Research method approach report

On the 23rd of June, 2016, Britons went to the polls in one of the most polarising and aggressive political events the country has seen in its centuries as a democracy. A tight, fiercely-fought political tussle that split typically solid partisan voter blocs with an in/out referendum on the UK membership of the EU. In the aftermath, the media of both sides was quick to lash out at reasons for the result, with ‘Remainers’ being comprised of the ‘liberal elite’ and the ‘Brexiteers’ composed of the anti-establishment and ‘uneducated’ vote. These stereotypes have lasted with some mainstream mass media outlets alleging that accusations of racism towards Brexiteers are still ongoing (Express, 2018). There is a lingering disillusion between Brexit and Remain voters, about the reasons behind voting one way or the other. This spreads across and between party lines, and between demographic lines. For instance, there is a clear trend in age where wards with older voters tended to vote leave and education was the most definitive bellweather of voter intention, with graduates and higher tending to vote remain (BBC, 2017).

That said, the question is still valid. Immigration was a hot topic for the media leading up to and after the referendum, with academic papers published by the CER (Centre for European Reform) claiming that: ‘If Britain votes to leave the EU it will be because of hostility to immigration.’ (CER, 2016). This rang true according to polls by Ipsos Mori – 33% of respondents listed Immigration as their number 1 issue (Ipsos Mori, 2016). It is clear that immigration a hot-button issue leading up to the referendum, but it remains to be seen just how responsible the mass media are in creating the discourse and bringing immigration to the fore. The concept for this question arises from the Noam Chomsky book ‘Manufacturing Consent’, in which (whilst it predominantly looks at America) it describes how monopolised media ownership can, in a co-ordinated fashion, ‘set the definition of what is news-worthy’ (Chomsky,1988). Under this theory of manufacturing consent, the report will set out both to assess whether xenophobia was a leading factor in the remain vote, and also whether the media’s role in the Brexit referendum was a significant factor in producing the xenophobia, or if there was an already-underlying problem in leave communities around this.

Because this was a referendum, not a General Election, the decision was not as predictable as votes typically are in Britain (not that they’re predictable with any certainty anyway). The vote crossed party lines and so typical factors like familial partisan alignment were not bellweathers of voter intention. This also rings true in newspapers, with politically aligned newspapers finding themselves with different messages to their readerships in some cases. Chomsky would posit that because a vast amount of the UK Newspaper media is monopolised, with several media barons having more power than most MP’s (London, 2013) that the media played a significant role in the outcome of the referendum, however this is an interesting angle to approach the psephology of Brexit from, since the entrenched political establishment did not get the result it seemed to desire. This report will discuss the methodology undertaken in pursuit of the above question, the theoretical framework used, the design of the research, and the ethics of said research.

Methodology:

Due to the open-ended definition of newspaper xenophobia (the research will have to define in advance to what extent a newspaper condones and promotes xenophobic content/views), the research will be Qualitative. It will analyse the subjective interpretation of xenophobia in Newspapers across a wide spectrum of sources. Quantitative analysis of xenophobic content in newspapers would be possible, however it would struggle to explain the impact it had, since the numerical data would not show how intentions changed with the newspapers setting the discourse. The epistemology of Qualitative analysis is post-positivistic, and since media and referenda are social constructs, the superior method of analysing the topic, being as it is limited in theoretical structure and therefore flexible. Quantitative analysis (which is better suited to empirical, scientific analysis) is much better suited to looking into natural events, whereas the flow of voter intention within the confines of the question is a socially constructed phenomenon. The conclusions that the research comes to would not (being analysed interpretively) be reproduceable in the same exact way that a Quantitative approach would allow and are subject to the logic processes of the researcher who conducts/reviews the data.

To validate the research, it must be conducted in as transparent a manner as possible, as well trying to remain objective. Whilst in interpretive research, bias cannot be completely avoided, the impact or implication of bias can be mitigated by establishing clear standard procedures and limits for the research to take. One criticism of Qualitative analysis that covers the above is Gidden’s papers suggesting the ‘Double Hermeneutic’ – the notion that the social sciences and commonly-held knowledge are intertwined and affect each other (Burnham, 2008). This can be applied to social science research, where the researcher and the research topic can alter each other – the researcher’s pre-conceived position can affect the interpreted conclusion, and the research can alter the researcher’s views. This leads into one of Gidden’s critical points on both epistemological approaches: ‘Hermeneutics and positivism are thus both partial philosophies, flawed by their incomplete character’ (Giddens, 1977). Both Positivist and Post-positivist approaches would yield results, however due to the subjective content of the question, Post-positivist, anti-foundationalist analysis would yield a higher quality of data. The reason for choosing anti-foundationalism as the ontological approach to the research is because the research will seek to analyse the extent of the media’s impact to voters - there being no one absolute truth about its impact (Kellstedt & Whitten, 2013) - which is subjective and may or may not have affected voters in the run-up. Whilst it is assumed that the media is acting within the confines of Chomsky’s theory of Manufacturing Consent (it wields a lot of influence over the common discourse of the nation), it is not assumed that the effect of the media was universally equal. This also fits into both Foucault and Gramsci’s belief that socialisation comes about through discourse rather than material structures. It isn’t the institution of the media that promotes and enforces normative behaviour, it’s the discourse that the media propagates (Marsden & Savigny, 2011). The research will involve numerous qualitative case studies, which will allow the generation of a hypothesis that will (as is the case with flexible post-positivist research) not necessarily be the one outlined in this report. This is a benefit of anti-fundamentalist, post-positivist research, because it allows a truth of the question to be more closely sought, however unreproducible.

