Social science research dictionary

sage dictionary of social research methods and dictionary of social research methods
Dr.DavidWllker Profile Pic
Dr.DavidWllker,United States,Professional
Published Date:11-07-2017
Your Website URL(Optional)
Comment
Jupp-Prelims.qxd 3/8/2006 10:37 AM Page xi Editor’s Introduction Organizing themes Although presented in alphabetical order, the concepts covered in this Dictionary were selected on the basis of several key themes which are embraced within the term ‘social research’. These are: (1) Philosophy of science, for example issues of ontology (what is the essential nature of reality?) and epistemology (whether or how we can gain knowledge of that reality). (2) Research paradigms, for example positivism (which in general terms is taken to include the scientific study of some objective social reality) and construc- tionism (which is concerned with the study of ways in which the social world is constructed through social interactions). (3) Research designs, for example the experiment (the attribution of outcomes to the controlled administration of a ‘treatment’ to one group and not another) and social survey (the systematic collection of data from or about units of analysis, usually individuals, often using sampling techniques). (4) Specific aspects of data collection, for example participant observation (partic- ipating in a group in a covert manner in order to study that group) and spe- cific aspects of data analysis, for example multivariate analysis (a set of statistical techniques to examine the relationships between several variables). (5) Issues to be addressed when carrying out research, for example ethics (what standards should be adopted, say in relation to obtaining informed consent from subjects?) and politics and research (the extent to which research is con- tributing to the oppression of certain groups in society). (6) The role of research in terms of function, for example policy-related research (research to evaluate the impact of social policies) and in terms of context, for example marketing research (the systematic collection of data about con- sumers of products and services in order to make informed decisions). Structure of the contributions The term ‘Dictionary’ is used to be consistent with the Sage Publications series of Dictionaries but, as with others such as The Sage Dictionary of Criminology it is more encyclopaedic in nature. Each of the contributions is structuredJupp-Prelims.qxd 3/8/2006 10:37 AM Page xii THE SAGE DICTIONARY OF SOCIAL RESEARCH METHODS according to a standardized format. First, there is a very brief definition of the concept. Second, this is followed by a longer elucidation of distinctive features, which could include historical background, disciplinary background (for exam- ple, sociology, psychology, economics), key writers, applications (where appro- priate) as well as main features. Authors were encouraged to think in terms of writing a critical and reflective essay. Therefore, for each concept, there is an evaluation in which authors raise some of the key issues and problems relating to the concept under consideration. The issues and problems which are raised are those chosen by each author rather than as a result of prescriptions laid down by the editor. It is the sections on distinctive features and on evaluation which give the publication its encyclopaedic character. For each entry, cross- references are made to associated concepts within the Dictionary. Some of these are associated by ‘similarity’ and ‘mutuality’ and others because they represent ‘challenges’ and ‘rivalry’ to the concept under consideration. The cross-references facilitate a mapping of concepts in terms of similarities and differences as described below. Finally, a brief list of key readings is provided. How to use the Dictionary The text can be used as a conventional dictionary or encyclopaedia to clarify the meaning of a term. However, more usefully it can be used in almost textbook fash- ion as a means of learning about the field of social research, and in the construc- tion of an essay or dissertation, by making use of the cross-referencing provided by the associated concepts. The latter provide a mechanism for mapping connections between concepts in terms of similarities and differences. Associated concepts relating to any given definition have been chosen to direct the reader not solely to other concepts that share common features or underlying themes and principles but also to concepts that differ – often sharply – in terms of such features, themes and principles. The features of two of the definitions in this Dictionary can be adapted to assist in this endeavour. First, network analysis is a technique that exam- ines the relationships between units of analysis. It was in part based on sociometry, a method founded upon asking children about their friendships. Network analysis is now more sophisticated and permits the examination of the strengths of rela- tionships and the degree of density and interconnectedness of networks. By fol- lowing cross-references it is possible to construct and examine networks of concepts: which concepts relate to one another, how they relate in terms of close- ness, strength of relationship, similarities and differences. The second key concept is called the constant comparative method, which is a form of analysis in qualitative research and includes the process of minimal and maximal comparison of units of analysis in order to further understanding. Minimal comparison involves examina- tion of cases which are as similar as possible and maximal comparison involves examination of cases which are as different as possible. This idea can be adapted to further the understanding of the territory of social research by listing the ways in which certain concepts in the network are similar and how other concepts differ. In this way the breadth and depth of social research can be uncovered. xiiA-Jupp-3373.qxd 3/8/2006 10:41 AM Page 1 A setting or social group, and the relative standing ACCESS of different social actors within