One of the blogs I follow on a regular basis is Jeffrey Keefer’s Silence and Voice which is currently concerned with some issues related to research design and formulating of research questions in the wider context of auto-ethnography as methodology and identity construction as the subject of interest. Recent posts I found very interesting and commented on are the one on broader research design questions and the one on auto-ethnography and reflexivity which are worthwhile having a think over – they provide very good ground for some reflection on dilemmas and politics entailed in the underlying epistemological questions of research-related decision-making.
I am fascinated by the way crowd-sourcing can work in academic blogging, it’s a great way of engaging broader audiences and gaining some insights from outside immediate areas. I like and do value the fact that academic bloggers, busy with studies and work, research and other things, take the time and effort to reflect publicly online, open up to questions and critiques – and respond to comments and ideas. Personally, I enjoy the challenge to think about issues and see whether I can contribute some ideas and to what extent I need to improve on gaps and communication of insights.
Jeffrey’s research made me recall some podcasts featuring Stuart Hall et al. debating questions related to identity construction. They form part of the Open University’s post-graduate course D853 Identity in Question which I studied in 2008 – the interviews cover Lacanian Theory, Language Approach, subjectivity and legal definitions of personhood as well as some comments on Michel Foucault’s genealogical perspective. They last 2 to 9 minutes and deliver some great food for thought. I also came across James Schirmer’s paper on Scribd which I recommend as thought-provoking read in this context The Personal as Public: Identity Construction/Fragmentation Online
This year’s annual conference of the Association of Internet Researchers (AoIR) Internet Research 10.0– Internet: Critical will be held 7-10 October 2009 in Milwaukee, WI, USA. I will be attending the preconference workshop on Multidisciplinary Internet Research which participants were asked to prepare for. The preparation covered a list of [early-stage] research questions, theoretical and methodological frameworks and key literature drawn upon in the reflection on interdisciplinary and multidisciplinary research design.
The workshop organisers have set up an already quite comprehensive wiki which is available on sociotech.net and contains my summary that is also available on Slideshare where you will be able to find a transcript of the 2-pages PDF. The wiki will be updated in due course, so keep watching if that field interests you.
This belongs to the revision of social research strategies, I am going to summarise the key differences between inductive and deductive research approaches – but first what they’ve got in common. Both strategies are rooted in a positivist assumption in terms of epistemology and ontology. The underlying empiricism, i.e. the notion that only knowledge gained through experiences and senses is acceptable, is implemented by rigorous testing. Enlarging the number of instances observed (samples) increases plausibility and the number of regularities being identified. The accumulated ‘facts’ provide basis for general laws of cause and effects. Those are depicted in models as dependent (predictor) and independent (outcome) variables.
Inductive theory is being derived from the observations made. This approach cannot test hypotheses but generates them. In contrast, deduction is theory-driven, it’s based on preconceptions and aims to overcome the limitations of induction. It puts theories to the test, that means hypotheses can be falsified and disproved. The aim is to move closer to the truth, hence the gradual elimination of false theories implies that theories tested and not disproved can only be considered provisional.
Ideally, a deductive approach starts with a theoretical framework (for instance based on Erving Goffman’s ‘stigma’ or Pierre Bourdieu’s ‘social capital’) and the formulation of hypotheses. Usually, this includes an alternative hypothesis (also called experimental H., which states the effect assumed) and the null hypothesis (which states the effect is absent). What follows is the data collection which delivers findings that either result in confirmation or rejection of the null hypothesis and a subsequent revision of the theory.
In practice, though, deduction often entails an element of induction and vice versa. This is rooted in theoretical reflection once the data has been collected or the desire to establish conditions which allow the theory to hold (or not). This continuous weaving back and forth between data and theory and is called an iterative strategy, particularly evident in qualitative research which takes a grounded theory approach and a way to add to the validity of research. In quantitative research, it is advisable to carefully distinguish between the more complex development of theory and the generalisation of empirical findings.
Ontological and epistemological positions provide fundamental aspects of research as they concern the philosophical questions what counts as reality and how beings come into being as well as what constitutes knowledge and how knowledge comes to be established. Two core positions can be distinguished in either area: positivist and constructionist.