The research project will have a small sample size due to its qualitative nature, and the sampling technique will be a form of stratified semi-random sampling. Creswell points out that the best qualitative information is garnered when the participants of a focus group agree with each other (Creswell, 2009). As such, whilst the sampling process will be random, it will be randomly picked from a specific stratum – Brexit voters. This will ensure transparency in the research process, lessening any chance of cherry-picked data which with a 2-way Hermeneutic scenario is very possible. Additionally, it will allow data collected to be stratified amongst several differing factors: age, wealth, employment, education etc. This will allow a fuller picture of not only the medias effect in different geographical locations, but also between the differing demographics of the samples. The samples will consist of people who voted for Brexit and will be taken from 6 locations: The ward that voted most overly for Remain, the ward that voted most for Brexit, the 2 most closely-contested wards, and the 2 that fall at the median between the 2 most closely-contested wards and the most Remain/Brexit. This ensures a cross-section of society geographically and the differing results will provide an insight into nation-wide trends, not regional trends where wards that voted Brexit were subject to the same socio-economic factors. After stratification, random sampling will be used to again, maintain transparency and validity.

The research will consist of focus groups of 5 people in each ward, invited again through semi-random sampling and asked to discuss a series of structured questions, but with discussion allowed the flourish. The interviewer will merely be an observer, keeping the discussion topical but otherwise completely passive. This will prevent the researcher’s personality from causing erroneous results, since people are likely to react differently to an interviewer dominating a focus group. Qualitative discourse analysis will be more valid without the presence of an authority figure in the room. Recorded and transcribed, the data will still be valid, and negates any power asymmetry, is an unknown variable in the research (Kvale, 2006).

A variety of narratives will emerge from the focus groups, from which some norms across demographics could emerge. Analysing this, Foucault posits that discourse is the means through which discussion frames an object or an event. Baker-Beall shows that a constructed discourse projects a ‘conventional wisdom’ (Baker-Beall, 2016) which will reveal itself through repeated statements in the narrative that convey societal norms regarding the reasoning behind Brexit.

For the discourse analysis, the texts to examine will be the transcripts from the focus groups that are hosted, and then the discourse will be mapped, identifying the conventional wisdom of the sample by identifying commonly repeated words, phrases and themes in the text. This will provide the reasoning behind the various groups’ vote. Looking to see if the discourse fits within a neo-liberal perspective of IR, the discourse ought to indicate that the inherent structural power of the media in a system of complex interdependence has countered the power of a political establishment using xenophobia, or whether it was other socio-economic factors that contributed more significantly to the vote. By asking people and taking active notes on participants age, race, economic background and level of education, I can also get a cross-section of how differing demographics reacted to the mass media. This will allow me to analyse with more depth than just researching the question from a quantitative, arguably shallow angle.

Theoretical Framework

The epistemological position that this research will be conducted from is from a critical constructivist position. The structures that give political actors power are not limited to those of ‘hard’ power, but economic gain and political influence are in the hands of artificially constructed and monopolised businesses and the mass media, which can (as the research may show) display. Human nature cannot explain the result of the referendum – being as it is a socially constructed event. This discounts Realism and Neo-Realism, and Structuralist theory, but suggests that that socio-economic factors in a commercialised world are what led to xenophobia in the media being emphasised. Additionally, the research will demonstrate how different societal constructions such as gender and economic status interact with other constructs and since Constructivism is a critical theory, the research’s theoretical framework is geared towards finding a solution to the systemic problems. (Marsden & Savigny, 2011).

Additionally, since the research will be spread out geopolitically and across the spectrum of Brexit – it is unreasonable to assume that the conventional wisdom is universally shared amongst Brexit voters, since as described in the abstract, the Brexit vote did not follow usual voter blocs in the UK. The cultural experience of, for instance, a north-eastern working class Brexiteer will be different than that of a city banker in the City of London. In this way, anti-functionalism dictates that there is no one singular truth, and so post-positivist research methods prevail.

Research Design

The purpose of the research is to understand the extent to which the mass media affected the Brexit vote with xenophobic discourse creation, and how this fits into the theory of social constructivism. The research will use primary sources and will seek to link the discourse markers from those sources to newspapers in the month leading up to the election, searching for the extent of the media’s involvement.