the field. There are also occasions when access will need Definition to be negotiated with a variety of different social actors within the setting. This can mean The process of gaining and maintaining entry paying particular attention to the role(s) that to a setting or social group, or of establishing the researcher adopts during social research, working relations with individuals, in order and the ways in which these roles are managed that social research can be undertaken. throughout the research encounter. This role management can include adopting certain Distinctive Features styles of dress or personae in order to fit into the setting. Access is part of the initial phase of social Complex and protracted access negotia- research, and is usually negotiated at the start tions are most often associated with qualita- of the research project. Access can also be tive or ethnographic work (de Laine, 2000). conceptualized as a continuous process (Lee, However, there may be similar processes 1993). Relationships may need to be renegoti- involved in gaining access to secondary data, ated throughout the course of research. It is documentary sources, or large research sam- sometimes unclear exactly what access is ples. Establishing trust and gaining a working required once research is under way. familiarity of the research field are essential There are a variety of practical and theo- components of undertaking social research of retical matters associated with ensuring all kinds. research access. There are certainly practical issues involved in gaining entry to settings or data, and to establishing rapport with Evaluation research participants. These might include formal approaches or applications (by letter Access is not a single event to be undertaken or direct contact), arranging initial meetings only at the beginning of the research process. and providing descriptions of the research However, advice tends to focus on the initial for potential settings and participants. It is negotiation of access, rather than on the main- usually appropriate to offer reassurances of tenance and renegotiation of access over time. confidentiality and trust as part of initial Access is more than seeking formal permission access negotiations. or gaining informed consent. It is also part of Identifying and establishing rapport with a more general process of active engagement key gatekeepers or (informal) sponsors within with settings and social actors, and of recog- a setting can be important to gaining and nizing the need to work at ethical research maintaining access (Hammersley and relationships (Denscombe, 2002). Atkinson, 1995). This can entail developing a How much information to disclose during theoretical or analytical appreciation of the access negotiations is a matter for carefulA-Jupp-3373.qxd 3/8/2006 10:41 AM Page 2 THE SAGE DICTIONARY OF SOCIAL RESEARCH METHODS consideration. It may not be possible to through change interventions involving a inform potential research participants about process of collaboration between researchers the full extent of the research, particularly and participants. The process is seen to be both if an exploratory study is envisaged, with educational and empowering. Action research new research questions emerging over time. should not be confused with evaluation Equally, full disclosure may not be desirable, research which attempts to measure the perhaps because this may jeopardize access, impact of interventions without the active col- or potentially change the behaviour of infor- laboration of participants. mants. Sometimes it is easier to give a fuller account of the aims of the research once Distinctive Features initial access has been secured and a level of In a systematic review based on UK health- rapport and trust established. However, care settings, Waterman et al. (2001) identi- where possible, those concerned should be fied two distinguishing features of action told about any implications of the research research: first, the cyclic process, and second, from the outset. Deception should certainly the research partnership. The action research be avoided. cycle begins with the analysis of a social situ- ation or the identification of a problem. This Amanda Coffey is typically followed by the formulation of some kind of intervention (for example, nurses in a clinic may change the way they Associated Concepts: confidentiality, covert carry out some aspect of professional prac- research, dangerous research, ethics, gate- tice), which is then evaluated. The planning, keeper, impression management, informed action, reflection and evaluation may lead to consent, organizational research, participant new rounds of intervention and evaluation. observation, reflexivity, research bargain, unob- The research element often focuses on the trusive measures, validity of measurement process of change and the achievement of planned objectives. Key Readings The second distinguishing feature of action de Laine, M. (2000) Fieldwork, Participation research is a partnership or collaboration and Practice: Ethics and Dilemmas in between the researcher(s) and the researched. Qualitative Research. Thousand Oaks, CA: In traditional research there is a clear separa- Sage. tion between the researcher(s) and the Denscombe, M. (2002) Ground Rules for researched, which is seen as essential to pre- Good Research. Milton Keynes: Open serving objectivity. In action research, how- University Press. ever, there is a deliberate attempt to involve Hammersley, M. and Atkinson, P. (1995) participants as a way of promoting change and Ethnography: Principles in Practice. London: as a device to reduce the social distance Routledge. between researchers and subjects. This Lee, R. M. (1993) Doing Research on Sensitive involvement is felt to be both educative and Topics. London: Sage. empowering. As Greenwood and Levin (1998: 6) say: ‘Action research aims to increase the ability of the involved community or organiza- tion members to control their own destinies more effectively and to keep improving their ACTION RESEARCH capacity to do so.’ It is clear that underpinning these distin- Definition guishing features are certain value com- Action research is a type of applied social mitments. The first of these is what could research that aims to improve social situations be called a democratizing motive, which 2A-Jupp-3373.qxd 3/8/2006 10:41 AM Page 3 ANALYSIS OF VARIANCE (ANOVA) reverses the conventional relationship between Evaluation should also take into account how researchers and researched, which some action research aims to change the way people action researchers see as elitist and exploita- do things. Its outcomes are not necessarily tive. The second commitment is to partici- ‘findings’ in the conventional sense of theoret- pation. Drawing on a long heritage of ical progress, but in terms of new practices, pragmatist scholarship Peter Reason (1994) changed behaviour patterns or improvements has argued that traditional social science in organizational processes. alienates subjects from their own under- Given the rapidly changing nature of orga- standings of the world. Action research nizational settings and the continuing pressure attempts to transcend this alienation through on organizational members to improve their the active involvement of people in trans- performance it is likely that action research has forming organizations or social groups. The a considerable contribution to make in the third commitment is to forms of inquiry that management of change. As Waterman et al. rely on subjective understanding. Action (2001: 57) suggest, action research has the researchers believe that it is only through potential to go ‘beyond an analysis of the status action and reflection that participants can quo to directly consider questions of “what understand their social situation and through might be” and “what can be”’. formulation of interventions arrive at new understandings. John Newton Evaluation Associated Concepts: applied research, If judged by the standards of conventional emancipatory research, evaluation research, academic research, action research might messy research, participatory action appear to be unscientific. The close and research, policy-related research, practitioner collaborative relationship between researchers research and researched, for example, could be seen as a source of bias because the researcher is Key Readings no longer independent of the research set- Greenwood, D. J. and Levin, M. (1998) ting. The flexible design features of action Introduction to Action Research: Social research projects might also be an anathema Research for Social Change. London: Sage. to the mainstream social researcher. In con- Reason, P. (ed.) (1994) Participation in Human trast to the clear specification of research Inquiry. London: Sage. questions or hypotheses to be found in Waterman, H., Tillen, D., Dickson, R. and conventional empirical studies, action de Koning, K. (2001) ‘Action research: a research is characterized by a fluid and ongo- systematic review and guidance for assess- ing process of formulation, implementation, ment’, Health Technology Assessment, 5 (23). adaptation and evaluation in which the iden- tification of stages or project milestones is often difficult. Research design in action research is evolutionary rather than specified beforehand in a research protocol. ANALYSIS OF VARIANCE (ANOVA) Since there is such a marked difference between conventional and action research Definition it would probably make more sense to judge the latter on its own terms. Such a judgement A set of procedures that estimate and attribute would need to consider the action research variance in a data set to different sources process, with its emphasis on participation and determine the probability, under the to bring about change and the role of reflec- null hypothesis, of obtaining the differences tion and self-evaluation in that process. between the variance estimates by chance. 3A-Jupp-3373.qxd 3/8/2006 10:41 AM Page 4 THE SAGE DICTIONARY OF SOCIAL RESEARCH METHODS Distinctive Features matrix algebra terms, a commonality is evident. Indeed, the same matrix algebra equa- In ANOVA, variance attributable to differ- tion is able to summarize all three of these ences between groups of scores is compared analyses. As regression, ANOVA and ANCOVA with an average variance attributable to can be described in an identical manner, differences between the scores within each clearly they follow a common pattern. This group. Between-group variance is defined common pattern is the GLM (general linear with respect to differences between the modelling) conception. It is said that regres- group means, and within-group variance is sion, ANOVA and ANCOVA are particular defined with respect to differences between instances of the GLM or that the GLM the individual scores and their group mean. If subsumes regression, ANOVA and ANCOVA. the nature of the groups influences the scores Unfortunately, the ability of the same matrix more than chance fluctuation, then the algebra equation to describe regression, between-group variance estimate will exceed ANOVA and ANCOVA has resulted in the inac- the within-group variance estimate. If this curate identification of the matrix algebra difference between the variance estimates is equation as the GLM. However, just as a sufficiently large, then the null hypothesis particular language provides a means of that all group means are equal is rejected. expressing an idea, so matrix algebra provides The t-test compares the means of two exper- only one notation for expressing the GLM. imental conditions. However, when there are more than two groups or conditions, more than Andrew Rutherford one t-test is needed to compare all of the condi- tions. Unfortunately, the likelihood of obtaining Associated Concepts: experiment, general a significant result by chance increases with the linear modelling, inferential statistics, multi- number of statistical tests carried out (that is, variate analysis hypotheses tested). A Type 1 error is committed when the null hypothesis is rejected erro- neously. Therefore, the Type 1 error rate