Health and Social Research Methods
Researchers can therefore be said to benefit those who have the most influence on the way in which data is analysed and those that have greatest access to the results of that analysis. In general these are not the people subjected to the research process and again the argument can be made that the research process itself disadvantages groups of people because of the power inherent in the position of ‘researcher’. This is a particularly thorny issue for youth and community workers who seek to work in ways that are anti-oppressive and who have a commitment to the challenging oppressive practices.Health and Social Research Methods
What stance do we take when faced with a need to research, an inherent quality of which involves disadvantaging the very groups we seek to serve. Do we have the right to conduct research ‘on’ disadvantage groups in the name of knowledge especially when that knowledge is only needed for the purposes of academic qualifications? Ethically then CDA and all other forms of social research must be carried out in such a way that the potential for reconstructing oppressive practices and perpetuating the interests of the dominant and powerful, is constantly on the agenda.
ORDER A PLAGIARISM -FREE PAPER NOW
Researchers therefore have a moral and ethical obligation to those involved in or affected by the research undertaken to the point that ‘being ethical limits the choices we can make in pursuit of the truth. Ethics say that whilst truth is good, respect for human dignity is better, even if in the extreme case, the respect of human nature leaves on ignorant of human nature’ (Cavan in Cohen et al, 2000:56). Ethics in social research is a subject of much debate not least concerning the need for guidelines. May identifies two approaches to the issue of guidelines, deontological and consequentialist.
The deontological approach, based on the work of Kant, says ‘ethical judgments in social research would – follow a set of principles which guide the conduct of the research itself. Research ethics takes on a universal form and is intended to be followed regardless of the place and circumstances’ (May, 2001:60). Consequentialist however argue that such guidelines would never cover all eventualities and that ethical considerations arise from particular instances. Decisions therefore need to be made in the context within which research is being conducted and with the consequences of research firmly in mind.
Both points of view are however problematical. Whilst consequentialists may be correct in assuming that guidelines can never cover all eventualities, their ‘leave it to the researcher’ approach leaves the way open for unscrupulous practice. But the fact that there is no one right or wrong way of determining ethical behaviour does not mean that ethical issues can be ignored. On the contrary the very multiplicity of ethical issues that may arise in research means that ethical considerations should be the basis on which most research decisions are made.Health and Social Research Methods
In general most ethical research debates concern informed consent, respect for privacy, confidentiality or anonymity, an awareness of the possibility of harm to the researched and the use of deceit or covert research. Informed consent advocates that participants in research should be fully informed of the work in which they are being asked to take part and that they should be given the option of saying no or withdrawing at any time. The background, purpose and method of research should be clearly explained together with an explanation of the possible uses to which results may be put.
Unfortunately the power relationships between researchers and researched can affect the quality of informed consent, since it assumes an equal relationship in which people feel able to say no. Cohen et al argue that informed consent perpetuates the one sided nature of research since ‘it may leave relatively privileged groups under-researched (since they will say no) and underprivileged groups over-researched (they have nothing to lose and ‘yes’ in hope)’ (Cohen et al, 2000:58).
There are also difficulties involved in obtaining the informed consent of parties other than the intended research group, as for example in the case of children, where the informed consent of parents is needed before they can be approached. Children can therefore be effectively silenced by the need for informed consent since adults have the power to say no on their behalf. This is therefore particularly problematic for those involved in the field of youth since much of their research will be carried out with young people for which parental consent will be necessary.
Respect for privacy raises issues concerning sensitive material and the balance needed in reporting private issues in the public domain. Anonymity appears fairly unproblematic on one level, but on closer inspection difficulties life in the extent to which a piece of research can maintain this. Small scale studies (such as those carried out by discourse analysts) have the potential to identify participants without naming names, and researchers need to be aware of this.