The critique of overly-localised generalisation is mitigated by the use of multiple focus groups in various diverse locations, allowing the research to be applied to the population as a whole. Sampling not fully at random to ensure a more useful set of data from the focus groups, the discussions will be recorded and transcribed, then subject to discourse analysis whereupon normative discourse key words relating to Xenophobia will be discovered. The research will ensure validity through the following; face validity (wherein a piece of research upon cursory glance appears to be researching for the question it is aiming to answer. Content validity is ensured through the removal of a dominant or imposing researcher guiding discussion, and by only collecting this data from Brexiteers. Finally, the construct validity of the research is defined as ‘the degree to which the measure is related to other measures that theory requires them to be related to.’ (Kellstadt & Whitten, 2013). The data collected will also satisfy Lincoln and Guba due to the method in which reliability is ensured (Bryman, 2004,): Credibility will be ensured through good research practices, Transferability will be catered for by the data being taken from multiple subsets. Dependability will be taken into account when the report and all documents are written, in a clear manner such that another researcher can follow the logic of the research and come to a similar conclusion. An attempt to remain objective will be accomplished through the removal of dominant interviewer so as not to inadvertently sway the results. Complete objectivity in such a polarising issue will be very difficult but must be done to satisfy the four criteria of a reliable study.

Ethical Considerations

The research will be conducted with responsible collection and analysis of unbiased data – the discourse analysis, whilst subject to logical interpretations under the theoretical framework, is to be done away from participants and is not to be cherry-picked or forced into a theory. Being qualitative, the theory can change to fit the data.

Participants will be briefed, in full, on the details of the project, and every care must be taken to ensure that all participants are willing, sane and in full control of their participation, free to withdraw their consent to research at any time. Participation is entirely voluntary and based on informed consent, wherein the participants are appraised of the research parameters and give their consent to be researched upon (Callan & Harrison, 2013). Written consent must be obtained, and additional clauses of that will involve the viewing of transcripts prior to publication and contact information for the researcher, given any questions or queries the participants may have. Being such a polarising subject, the focus group could get heated or personal, the observing researcher will have a duty of care to the participants and will ensure that no discomfort is caused to participants, either by the researcher’s own doing, or that of another participant. All information will be held under the Data Protection Act 1998, and permission to run the focus groups will be subject to review by the university ethics committee.

Bibliography:

· Baker-Beall, C. (2016). The European Union's fight against terrorism. Manchester: Manchester University Press, p.16.

· Bosotti, A. (2018). 'Nothing to do with colour!' Vince Cable grilled over 'racist Brexiteers' accusations. [online] Express.co.uk. Available at: https://www.express.co.uk/news/uk/930678/vince-cable-brexit-news-european-union-latest [Accessed 17 Mar. 2018].

· Burnham, P. (2008). Research Methods in Politics. Warwick: Palgrave Macmillan.

· Giddens, A. (1977). Review: Habermas's Social and Political Theory. American Journal of Sociology, [online] 83(1), pp.198-212. Available at: http://www.jstor.org/stable/2777774 [Accessed 15 Mar. 2018].

· Harrison, L. and Munn, J. (2013). Key research concepts in politics & international relations. 1st ed. London: Sage.

· Herman, E. and Chomsky, N. (1994). Manufacturing consent. 1st ed. New York: Vintage.

· Kellstedt, P. and Whitten, G. (2013). The Fundamentals of Political Science Research, 2nd Edition. 2nd ed. New Yourk: Cambridge University Press, p.102.

· Kvale, S. (2006). Dominance Through Interviews and Dialogues. Qualitative Inquiry, 12(3), pp.480-500.

· London, T. (2013). Everyone should know who owns the press – for the sake of our democracy. [online] Left Foot Forward. Available at: https://leftfootforward.org/2013/06/everyone-should-know-who-owns-the-press-for-the-sake-of-our-democracy/ [Accessed 12 Mar. 2018].

· Pierre, J. and Peters, B. (2000). Governance, politics, and the state. New York: St. Martin's Press, p.60.

· Rosenbaum, M. (2017). Voting data sheds light on referendum. [online] BBC News. Available at: http://www.bbc.co.uk/news/uk-politics-38762034 [Accessed 15 Mar. 2018].

· Savigny, H. and Marsden, L. (2011). Doing political science and international relations. Houndmills: Palgrave Macmillan.

· Skinner, G. (2016). Immigration is now the top issue for voters in the EU referendum. [online] Ipsos MORI. Available at: https://www.ipsos.com/ipsos-mori/en-uk/immigration-now-top-issue-voters-eu-referendum [Accessed 21 Mar. 2018].

· Tilford, S. (2016). Britain, Immigration and Brexit. Centre for European Reform, [online] (105), pp.1-2. Available at: https://www.cer.eu/sites/default/files/bulletin_105_st_article1.pdf [Accessed 18 Mar. 2018].