Research In the Social Sciences

Introduction

The social sciences branch out into different fields ranging from economics to anthropology, with each respective field employing different research models and methodologies varying in degree of specificity. Strengths and weaknesses of methodologies, as well as characteristics of research findings differ depending on this degree of specificity and generality.

Moving towards a higher degree of generality equips us with broader perspectives, but also results in us compromising on accuracy and sacrificing the detail that comes with specificity. Though trade-offs exist between generalization and specificity, they do not negate each other’s individual values and merits, and both remain equally important. Hence, it is imperative that we are aware of the limitations and trade-offs associated with research methods adopted (based on their degree of specificity), so that we can take measures to combat these limitations and minimize the trade-offs, and conduct research in a balanced way that would work best in line with our goals in varied contexts, whether in policy planning or academia.

One trade-off that arises is largely related to efficiency. As social science research gets increasingly specific, it becomes time-consuming to focus on individual cases. (Neill, 2007) Also, going into excessive detail requires large amounts of resources, and may not be worth the cost depending on what the research findings can ultimately offer. On the other hand, if we over-simplify and generalize across cases, we may overlook details crucial for policy implementation success, and inefficient outcomes can still arise.

Conventional ethnography is a highly precise, qualitative research process involving long term field visits, participant roles and time extensity to gather insider information about the way of life and values in the field of study. (Knoblauch, 2005) As some believe that an observer is unable to be neutral and objective or operate outside their own value systems and assumptions, conventional ethnography is considered inefficient to some, since this experimentally intensive research could end up producing highly contextualised and subjective findings that are inaccurate and unreliable. (Laragy, 2006) Furthermore, findings may be exclusive to that specific area of research, limiting its application across the larger population.

In Rita Astuti’s case study (Astuti, 1999), ethnographic findings of one Vezo woman’s day-to-day life offer in-depth, intimate knowledge about her motivations and values on a personal level that the application of general research methods might have missed out on. However, such knowledge is limited in providing a broader understanding of overall trends across the whole diverse Vezo population, and the study of just one woman’s activities might be too small a sample size to generalize findings across the entire population.

Lack of awareness of such a trade-off can lead to the implementation of myopic policies. In Madagascar, a tension exists between the conservationist and the Malagasy ethos. The Malagasy ethos is founded on having an ever-growing net of descendants, which, based on the conservationist ethos, is overpopulation and a threat to achieving equilibrium among different species on the planet. The politically, and immensely more powerful conservationist ethos works towards limiting the unsustainable growth of the Malagasy, to create a broad equilibrium among all the different species in the Masoala peninsula. (Keller, 2008) However, simply adopting this general stance without assigning value to the unique nature of Malagasy kinship, can lead to policy resistance and serious obstacles in policy implementation. Brute-forcing conservation policies could also lead to the devastation of local culture. By recognising what Keller’s specific ethnographic research unearths about the Malagasy’s views of conservationist policies, measures like educating local farmers about the necessity of having fewer children to safeguard their future becomes essential. Thus, what one might criticise as ‘time-consuming/ inefficient’ and ‘overly-specific’ research might in actual fact help craft policy solutions that address contextual problems, avoiding ineffectual and inefficient one-dimensional policies, and ultimately, it depends on what the country prioritises.

Human understanding also constitutes part of the tension between generality and specificity. In social science research, general measures and data are more comprehensible and comparable than specific, complex ones that could require specialized knowledge.

An implication of this is that measures should be chosen to suit goals for which the data measured is being used to achieve. For example, for poverty measures, the UN adopts the simple US$1-a-day indicator. This universal measure is useful in raising awareness and generating political momentum, as it aggregates and standardises the picture of poverty at an international level and can mobilise mass global effort. (Maxwell, 1999) This indicator points to symptoms of poverty but omits other aspects, like the social side of poverty, and might thus be too thin a measure. The Human Poverty Index (HPI) however, is a measure that delves into the comprehensive specifics of relative poverty. (UNDP HDR, 2011) It is sensitive to local perceptions of poverty in different areas, and addresses the causes of poverty more directly, though the dimensions involved are more subjective and hence complex to consider. Nevertheless, the UN’s choice of ‘US$1-a-day’ as a headline measure can be attributed to how it works best in helping the UN achieve a realistic goal. A thick, qualitative measure like the HPI, would involve more complications like how much weight to assign to different components of the index and would perhaps be more useful for analysis and detailed planning of tackling diverse poverty causes in particular regions.