Confidentiality is another area in which great care is needed since child protection policies can specify clear instances in which confidentiality, and for that matter anonymity and privacy, could and should be breached. An awareness of the possible harm that taking part in research studies can do to participants needs to be considered at every level and honesty about the extent to which the researcher can control this is vital. Informed consent is often viewed as a ‘get out’ clause for some researchers (I am not one of them) who would argue that participants knew what they were getting into since consent Health and Social Research Methods
‘places some of the responsibility on the participant should anything go wrong in the research’ (Cohen et al, 2000:51). But the argument about possible harm becomes less clear when one considers the extent to which the researcher has control on the use to which his/her research is put. Funders may use the research for their own purposes irrespective of the researcher’s intent and responsibility to funders/sponsors of research involves another delicate balancing act in serving the interests of both parties.
The use of deceit and covert observations as a research method are generally assumed to be unethical although there is an acceptance that some forms of research, for example, those which may represent a risk to the researcher (i. e. undercover investigations of criminal behaviour) and those in which it is almost impossible to obtain information openly, can be seen as acceptable because the researcher has put themselves at risk. However, this aspect of ethical practice is more involved that is first thought especially for CDA research methods that need to research ‘naturally occurring’ language.
If our social world is created within the relationships between people then it stands to reason that interaction between the researcher and the researched will change the way people behave and any language used as research can only be viewed as that which occurs when participants know they are being observed. This will prevent the acquisition of ‘naturally occurring’ talk and therefore has the potential to invalidate the research. That said CDA also uses a wide range of text and talk to provide data for analysis.
Some of these, newspapers, reports, minutes of meetings etc. , separate the researcher from the people creating the text, significantly reducing or eliminating any interaction at all. But does published writing, parliamentary debates and other written texts do away with the need for informed consent and are there issues concerning copyright laws. Some would argue that where copyright is not an issue and when writing is within the public domain, consent isn’t an issue and researchers have the right to analyse what is being achieved with texts.
This makes the use of CDA a particularly attractive research tool since it can use political writings and texts to deconstruct issues o power, hegemony and ideology which is a significant aspect of the work of youth and community workers. However, the positioning of the research in the acquisition of research data and the role they play in the research process is an acknowledged aspect of qualitative research in general since ‘the researchers self plays a significant role in the production and interpretation of qualitative date’ (Denscome, 2000:208).Health and Social Research Methods
In critical discourse analysis this is more specific since ‘the data do not exist out there waiting to be discovered – but are produced by the way they are interpreted and used by the researcher’ (Denscombe, 2000:208). Discourse analytic research simply cannot escape the fact that the researcher is indulging in an exercise of power over other peoples works by imputing their own understanding and analysis as to its import and meaning.
Again the power relationship between researcher and researched is a significant factor especially in the field of CDA with its commitment to understanding the use of language in dominance and particularly for youth and community workers who may wish to use it whilst attempting to work anti oppressively. The claim that ‘discourse analysts conduct research in solidarity and cooperation with dominated groups’ (Schiffrin et al, 20001:352) is very much open to challenge.
In fact Deborah Marks argues that ‘the exercise of power is all the worse when covered by an illusion of ‘democratisation’ and the disingenuous fantasy of ’empowerment” (Marks in Burman & Parker, 1993:157). This means that the researcher’s standpoint, the philosophical beliefs, political motivations and purpose in undertaking researcher should be ‘expressed openly so that the research does not suffer from hidden subjective motives of the investigator, and the readers are not misled in any way’ (Saratakos, 2000:123).
During the course of the research, the researcher needs to engage in the reflexive practice of scrutinising their role in the research process since ‘the researchers identity, values and beliefs cannot be entirely eliminated from the process’ (Denscombe, 2000:208). That values enter the field of research is not an issue confined to qualitative research in general nor critical discourse analysis in particular. All research is subject to values from inception to conclusion.Health and Social Research Methods
ORDER A FREE PAPER NOW
The values of the funders or sponsors, the researchers and the groups researched will affect all areas of the research process and although there are researchers who advocate objectivity and a search for the ‘truth’, critical discourse analysts take a political and therefore value laden position in their analysis of data. Teun Van Dijk says ‘unlike other discourse analysts, critical discourse analysts (should) take an explicit socio-political stance: they spell out their point of view, perspective, principles and aims both within their discipline and within society at large.
