Lecture 2 Research and Research Strategies 1.Social Science Method: Concepts, Theories and the ‘Circuit of Knowledge’ 2. 2.Research Strategies ¨ ¨ > Research Strategies nWhen starting out research we need to decide on some kind of strategy or overall approach nIn social research a range of approaches are open to us. The choice of strategy depends on such things as: ¨Nature of the object being studied ¨Purpose of enquiry and questions we want answering ¨Practicality, cost and time ¨Methodological, political and ethical stances ¨ ¨ > Research Strategies nResearch strategies are overall approaches to research. They include experiments, surveys, case studies, action research & ethnography. nStrategies do not necessarily determine the methods used. nThat said, strategies tend to be linked with different paradigms: ¨ Experiments/surveys (naturalist/scientific) ¨ Case studies – (anti-naturalist/ethnographic) ¨Action research- (critical theory) ¨Such approaches may well favour certain methods ¨ > Experiments/Quasi experiments nAim to test theories and identify cause nClassic ‘science’ model - control, observation/measurement, replication nRequires a closed system - to establish internal validity nUses treatment and control/comparison groups chosen by matched and random sampling. nTests pre and post intervention or treatment to ascertain change. n > Controls include Eliminating other causal variables – the classic laboratory experiment Eliminating the causal variable itself (e.g. removing fats from diet to observe effects on health) Holding factors constant – typical in social science where its practically impossible to create closed systems e.g gender, educational background, class, ethnicity can all be controlled e.g. by say selecting a group of people of similar age and income Random allocation of groups in research. The theory here is that if individuals involved in the study are randomly chosen then not relevant factors should cancel themselves out. Control groups. The most common way of exercising control. Works by selecting matched groups/sample. The factor is introduced to one group but not the other. The key is to have both groups matched at the outset. A pre factor measure is take of both groups and a post factor measure. In theory any change noticed should be due to the factor. Experiment/Quasi experiments nClosed systems impossible to produce in social world. Most social science settings use statistical control & adopt logic of the experimental method nThreats to internal validity - history, maturation, statistical regression, selection bias nThreats to external validity – Hawthorne effect, representativeness, operationalisation problems, context specific nature of behaviour nThese problems reflect bigger issues e.g. nature of human behaviour, models of cause, nature of measurement > In education tend to lack control over when events take place and for whom Surveys nSurveys aimed at breadth, provide a snapshot, rely on empirical data – to survey is to look nCan be used longitudinally nTend to rely on sampling, measurement, statistics and quantitative analysis nSurveys include postal questionnaires, face to face interviews, telephone interviews, observation, documentary surveys > Surveys nRepresentativeness is crucial if surveys are to be generalised – hence need for random selection nOne off snapshot may tells us little about processes nCan tend toward empiricism – view that measures or data simply reflects world out there. Raises complex issue of operationalising concepts and measurement nAre detail, depth and contextual understanding sacrificed in search for breadth? nRely on correlations as cause but does this capture cause? n > Case Studies nA case study is a specific instance of the thing to be investigated – person, an organisation, a country. nTends to be descriptive and analytical – exploring specificity in depth - can be used to illustrate a general theory/principle or to test/generate theory nOccurs in ‘natural settings’. Focuses on processes and relationships. Often uses multiple sources and methods nCases selected for suitability – typicality, extreme instance, test for theory, least likely instance, availability > Case studies nHow reliable and valid are case studies? nCan case studies be generalised? How representative are cases? nHow do case studies link to wider processes e.g. effects of new health policy in a hospital – attitude of staff, style of implementation, etc? nWhat should count as a case? The activities/processes in ‘bound’ areas like hospitals extend beyond the hospital (e.g. government policy impacts) > Ethnography nEssentially the study of peoples, groups, cultures nAim is to authentically capture the way in which those studied see and construct the world – usually this means joining the group/setting nMeanings and social processes are crucial - capturing the world as it is seen, experienced and actually happens - understanding is the goal nit is holistic, studying people, events and processes in context – direct observation and participation are its empirical base n > Ethnography nIts very difficult to do nWhat effect does research have on setting – reactivity and going native nDo ethnographers describe the real world or construct it. Is ethnography story telling? nAre ethnographies unique? Can we generalise? How reliable and valid are they nCan ethnographies be linked to theory and used to explain? > Action Research nAttempts to bridge a gap between theory/practice so research impacts on/improves practice nResearch is not ‘done’ on others but involves researching ones own work - it is participatory nAR is collaborative involving those responsible for action in improving it nEstablishes self-critical communities participating and collaborating in all phases of research nSystematic learning process - ideas, assumptions, practices tested by gathering evidence in practice – plan, act, observe reflect n > Cohen and Manion define action research as ‘a small scale intervention in the functioning of the real world and a close examination of the effects of such an intervention (1994 p 186) Action Research nCan you fulfil two roles at same time? Clash of interest – does AR produce self serving results? nResearcher as practitioner limits scale and just how much research findings can be generalised nHow rigorous is the research process? nIs it a political or research strategy? nIs there a tension between AR as improving practice, AR as critical/emancipatory > Ethics nArise at all stages of research nChoice of subject, purpose, practice, outcomes, use nEach strategy poses generic and specific ethical issues nIssues to consider: ¨ deception ¨ trust ¨ confidentiality/anonymity, §privacy, §role, §Purpose > n PURPOSES PURPOSES QUESTIONS LITERATURE REVIEW APPROACH DESIGN FRAME, METHODS AND ANALYSIS > n“There are known knowns. There are things that we “ know we know. There are known unknowns. That is to say, there are things that we now know we don’t know. So when we do the best we can and we pull all this information together and, and we then say well that’s basically what we see as the situation that is really only the known knowns and the known unknowns, and each year we discover a few more of those unknown unknowns and …. and I, I … it, it sounds like a riddle … “ n (Donald Rumsfeld, U.S. Secretary for Defense, United States of America Department of Defense news briefing, 12 February 2002) >