Key Readings increases with the number of statistical tests Keppel, G., Saufley, W. H. and Tokunaga, H. applied to any data set. ANOVA was developed (1992) Introduction to Design and Analysis: to assist in the analysis of data obtained from A Student’s Handbook, 2nd edn. New agricultural experiments with any number of York: Freeman. experimental conditions without any increase Kirk, R. E. (1995) Experimental Design: in Type 1 error. ANOVA procedures appropriate Procedures for the Behavioral Sciences. for an extensive variety of experimental designs Pacific Grove, CA: Brookes/Cole. are now available (e.g. Kirk, 1995). Rutherford, A. (2001) Introducing ANOVA and Nevertheless, despite ANOVA being devel- ANCOVA: A GLM Approach. London: Sage. oped for use in experimental research, it may be applied to any data organized by categories. However, as with any statistical procedure, interpretation of the ANOVA results will depend upon the data collection ANALYTIC INDUCTION methodology and the conformity of the data to the statistical assumptions underlying the Definition analysis (e.g. Rutherford, 2001). A research strategy of data collection and analysis which explicitly takes the deviant case Evaluation as a starting point for testing models or theo- When regression, ANOVA and ANCOVA ries developed in research. It can be character- (analysis of covariance) are expressed in ized as a method of systematic interpretation 4A-Jupp-3373.qxd 3/8/2006 10:41 AM Page 5 ANALYTIC INDUCTION of events, which includes the process of the development of grounded theory. Three generating hypotheses as well as testing types of results are obtained with analytic them. Its decisive instrument is to analyse the induction: forms of activities (how something exception or the case that is deviant to the is done normally), accounts of self-awareness hypothesis. and explanations, and motivational and other reasons for specific behaviours are all analysed and presented. Distinctive Features Analytic induction does not start from This procedure, introduced by Znaniecki in conventional definitions or models of what is 1934, of looking for and analysing deviant studied. As it implicitly assumes that hypothe- cases is applied after a preliminary theory ses, theories and models are not immediately (hypothesis, pattern or model) has been devel- perfect, it is a strategy to refine interpretative oped. Analytic induction, above all, is oriented conclusions from data. By searching for nega- to examining theories and knowledge by tive cases and by testing models and theories analysing or integrating negative cases. against them, it is a strategy to define their The procedure includes the following limits and to make explicit under what time, steps: (1) a rough definition of the phenome- local and social conditions they are not valid. non to be explained is developed; (2) a hypo- Therefore, analytic induction is a way to gen- thetical explanation of the phenomenon is eralize and to delimit qualitative, case-based formulated; (3) a case is studied in the light findings. of this hypothesis to find out whether the Analytic induction has been criticized as hypothesis corresponds to the facts in this it does not – as originally intended by case; (4) if the hypothesis is not correct, Znaniecki (1934) – provide a means for estab- either the hypothesis is reformulated or the lishing causal laws and universals. There are phenomenon to be explained is redefined in a question marks against the generalization way that excludes this case. Thereafter, the of case studies and external validity in gen- researcher actively searches for negative eral. Nevertheless, analytic induction has cases to discredit the hypothesis, model or its own importance as a procedure for assess- typology. ing and developing analyses by the use of Practical certainty can be obtained after a negative cases. small number of cases has been studied, but the discovery of each individual negative case Uwe Flick by the researcher or another researcher refutes the explanation and calls for its refor- Associated Concepts: case study method, mulation. Each negative case calls for the constant comparative method, deviant case redefinition of concepts and/or the reformu- analysis, grounded theory, hypothesis, induc- lation of hypotheses. Further cases are stud- tion, validity ied, the phenomenon is redefined and the hypotheses are reformulated until a universal relation is established. Key Readings Cressey, D. R. (1950) ‘Criminal violations of Evaluation financial trust’, American Sociological Analytic induction is not based on enumera- Review, 15: 733–48. tive argumentation. As it focuses on the Flick, U. (2002) An Introduction to Qualitative single deviant case to test a more or less gen- Research, 2nd edn. London: Sage. eralized model, it is a genuinely qualitative Lincoln, Y. S. and Guba, E. G. (1985) way of assessing the stability and limitations Naturalistic Inquiry. London: Sage. of research findings. With its emphasis on Znaniecki, F. (1934) The Method of Sociology. testing theories it goes one step further than New York: Farrar & Rinehart. 5A-Jupp-3373.qxd 3/8/2006 10:41 AM Page 6 THE SAGE DICTIONARY OF SOCIAL RESEARCH METHODS and social practices that draw directly or ANTI-RACISM RESEARCH indirectly on the conviction that there are racial groups which have distinct physical or Definition cultural characteristics which are usually but not only defined in negative terms. Social Research that focuses on studying the belief scientists distinguished racial prejudice, (and social practices) that there are racial which is hostile and negative attitudes groups which have distinct physical or towards minority races and ethnic groups, cultural characteristics usually, but not only, racial stereotyping and racial discrimination, defined in negative terms. defined as unjustified negative or harmful action towards minority races and ethnic Distinctive Features groups. The conclusions were that racial prej- The desire to categorize people into racial udice and racial stereotyping premised upon types based on