Society and the British Educational Research Association. The British Educational Research Association’s
(2004) Revised Ethical Guidelines, used for this research, includes sections covering:
responsibilities to the research profession;
responsibility to the participants (including working with children, informed consent and rights to withdrawal);
responsibility to the public;
relationships with funding agencies;
publication;
intellectual ownership; and
relationship with host institutions.
Seedhouse (1998) suggests that codes of practice are an example of the first of four layers of ethical decisions for researchers, the external layer. The remaining three, built on this, are consequential (for individuals, groups or society), deontological (what is one’s duty to do) and, individual (the core rationale of respect for individual freedom and autonomy).
Ethics has been defined by Cavan (1977) as ‘a matter of principled
sensitivity to the rights of others’ (p810). This seems to suggest that within the form of social research proposed for this study, the most significant consideration is therefore that of informed consent. Howe and Moses (1999) suggest that informed consent is the cornerstone of ethical behaviour because it respects the rights of individuals to make decisions about and exert control over their lives. Diener and Crandall (1978) define informed consent as:
‘the procedures in which individuals choose whether to participate in an investigation after being informed of facts that would be likely to influence their decision.’ (p57)
There are seven aspects to informed consent that seem important to incorporate and which guide this research. The first four are taken from the work of Diener and Crandall (1978):
competence, implying that participants are able to make their own informed decisions;
voluntarism, allowing participants to choose freely whether they take part;
full information, ensuring that consent is fully informed; and
comprehension, referring to participants’ full understanding of the research.
LeCompte and Preissle (1993) identify risk and vulnerability as two additional issues to be considered. Vulnerability in this context refers to limitations of the participants’ freedom to choose, such as age or limited mental capacity, which in this research is low. Equally, the research does not contain the elements of risk that, for example, medical research might. The researcher must ensure that no harm comes to the participants
physically, psychologically, emotionally, professionally or personally (non- maleficence). The issue of embarrassment is the final consideration added to the list and identified by Bogdan and Biklen (1992, p54). It has been addressed in this study through confidentiality (see below). The gaining of informed consent is appropriate for a case study approach as it is possible to identify in advance those individuals who will potentially be involved.
What is considered to be ‘harm’ is subjective. A careful weighing of this will need to be ‘situated’ (Simons and Usher, 2000) (interpreted in specific, local situations) and could potentially include the question of whether the end justifies the means within a given study. Cohen et al. (2011) suggest that ‘researchers should never lose sight of the obligations they owe to those who are helping’ (p86) and this view is incorporated in the principles proposed by the American Psychological Association (APA) (2002) which attempt to strike a balance between the rights of investigators to seek an understanding of human behaviour, and the rights and welfare of
individuals who participate in the research. This represents one of the potential dilemmas facing researchers.
Other ethical issues potentially relating to this research are now
considered, along with possible strategies that can be used in an attempt to overcome them. A participant’s right to privacy is a fundamental ethical consideration within research. The greater the sensitivity of information, the more safeguards are required to protect participants’ privacy. As an
issue, it can be addressed through informed consent, discussed previously, or through anonymity and/or confidentiality, outlined below. Cohen et al. (2011) state that, ‘the essence of anonymity is that information provided by participants should in no way reveal their identity’ (p91). This is possible where data collected contain no personal details, such as an anonymous questionnaire. However, Cohen et al. (2011) suggest that, ‘a subject
agreeing to a face-to-face interview can in no way expect anonymity’ (p91). Oliver (2003) identifies also that, ‘there are no absolute guarantees of anonymity, particularly in the case of people who hold named posts’ (p80). For qualitative research, particularly where a study is carried out in a single setting, or a very small number of settings, it is difficult to ensure that data are totally unattributable: interview transcripts and field notes inevitably record sufficient detail to make participants identifiable. Even where anonymity is preserved beyond the setting, the participants are likely to recognise themselves and one another (Ellis, 1995). This raises the possibility that publication will cause private (or community) shame or embarrassment, even where it does not lead to public humiliation (Ellis, 1995; Hopkins, 1993).
Confidentiality involves ‘not disclosing information from a participant in any way that might identify that individual or might enable them to be traced’ (Cohen et al., 2011, p92). In order to protect those involved in qualitative research, assurance that confidentiality will be maintained through the deletion of identifiers, such as, changing of individual and institution names, can be provided. The informed consent process can be used to provide assurance that confidentiality will be maintained where this is a possibility. Howe and Moses (1999) argue however that in order to create the ‘thick descriptions’ that give strength to qualitative research a level of detail is needed ‘that cannot be obtained if privacy, confidentiality and anonymity are required’ (p44). The balancing of these potentially conflicting
requirements in a ‘costs/benefits ratio’ is a further ethical dilemma that needs careful consideration.
Deception is another ethical dilemma. It has been suggested that it exists at some level within all sociological research (Aronson et al., 1990; Bailey, 1994; Cohen et al., 2011). At its most extreme, deception involves covert
can also involve not telling the truth, telling lies, or compromising the truth. Evidence suggests that there are no simple answers to the ethical issues of covert research (Bulmer, 1982) and it can do much harm (Erikson, 1967; Hornsby-Smith, 1993). Conversely, it might be considered that covert methods have advanced our understanding of society (Douglas, 1976; Bulmer, 1982).