- positivist ontology: the world is ‘out there’, it operates in a systematic and lawful manner, discrete and observable events, reality is separate from human meaning-making;
- constructionist ontology: assumes the world we can study is a semiotic world of meanings, represented in signs and symbols, language is central to this position;
- positivist epistemology: knowledge can only be gained by gathering facts in a systematic and objective manner, predominantly by the experimental method and by testing of hypotheses in order to gradually build laws. The aim is to refine them and achieve applicability on a universal level;
- constructionist epistemology: knowledge is constructed rather than discovered, it is a representation of the ‘real world’ and interpreted by the researcher. Knowledge is subject to time-space configurations and a means of power (e.g. doctors as ‘architects of medical knowledge’). Scientists and their institutions shape the production of knowledge by their choices and values.
These positions significantly shape research designs and methodologies.
The Virtual Knowledge Studio in Amsterdam offered an Ethics of (e)research Workshop on
Monday 15 June, which brought together post-graduate/doctoral students and researchers from various fields and a range of cultural backgrounds.
Below are the ethical dilemmas I anticipate to encounter in due course of my future research project which will investigate Digital Technologies [as research tool and objects] in the context of informal cognitive processes embedded in online social interaction which have repercussions on real world settings and experiences.
Methodology: based on a triangulatory approach, it will include a self-completed online survey, auto-ethnographic work as well as semi-structured focus-group interviews and content analysis. Inevitably, in particular the auto-ethnographic work conducted in the blogosphere and online social networks entails a range of possible ethical conflicts which I fear an Ethics Committee may subject to a one-size-fits-all policy that won’t take into account the following particularities:
- Participant Consent –fully informed and voluntary (FIV) – in retrospective?
Conversations and comments on blogs, tweets and retweets on Twitter and comments on Facebook status updates or semi-public debates via Facebook’s wall-to-wall feature: they cut across the public/private boundaries. Given that participants provide FIV consent, Ethics Committees should accept this as ethical research. However, danah boyd et al. have experienced considerable difficulties with retrospective consent in recent projects. Hence, a more ‘dynamic’ and contextualised/non-static model of ethical guidelines is still something we cannot take for granted when submitting our forms to the Ethics Committees.
- Public versus private, blurred boundaries and imagined risky/secure spaces
Are Facebook status updates private, semi-public or public? If forwarded by applications that support Twitter boundaries become blurred and even participants may differ in their perceptions, resulting in different participants demanding different levels of privacy (at different stages in the research) – or, maybe also requesting to categorise rather private messages as public for they may want to be heard and gain higher ‘online status’ (for instance on QDOS which calculate your virtual footprint). Imposing privacy might indeed cause harm when participants do rather desire publicity.
- Confidentiality and Anonymity
Are aspects closely related to point 2 above. Can we safely assume all participants desire anonymisation of their real name or pseudonym? How can I deal with texts/images and other media that evolve over time and contain various levels of confidentiality, for instance participant comments in 1:1 conversation (think Twitter DM [direct messaging], forwarded automatically to email, responded to by public tweet) and also in focus group follow-up interviews. I.e. naturalistic research in the first case vs. participatory research in the latter.
Moreover: cultural differences, expectations and needs may vary across age groups, perhaps even gender, and depend on social class background/educational level. Ideally, we are giving a voice to the interviewee/participant and promoting a level of equality, i.e. avoiding misrepresentation, paternalistic attitude and harm by all means – yet, we need to understand that positions are highly contextual and depend on subjective needs of participants rather than universal model of research ethics. Have ethics committes already arrived at that point?
- Power and Equality
Conducting research, collecting data and distributing findings may be greatly facilitated by online channels. The level of transparency can be high, and research participants may want to claim part-authorship for instance by using excerpts of the research report to be posted on their blog or website (or used in other media). A continued dialogue with participants, post-debriefing, may require further ethical decision-making beyond the levels common in other contexts. Again, not a one-size-fits-all ethical guideline but rather a case-by-case-based ethical decision making might be required. Will participants become involved in future amendments of ethical guidelines?
Finally, the participants’ levels of reflexivity and general awareness of research processes do seem to increase continuously while access to paths of personal and professional development of the researchers become ever more transparent and accessible. Will we need to learn to remind participants that they also need to behave ethically towards researchers? Are we progressing towards a more equal research-driven community and wider – globalised – society? Moreover, the researcher as the researched: my blog, my SNSs, my microblog, all the many profiles, traces left – can I expect research participants to act in an ethical manner in case they won’t agree with my findings (interpretations of findings, to be precise)?