physical appearance (or skin notions of white superiority were widespread colour) has a long tradition in Western society and pervasive in the United States and the (see Banton, 1987). Darwin’s evolutionary UK. There was also evidence of racial dis- theory of 1859 instituted the idea that there crimination across a variety of social and was a racial hierarchy based on different economic settings, and of racial violence, the species or races with natural, discernible most extreme form of discrimination. biological characteristics. Darwin also talked Researchers realized that one problem with of the ‘civilized races of man’ exterminating survey and questionnaire data was that it only and replacing the ‘savage races’ and was at provided information on the racism that peo- one with Galton in demanding control of how ple were willing to admit to. In addition, as a different races breed (eugenics). For many result of social changes, it became apparent decades, in a variety of national settings, that fewer people were willing to express positivist ‘race scientists’ researched for the overtly racist attitudes to researchers. This genetic base of race. These investigations posed a key question: were white people were originally encouraged by European becoming less prejudiced or were an increas- colonization and the development of the ing number of people much more knowing European and American slave trade and and therefore guarded about revealing their reached their high point in Nazi Germany, true sentiments on racial matters to with extensive government resources being researchers? Researchers concluded that allocated to verify scientifically the superiority subtle more flexible forms of racism had of the Ayran race. However, social scientists replaced the old-fashioned racism expressed increasingly argued that we needed to put by previous generations of respondents quotation marks around the word ‘race’ to (McConahay et al., 1981). This is likely to be indicate that we are dealing with a social the case because in several countries the construction of individuals and groups, rather direct expression of racist views has been than an established unproblematic scientific outlawed. fact (Wetherell, 1996). This rejection of the One of the key proponents of this argu- scientific validity of the concept of race ment is US social psychologist David Sears opened up the possibility of more sophisti- (see Sears, 1988; Sears et al., 1999). His work cated research designs examining how and on the new racism describes a more elusive, why ‘race’ matters. abstract symbolic language of race that avoids A substantial programme of research has blatantly negative racist statements in favour been carried out based on standardized social of political code words and symbols. This new surveys and tests with the acknowledgement racism being expressed is partly based on a in the United States and the UK in the 1970s view of racial discrimination as being out- that white racism was a serious social prob- dated and puts the onus of achievement and lem. ‘Racism’ can be said to refer to beliefs equality on blacks and other minority groups. 6A-Jupp-3373.qxd 3/8/2006 10:41 AM Page 7 ANTI-RACISM RESEARCH The new racism asserts that it is black people’s A separate sociological research programme own deficiencies that are the cause of their has focused on analysing official data to identify problems, not the history of slavery segrega- racial and ethnic inequalities, most noticeably tion, discrimination, prejudice and racism in the labour market, government agencies, that is assumed to have come to an end. The housing, education, the media and criminal new racism is thought to be most obvious in justice. This has concentrated minds on struc- white people’s views on affirmative action tural and systemic rather than individual and and various social problems. Conversations group dimensions of racism. In the UK context on these topics are often framed by an unspo- this debate received renewed focus as a result ken subtext of racist attitudes and negative of the conclusion of the judicial inquiry into the associations. Researchers have devised sophis- murder of black teenager Stephen Lawrence ticated modern racial prejudice scales to that the murder investigation was hindered examine various covert forms of contempo- by ‘institutionalized racism’ within the rary racism (see Sears et al., 1999; see also Metropolitan Police (see Macpherson, 1999). Bobo, 1999; Dovidio and Gaertner, 1998). Originally coined in 1967 by the US Black Wetherell and Potter (1992) have argued Panther Stokely Carmichael and Charles V. that the modern racism approach in social Hamilton, the inquiry defined institutional psychology is limited by its dependence on racism as: ‘the collective failure of an organiza- the prejudicial model of racism. They argue tion to provide an appropriate and professional that racism should not be narrowly equated service to people because of their colour, cul- with a particular psychological complex of ture, or ethnic origin. It can be seen or detected feelings, thoughts and motives. Their inter- in processes, attitudes and behaviour which views focused on how the taken-for-granted amount to discrimination through unwitting discourse of white New Zealanders rational- prejudice, ignorance, thoughtlessness and racist izes and justifies Maori disadvantage, how stereotyping which disadvantage minority inequality is normalized and rendered ethnic people’ (Macpherson, 1999: 28). This unproblematic and conflicts subdued. finding has provided a challenge to social scien- Sociologists, suspicious of the psychologizing tists to develop the methodological tools to of racism, have undertaken ethnographic research the British government’s programme case studies of community attitudes towards of action to identify and eradicate ‘institutional race or have accidentally uncovered everyday racism’. Given the nature of the Macpherson talk about race to be a central issue in the inquiry, it is not surprising that a considerable course of research. Some of the most