Although not in each stage of theory formation and analysis, their work is admittedly and ultimately political. Their hope, if occasionally illusory, is change through critical understanding. Their perspective, if possible, that of those who suffer most from dominance and inequality’ (Van Dijk in Wetherell et al, 2001:383). Youth and community workers also take an explicit political stance which CDA therefore supports since it does not make unrealistic demands of neutrality. But this also raises issues concerning intent because when commissioned to undertake research on behalf of someone else (i.
e. employers) their intent and political stance needs to be examined carefully and considered in the light of anti-oppressive practice. This means that whatever method of research employed, the intent, political stance and values of researchers and sponsors will form an integral part of the process. This is not only important from an ethical point of view but also because they involve issues of validity and reliability. Willig describes validity as ‘the extent to which our research describes, measures or explains what it aims to describe, measure or explain’ (Willig, 2001:16).
Whilst Cohen et al, describe the variety of forms validity can take and they talk of the ‘honesty, depth, richness and scope of data achieved, the participants approached, the extent of triangulation and the disinterestedness or objectivity of the researcher’ (Cohen et al, 2001:105). For CDA validity is an important point but it can only be approached from within the research paradigm itself which can unfortunately lead one round in circles since ones own analysis is another construction (and not the only one) which can be used as data.Health and Social Research Methods
As Margaret Wetherell points out researchers can set up a ‘potentially infinite interpretative regress, analysing their own analysis and then analysing their own analysis of their own analysis to demonstrate the layers of construction’ (Wetherell et al, 2001:937). Critical discourse analysis, like much qualitative research, works with very small scale studies and it actively acknowledges the part the researcher plays in interpretation. This means that the disinterestedness or objectivity of the researcher is not an appropriate concept with which to look at the issue of validity in respect of CDA research.
In fact the acknowledgement of the intrinsic role played by the researcher and their influence on the data, is one of the measures of validity that CDA actually advocates. This coupled with a commitment to work with naturally occurring data as opposed to artificially created, laboratory controlled methods, an emphasis on the historical, cultural and social specificity of data, an awareness of the power, social, political and personal inherent in talk and text, all combine to produce a test of validity that is true to the principles of the constructionist paradigm.
These principles can also be used to form the basis of theoretical validity since any CDA analysis will be looking at and for, certain aspects of language use (i. e. the perpetuation of power and dominance). Internal validity in the case of critical discourse analysis, is concerned with the nature of the data itself and whether or not it is reasonable for the researcher to come to the conclusions they do based on the data to hand. In other words findings must be sustained by the data itself.
Discourse analysis focuses specifically on particular pieces of text and talk and illustrates its analysis by using the actual words of the participants. It should therefore be possible to measure accurately the internal validity of CDA. However, CDA cannot lay claim to external validity if this is understood as the ‘degree to which the results can be generalised to the wider population, cases or situations’ (Cohen et al, 2001:109). But the very paradigm upon which critical discourse analysis rests involves a generalised account of how power and dominance is created, sustained and maintained.Health and Social Research Methods
External validity then can be achieved by comparing similar text and talks, found in similar situations, analysed by researchers using the same theoretical framework. However, the very specificity of CDA makes this very difficult and Cohen et al (2001) advocate a detailed, in depth description of the research process so that others can decide whether or not the findings could be applied to other situations. For example, whilst investigating the use of CDA as a research method I had several conversations with another student who identified a particular line with the government’s new Connexions paper.
‘youth work interventions have demonstrated over many years the value that they can add to young people’s lives’ (Connexions Booklet, 2002:7). Having analysed the line, she had come to the conclusion that the word ‘can’ was being used strategically, i. e. as a criticism of youth work interventions because using the word ‘can’ as opposed to ‘does’ suggests that youth work interventions do not always demonstrate the value they add to young peoples’ lives.
Bearing in mind the specificity of the line i.e. within a document that was proposing a radical overhaul of existing youth work provision, this did not seem an unreasonable conclusion to reach. A couple of weeks later, I was invited by the Housing Committee of Charnwood Borough Council to take part in the development of their Homelessness Strategy which was being altered to fit the criteria now being set by the government. As I read through the existing document, I was surprised to find the same kind of linguistic device being use.