Another trade-off between generality and specificity is predictability. Generalization allows for greater predictability, as the overarching model formed based upon recurring trends and past events enable transferability of understandings to novel situations. (Shiveley, 2009) Generalization also allows one to make and test bold statements in different contexts, spurring new findings on how applicable the theory is in predicting certain events. (Eyler, 2010)

However, generalization may fall apart with an unexpected event, leading to vastly different outcomes. To prevent hasty generalizations, one could work exclusively around individual cases. However, doing so does not guarantee that in repeating the research methodology, one will get back the exact same outcome (Lopez-Trueba, 2008), since firstly, the findings do not cover many cases, and secondly, human behaviour is non-static and inherently unpredictable. (Morse, 2002)

A study was conducted to investigate the correlation between ecological change variables and civil war incidences in Africa. (M. Burke et al, 2009) Results showed that there was a statistically significant correlation between regional temperature changes and civil war incidences, other variables controlled. Further studies of causal chains also reflect how drought had intensified underlying tensions and migration patterns, leading to a lack of social trust and civil strife. (Luc Bowens, 2011) Generalising across cases would lead to the explanation that ecological changes led to civil strife. Yet, though not incorrect, it is simplistic to do so, since other direct chains leading to civil strife of different origins completely unrelated to climate change exist. (De Waal, A, 2007)

In this respect, we see that generalization and specificity both have drawbacks in terms of their use for prediction, and determining the degree of generality to employ depends on the context. Generalization would work better in an environment with a record of recurring trends, whilst in examining small samples and random occurrences, it might be better to focus on specific cases rather than make broad assumptions that do not constantly hold true.

Often, the nature of data in social science complicates predictability due to the fact that human agency behaviour varies. (Leung, 2011) When people know about predictions, they change their behaviours and responses. (Goodhart, 1975) Hence, the implication is that a good method would use a combination of methods that cover all grounds (from general to specific) to try to predict outcomes more accurately.

Finally, there are pressures of academic specialization in social scientific fields. (Cox, 2011) Expertise knowledge offers detailed and useful insights exclusive to one’s own particular field, which can in turn generate further spin-offs in developing new ideas and knowledge to explain the causes of things. However, just solely concentrating on one’s own discipline would result in a narrow tunnel vision of events, causing one to lose out in breadth and scope of understanding.

For example, various disciplines explain the end of the Cold War differently. Historians emphasise on multiple actors and reject a simple bipolarity model. They explore specific details of all possible reasons that could have ended the Cold War, from Gorbachev’s ‘heresthetics abilities’, the weakening of communism’s ideological appeal, to the European integration process. However, this gives rise to the danger of being involved in too much detail, and excess complexity could break a simple issue up into too many small and intricate parts. Historians could have 20-25 different discoveries and competing explanations for how the Cold War ended, but still fundamentally lose sight of the bigger, overall picture. (Ludlow, 2011)

Hence, communication between different specializations is essential to pull different reasons together for a more complete overview, so that explanations of causes would not be one-sided and skewed towards one particular niche field. It is also ideal to target more inter-disciplinary methods of research, since events in social sciences are usually caused by a combination and contingency of factors from different disciplines rather than one specific discipline. (Cox, 2011)

In conclusion, though trade-offs exist, we cannot solely focus on one methodology of research since both generalization and specialization have their individual merits and are important. Ultimately, how we conduct research boils down to what we prioritise or what works best in a given context, since some methodologies work better in certain environments than others. The social sciences are a dynamic and ever-changing system of interactions and interrelations existing as a framework of the whole, not as a sum total of the separate and individual. (Anguelov, 1984) This essential unity of social phenomena results in the need for integration of general and specific methodologies, since we can sometimes only generalize after obtaining specific data, or only narrow into specifics with the backing of generalizations that set the strategic direction of where and what the focus should be.