inter- focus of research is on the rules, procedures esting if controversial sociological studies and guidelines that produce discriminatory out- have involved research on white racists comes in the criminal justice system. groups (see Blee, 2002). A substantial body of sociological research findings also exists on Evaluation the media and racism. Research on race and the news media, for understandable reasons, Contemporary racism takes a bewildering tends to concentrate on studying how visible variety of forms and is spawning a new ethnic minorities are negatively portrayed generation of research questions that require and crudely stereotyped by white-controlled a rethinking of conventional methodologies. media organizations. A particular focus is This is also generating critical writings on the analysing the discourses used by sections of ethical dilemmas associated with the doing of the news media to racialize public debate anti-racism research (Twine and Warren, about law and order and other social prob- 2000). Researchers need, for example, to lems (Law, 2002). The possibilities for work through how their racial and ethnic researching media and racism have increased backgrounds influence the analytical lens dramatically with the development of the through which they view their research Internet. subjects. There are also ‘who’s side are we 7A-Jupp-3373.qxd 3/8/2006 10:41 AM Page 8 THE SAGE DICTIONARY OF SOCIAL RESEARCH METHODS on?’ questions and the temptation to expose Sears, D., Sidanius, J. and Bobo, L. (1999) Racialised Politics: The Debate about the views of racist respondents to advance a Racism in America. Chicago: University of particular political agenda. Finally, there is Chicago Press. the controversial question of whose racism is Twine, F. W. and Warren, J. W. (eds) (2000) researched? To date the primary focus has Racing Research and Researching Race: been on the racism of lower socio-economic Methodological Dilemmas in Critical Race white groups. Researching institutional Studies. New York: New York University racism will require, amongst other things, Press. accessing powerful government agencies and Wetherell, M. (1996) ‘Group conflict and the multinational corporations. In addition, there social psychology of racism’, in M. is also a lack of research on what we might Wetherell (ed.), Identities, Groups and describe as the rationales and dynamics of Social Issues. London: Sage. pp. 175–234. non-white racism. Wetherell, M. and Potter, J. (1992) Mapping the Language of Racism: Discourse and the Legitimation of Exploitation. London: Sage. Eugene McLaughlin Associated Concepts: critical research, dis- course analysis, discursive psychology, eman- cipatory research, Internet research, media APPLIED RESEARCH analysis, official statistics, politics and social research Definition Research that focuses on the use of knowl- Key Readings edge rather than the pursuit of knowledge for Banton, M. (1987) Racial Theories. its own sake. A motivation behind applied Cambridge: Cambridge University Press. research is to engage with people, organiza- Blee, K. M. (2002) Inside Organised Racism: tions or interests beyond the academic disci- Women in the Hate Movement. Berkeley, pline and for knowledge to be useful outside CA: University of California Press. the context in which it was generated. Bobo, L. (1999) ‘Race, interests and beliefs about affirmative action’, American Distinctive Features Behavioural Scientist, 41 (7): 985–1003. Dovidio, J. F. and Gaertner, S. L. (1998) ‘On This engagement with the ‘outside world’ (for the nature of contemporary prejudice: the example, government departments, commer- causes, consequences and challenges of cial organizations, pressure groups) gives aversive racism’, in J. L. Eberhardt and applied research some distinctive characteris- S. T. Fiske (eds), Confronting Racism: The tics. Bickman and Rog (1998), for example, Problem and the Response. London: Sage. argue that basic and applied research differ in Law, I. (2002) Race in the News. London: purposes, context and methods. Although the Palgrave. differences are presented in dichotomous Macpherson, Sir William (1999) The Stephen terms, the authors suggest that in reality they Lawrence Inquiry. London: TSO. should be seen as continua. The differences McConahay, J. B., Hardee, B. B. and Batts, V. of purpose can be described in terms of the (1981) ‘Has racism declined in America? goals of knowledge production. For the basic It depends upon who is asking and what researcher the production of knowledge is an is asked’, Journal of Conflict Resolution, 25: end in itself whereas for the applied 563–78. Sears, D. (1988) ‘Symbolic racism’, in P. Katz researcher knowledge is used to further other and D. Taylor (eds), Eliminating Racism. ends or goals. For example, a basic researcher New York: Plenum Press. might be interested in understanding how 8A-Jupp-3373.qxd 3/8/2006 10:41 AM Page 9 APPLIED RESEARCH customers make purchasing decisions with other economic phenomena) which frequently no other concern than the process of decision inform government decision makers. Outside making itself. An applied researcher might be government, university-based economists, have interested in the same kind of behaviour but a long-standing reputation for their forecasts on be primarily concerned with how the find- economic growth (Begg and Henry, 1998). ings of the research could be used, say, to Psychology, too, has always recognized the increase the sale of certain commodities. practical application of its knowledge base. In This points to an important difference in World War II psychological assessment tech- the context of applied and basic research. niques were used to recruit pilots, and after Applied research is often initiated by some- the war returning members of the armed one other than the researcher (it could be a forces brought home emotional problems government department, a pressure group or which were managed by a growing number a commercial organization). Basic research, of clinical psychologists. Today, the discipline in contrast, is