‘Homelessness can have a devastating effect on a person’s health, and on their economic, social and general well being’ (Folwell, 2001:1). The use of the term ‘can’ could again be seen to indicate that there were instances when homelessness did not have a devastating impact on a person’s health etc. One of the criteria for being statutorily homes and therefore entitled to social housing is being unintentionally homeless. If you are judged to have made yourself homeless by doing or not doing something, then you are not statutorily homeless and not entitled to housing.
The idea that homelessness does not always have a devastating effect on someone’s life, especially if it could be seen as their own fault, could provide and argument that would support this view and justify the withdrawal of services. I am happy to report that the introduction to the new document now contains the sentence that homelessness ‘does’ have a devastating impact on a person’s health etc. In this instance it is clear that although the researcher herself did not identify the extent to which her findings could be generalised, I, as another researcher, did.Health and Social Research MethodsHealth and Social Research Methods
Reliability is another problematic area for critical discourse analysis since research is said to be reliable if ‘reliability is a criterion applied to measurement and means that the tools or instruments being used can be relied upon to measure consistency’ (Wetherell et al, 2001:318). The measurement tool and/or instrument in the case of CDA is the researcher her/himself and the question would need to be changed to ‘if someone else did the research would he or she have got the same results and arrived at the same conclusion’ (Denscombe, 2002:213).
This is particularly difficult since there is an acceptance within the paradigm of CDA that there is always another understanding, another interpretation. Carla Willig raises the issue of reliability in the field of qualitative research in general and argues that ‘qualitative researchers are less concerned with reliability. This is because qualitative research explores a particularly, possibly unique, phenomenon or experience in great detail. It does not aim to measure a particular attribute in large numbers of people’ (Willig, 2001:17).
Denscombe offers a solution to this by advocating the use of clear ‘audit trail’ which should ‘be constructed and mapped out for the reader – allowing him or her to follow the path and key decisions taken by the researcher from conception of the research through to the findings and conclusions derived from the research’ (Denscombe, 2001:213). As with the issue of generalisability the answers proposed put the onus on other researchers to decide the matter for themselves and I am not convinced this settles the matter.Health and Social Research Methods
However, the usefulness or otherwise of all research is almost always left in the hands of others and an argument could be made that suggests that CDA is just being more honest than other forms of research. A number of researchers have advocated the use of ‘triangulation’ as a means of providing additional evidence or proof of reliability. In this way two or more research methods are used to analyse the same data and compare results. If the results or outcomes are similar, then research can claim reliability and validity.
In respect of critical discourse analysis however, triangulation can be challenged on the basis that it is positivistic instrument that assumes an ‘out there’ that can be separately measured because ‘the assumption that a single unit can always be measured more than once violates the interactionist principles of emergence, fluidity, uniqueness and specificity’ (Denzin in Cohen et al, 2001:115). Conclusion On the face of it critical discourse analysis appears to provide an ideal base from which youth and community workers committed to the principles of anti-oppressive practice can undertake research.
However, there are issues that cannot be overcome, concerning the power relationship between researcher and researched and special attention must be paid because ‘although the attractions of the qualitative paradigm include a more reflexive and potentially less exploitative attitude to the hierarchy of power between researcher and researched, experience has shown that the early optimism of feminist and other researchers about the democratising potential of ‘qualitative’ methods may be less justified than was at first hoped’ (Mayall et al, 1999:164).Health and Social Research Methods
However, the power relationship between researcher and research is a feature of all research methodologies and a sensitivity to, and understanding of, the issues involved can ensure that research is carried our in an ethical manner that seeks to reduce this inevitable element of research practice. However, of all research methods, CDA has an implicit understanding that power and dominance are features of peoples’ lives and has as its intent the exposure of the use of language in perpetuating these.
Thus it can be said to provide youth and community workers with an opportunity to further their understanding of how oppression is constructed whilst providing possibilities for action. Critical discourse analysis also provides the opportunity to challenge the covertly political aspects of oppression by concentrating its efforts on the institutional nature of oppression by using the written texts, policies, practices and procedures, they produce.
The analysis of political texts would likewise provide an opportunity to understand and challenge government hegemony at work in society which is being perpetuated through published reports and as such provides an ideal base from which youth and community work policy and practice can be developed.Health and Social Research Methods