References:

1. Anguelov, Stephan (1984). Discussion: Methodological problems of the social sciences, Studies in East European Thought Volume 27, Number 3, 263-265, DOI: 10.1007/BF00831907 http://www.springerlink.com/content/g57p5006h7304l47/ Retrieved on: 19 March 2011
2. Astuti, Rita (1999). At the centre of the market: a Vezo woman. In Day, S, Papataxiarchis, E. and Stewart, M (Eds) Lilies of the field. Boulder, Co: Westview Press (4) 83-95
3. Burke, M.B. et. Al. (2009). Warming increases the risk of civil war in Africa, PNAS 106(49), 20670-20674. http://gate2.library.lse.ac.uk/login?url=http://dx.doi.org/10.1073/pnas.0907998106
4. Cox, Michael (2011). Why did we fail to predict the end of the Cold WarLSE100 Module 3 Lecture 3 capture. Retrieved from LSE100 Moodle site. http://moodle.lse.ac.uk
5. De Waal, A. (2007). Is Climate Change the culprit for Dafurhttp://blogs.ssrc.org/sudan/2007/06/25/is-climate-change-the-culprit-for-darfur/
6. Eyler, Aaron (2010). The importance of generalizations in social studies. http://researchtopractice.wordpress.com/2010/04/25/the-importance-of-generalizations-in-social-studies/ Retrieved: 19 March 2011
7. Goodhart, C.A.E. (1975). Monetary Relationships: A View from Threadneedle Street. Papers in Monetary Economics (Reserve Bank of Australia) I.
8. Keller, E. (2008). The banna plant and the moon: conservation and the Malagasy ethos of life in Masoala, Madagascar. American Ethnologist, (35). 651-654, 656-659, 661-662
9. Knoblauch, Hubert (2005). Focused Ethnography [30 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research 6(3), Art. 44, http://nbn-resolving.de/urn:nbn:de:0114-fqs0503440. Retrieved: 19 March 2011
10. Laragy, Elizabeth (2006). The Imperial Archive: Key Concepts in Post-Colonial Studies: Ethnography http://www.qub.ac.uk/schools/SchoolofEnglish/imperial/key-concepts/Ethnography.htm#_ftnref4 Retrieved: 19 March 2011
11. Leung, Tim (2011). Asking your own questions- and asking the right questions LSE100 Special Lecture 3 capture. Retrieved from LSE100 Moodle site. http://moodle.lse.ac.uk
12. Lopez-Trueba, Mei (2008). Reliability, Validity, and Generalisability. Are they compatible in ethnographic research focused on Occupational Health(2). http://www.scribd.com/doc/4637204/Reliability-Validity-and-Generalisability Mei Lopez-Trueba Retrieved : 19 March 2011
13. Luc Bowens (2011). LSE100 Understanding the causes of things: A philosophy of science perspective on causality in the social sciences 1-5
    
14. Ludlow, Piers (2011) The end of the Cold War, An historian’s approach. Module 3 Lecture 2 capture. Retrieved from LSE100 Moodle site. http://moodle.lse.ac.uk
15. Maxwell, Simon. (1999). ODI Poverty Briefing: The Meaning and Measurement of Poverty Overseas Development Institute (3) 1-2
16. Morse, J. M. and Richards, L. (2002). Readme First for a User’s Guide to Qualitative Methods. Thousand Oaks, CA: Sage. A basic text and guide to the qualitative methods literature. (4)
17.  Neill, James (2007). Qualitative versus Quantitative Research: Key Points in a Classic Debate.
    http://wilderdom.com/research/QualitativeVersusQuantitativeResearch.html Retrieved on 19 March 2011
18. UNDP (2011). Human Development Reports : Human Poverty Index http://hdr.undp.org/en/statistics/indices/hpi/ Retrieved: 20 March 2011
19. Shiveley, J. M., & Misco, T. (2009). Reclaiming generalizations in social studies education. Social Science Research & Practice, 4 (2), 73-78. http://www.socstrp.org/issues/PDF/4.2.6.pdf Retrieved: 22 March 2011