more often than not conducted of psychology has become institutionalized as by the person(s) who formulated the topic or a professional activity with a variety of research question in the first place. Where applied specialisms, such as clinical, educa- research is ‘other initiated’, the researcher tional, forensic and industrial. Each specialism may have less control over various aspects of has its own conferences and peer-reviewed the research process (for example, the design journals under the umbrella of the relevant of the study, its scope and timeframe and, national association for example, the perhaps, whether the results will be made American Psychological Association and the publicly available). Where research is ‘self- British Psychological Society. initiated’ there is greater opportunity to ‘set Sociology, in contrast, has been slower to one’s own agenda’ and be less constrained develop as a form of practice, though this is about how the research is managed and perhaps more the case in the UK than the conducted. Even so, the standard conditions United States. Even in the US, however, the attaching to externally funded projects mean President of the American Sociological that basic researchers are also subject to a Association opened his 1980 presidential number of ‘other-initiated’ requirements (for address with the remark that ‘the stance of example, compliance with ethical guidelines). our profession toward applied work … has There are no specific research methods asso- been one of considerable ambivalence’ (Rossi, ciated with either basic or applied research, but 1980: 890). Various explanations have been Bickman and Rog (1998) suggest that applied put forward for this stance. The ‘opt out’ view researchers are more likely to pay greater atten- argues that sociologists have always seen their tion to issues of external validity. This does not discipline primarily as a form of scholarship imply a lack of attention to internal validity on in which the subject is seen as an accumula- the part of applied researchers; it is simply a tion of literature to be learned, debated and matter of usefulness. Applied researchers will developed. As such, sociologists have not been want to show that their results can be used to concerned to promote or advertise their work address a problem or issue in the ‘real world’. to others. The result, Martin Albrow has sug- gested, is that sociology has become some- thing of a ‘subterranean mystery’ (Albrow, Evaluation 2000); a craft rather than a profession. The Different social science disciplines have ‘sell out’ view argues that sociology’s identity engaged in applied research with different is borne out of radicalism. The ‘debunking levels of enthusiasm. Some, like economics, motif’ inherent in some views of the sociolog- have established a close relationship between ical enterprise has meant that sociologists the development of theory, techniques of mea- have been reluctant to become embroiled surement and statistical analysis in order to in applied research because this would entail explain the workings of the economy (and an endorsement of existing power structures 9A-Jupp-3373.qxd 3/8/2006 10:41 AM Page 10 THE SAGE DICTIONARY OF SOCIAL RESEARCH METHODS and would turn sociologists into state-funded listings of individuals, groups, institutions or ‘apparatchiks’. whatever. Sometimes a further sub-sample is These ‘anti-applied’ views are not taken in which case the phrase multi-stage completely dominant, however. A volume of sampling is used. essays published by the British Sociological Association called for an ‘active sociology’ Distinctive Features (Payne and Cross, 1991). Its editors argued With area sampling, data are collected from that no academic discipline can afford to or about all individuals, households or other become too introspective or passive towards units within the selected geographical areas. the world outside academia. What is Area sampling is sometimes referred to as required, they say, is ‘an applied sociology … block sampling, especially in the United which … will be an active participative soci- States. It is also a form of cluster sampling. ology which engages with society’ (1991: 2). Evaluation John Newton Area sampling is used a great deal in coun- tries where there are no adequate population Associated Concepts: action research, criti- lists, which are replaced instead by maps. cal research, econometrics, evaluation However, the boundaries on such maps must research, emancipatory research, policy- be clearly defined and recognizable, both at related research, politics and social research, the stage of sampling and at the stage of data practitioner research collection. This form of sampling is not appropriate where the aim of the research is Key Readings to follow a cohort of individuals over time Albrow, M. (2000) Review of Steele, S. F., because of the population changes within the Scarisbrick-Hauser, A. M. and Hauser, areas that have been selected in the sample. W. J., ‘Solution centred sociology: addressing problems through applied sociology’, Sociology, 34 (3): 596–7. Victor Jupp Begg, I. and Henry, S. G. B. (1998) Applied Economics and Public Policy. Cambridge: Associated Concepts: cluster sampling, Cambridge University Press. cohort study, cross-sectional survey, longitu- Bickman, L. and Rog, D. J. (1998) Handbook dinal study, sampling, social survey of Applied Social Research Methods. London: Sage. Key Reading Payne, G. and Cross, M. (eds) (1991) Sociology in Action: Applications and Opportunities for Moser, C. A. and Kalton, G. (1971) Survey the 1990s. London: Macmillan. Methods in Social Investigation. London: Rossi, P. H. (1980) ‘The challenge and oppor- Heinemann. tunities of applied social research’, American Sociological Review, 45 (Dec.): 889–904. ATLAS.ti AREA SAMPLING Definition One of a number of Computer Assisted Definition Qualitative Data Analysis Software (CAQDAS) A form of sampling in which the clusters that programs designed to facilitate the manage- are selected are drawn from maps rather than ment and analysis of