This research is completely overt in nature but the less extreme aspects of deception are identified. It could be considered that deception was used by suggesting to participants that the focus was on investigation of ways that primary schools facing challenging circumstances improve, without acknowledgment that decline could equally be the outcome examined. Whilst not intentionally seeking to deceive, but rather to be positive with participants of the study, this view could be considered to potentially
influence the responses and behaviour of those involved. However, it could be suggested that telling participants in detail what is being examined might equally bias the outcome (Plummer, 1983).
The ultimate decision with regards to deception for this study involves the costs/benefits ratio in combination with the issue of non-maleficence. While the omission of information shared with participants could be considered to be deception, harm was unlikely to result and this was also balanced against potential bias in outcomes being minimised.
Some ethical considerations are directly related to the nature of the
research technique employed. These, as well as those relating to aspects of data analysis, will be discussed as part of later sections in this chapter.
5.11 (i) The ethics of this research
Following identification of the research opportunity, and careful
consideration of the potential ethical issues involved, an application to the Ethics Committee of the University of East Anglia was made. Once accepted, a formal proposal was made to the Local Authority (with overall responsibility for the County’s schools) and then the Chair of Governors and headteacher of the initial school setting. Subsequent to that
gaining informed consent was completed at the point of entry to the field. This included providing information about the aims and context of the study, the level of commitment involved, the form of data to be collected and the opportunity to withdraw all or part of this, and the ability to withdraw from the research at any point. Access and consent were carefully negotiated and permission to digitally record interviews also gained. The same process was used for the subsequently identified case study schools. Copies of the Ethics Application and the individual Consent Form are provided for reference in Appendices 2 and 3 respectively.
5.12 Sampling
Sampling takes one of two main forms; probability (also known as random) sampling and non-probability (also known as purposive) sampling. In the latter, the researcher deliberately selects a particular section of the wider population. Examples of this could be schools facing challenging
circumstances from the wider population that includes all schools, or a narrower focus on particular schools within that group identified as facing challenging circumstances. Non-probability sampling is the approach taken more often within qualitative research. Cohen et al. (2011) state that, ‘each sample seeks only to represent itself or instances of itself in a similar population, rather than attempting to represent the whole, undifferentiated population’ (p155). It is considered an appropriate strategy for this
research.
Within qualitative research there are different levels of sampling.
Researchers can sample at the site level, event or process level, and at the participant level. Within this study, site level sampling will be followed by participant level sampling. There are also different types of non-probability sampling, such as, quota sampling, convenience sampling, snowball sampling, purposive sampling and volunteer sampling. Purposive sampling, also referred to within the literature as ‘purposeful’ (Creswell, 2013, p156), is considered appropriate for this study because, as Cohen et al. (2011) state:
typicality or possession of the particular characteristics being sought. In this way, they build up a sample that is
satisfactory to their specific needs.’ (p156)
This sampling therefore ‘enables the full scope of issues to be explored’ (Lincoln and Guba, 1985, p39). Lincoln and Guba (1985) go on to suggest that, ‘the natural mode of reporting [following purposive sampling] is the case study’ (p43). Related to the collective case study approach identified as appropriate for this study, Creswell (2013) suggests that ‘often the inquirer purposefully selects multiple cases to show different perspectives on the issue’ (p99).
As suggested above, there are a number of sub-divisions within purposive sampling. Teddlie and Tashakkori (2009, p174) provide a typology of those kinds of purposive sampling that are used to achieve representativeness or comparability, with which this study is concerned. They include within this list, typical case sampling, extreme or deviant case sampling, intensity sampling, maximum variation sampling, homogeneous sampling, and reputational case sampling. Of these, Miles and Huberman (1994, p28) identify ‘maximum variation sampling’ as being most often selected in qualitative studies. Sites or participants are selected because they possess characteristics that are quite different to one another in relation to identified overall criteria. An example of this could be a study examining the life experience brought by newly qualified teachers to their practice where young and more mature teachers in the early stages of their careers are selected. Maximum variation sampling is an appropriate sampling strategy for the later parts of this research as it allows for further case study schools to be identified that are facing challenging, but different, circumstances.
5.12 (i) Sampling for this research
The school used initially within this study was not selected through a process of sampling; it presented through circumstances unrelated to research. The two subsequent schools were however selected using purposive, maximum variation sampling. The overall criteria related to primary schools facing challenging circumstances with the second and third schools each providing a different example to the initial school; one with an
Ofsted Notice to Improve (and an acting headteacher) and the other a new school that was being created through the amalgamation of two
neighbouring but pre-existing schools and with a new headteacher of proven experience.
Sampling for the participants to interview within each school was considered during the initial stages of the research, forming part of the application to the Ethics Committee. Considerations included the number in the sample and the positions that they might hold within the school. A member of the senior leadership team, a class teacher and a local authority representative were chosen through maximum variation sampling to
represent potentially diverse perspectives of the same set of challenging circumstances. In reality, more than one individual representing each of the posts held was identified initially to allow for individuals declining the invitation to be interviewed.