qualitative data. It was 10A-Jupp-3373.qxd 3/8/2006 10:41 AM Page 11 ATLAS.TI originally developed as an exercise to support Searches can be filtered in a variety of ways grounded theorizing. and the Supercode tool allows search expres- sions to be easily saved and re-run. Some of Distinctive Features the more sophisticated search operators are based on the nature of links between codes. Like all CAQDAS programs, Atlas.ti is a tool A number of tools facilitate the organiza- for facilitating analysis rather than a method tion and management of team projects. These in itself and therefore can feasibly be used include a tool to merge projects, a facility to support a number of methodological or allowing shared data to be accessed by theoretical approaches. It supports the man- several projects (‘Hermeneutic Units’), sup- agement and analysis of textual, audio and port for East Asian and right-to-left languages, visual data and enables the creation of easy ways to back up and move projects and (semantic) networks to facilitate theory build- XML project export. ing processes. Atlas.ti also enables the integration of As well as the code and retrieve, organiza- quantitative and qualitative aspects of tional (for example, demographic variables) projects offering a flexible word count facil- and search (for example, Boolean, Proximity ity, the results of which can easily be etc.) tools available in most CAQDAS programs, exported to statistical or spreadsheet pack- Atlas.ti provides a number of additional tools ages. It is also possible to import (and export) which increase flexibility and facilitate analytic demographic information. development. The code margin view is fully interactive, Evaluation enabling codes and memos to be easily and quickly accessed, edited, merged, replaced The evaluation of any CAQDAS package must and linked. Not only can codes be grouped take into consideration a number of factors – into (for example, theoretical) families, but such as methodology, theoretical approach, the new Super Families tool enables addi- type of data, size of data set, project aims and tional hierarchical collections to be created requirements. Therefore certain packages and a particular code can belong to any may be particularly useful for certain types of number of families. The Autocoding tools are studies and researchers are advised to inves- also particularly flexible. tigate the various options before choosing. In addition to being able to directly code Atlas.ti is clearly amongst the more sophis- multi-media data, objects (for example, Excel, ticated options available, offering a variety of PowerPoint) can be embedded into rich text powerful and flexible means by which to format (rtf.) files. These objects can then be explore, work with and interrogate qualita- edited in-context as the functionality of the tive data. As such, it enables analysis to go corresponding application becomes available beyond code and retrieve processes. within Atlas.ti. Perhaps its most frequently cited advan- The Networking tool is extremely versatile, tage is that, in comparison to other CAQDAS allowing quotations (segments of text), codes, software, Atlas.ti keeps the researcher very documents and memos to be linked to each ‘close’ to the data. For instance, when navi- other in a variety of ways. It is also gating around coded data, if required, quota- possible to create hyperlinks between quota- tions will be illustrated in their source tions to, for example, track a story or sequence context – allowing easy in-context re-coding. of events within or between data files. As a Atlas.ti is therefore often seen by researchers presentation tool, the Networking facility is as particularly well suited for ‘grounded’ flexible, allowing, for example, quotations to approaches to qualitative analysis. be displayed as illustrations of research find- Atlas.ti is a multi-faceted software and the ings or theoretical processes to be visualized. user may employ only a small proportion of The Query tool allows both simple and the tools available for any given project. Its sophisticated interrogation of the data set. size and flexibility are both its advantage and 11A-Jupp-3373.qxd 3/8/2006 10:41 AM Page 12 THE SAGE DICTIONARY OF SOCIAL RESEARCH METHODS its disadvantage. It can be used easily at a occurs in longitudinal research when the basic level but possibly needs a more experi- subjects studied drop out of the research for enced and confident researcher to make inno- a variety of reasons, which can include: vative use of specific tools for investigating unwillingness of subjects to continue to par- and representing complex relationships. ticipate in research, difficulties in tracing original respondents for follow-up (for exam- ple, because of change of address) and non- Christina Silver availability for other reasons (for example, death, serious illness). A survey of major lon- Associated Concepts: CAQDAS, coding, The gitudinal studies in the United States found Ethnograph, grounded theory, NUDIST, that the average attrition rate was 17 per cent qualitative research, QSR NVivo (Capaldi and Patterson, 1987, cited in Sapsford and Jupp, 1996). Therefore, quite a lot of Key Readings cases may be lost. A second sense of attrition relates to the Atlas.ti website http://www.atlasti.de/ loss of data from secondary sources in the CAQDAS Networking Project website process of collection. A commonly used http://caqdas. soc.surrey.ac.uk/ example is the case of crime statistics, which Coffey, A., Holbrook, B. and Atkinson P. are the end product of processes of decisions (1996) ‘Qualitative data analysis: tech- by victims to report crimes to the police, nologies and representations’, Sociological police decisions about recording and deci- Research Online, 1 (1). sions by various agencies in the criminal Fielding, N. and Lee, R. (eds) (1991, 2nd edn justice process which may or may not result 1993) Using Computers in Qualitative in prosecution, conviction and the imposition Research. London: Sage. of a particular sentence. Using the British Kelle, U. (ed.) (1995) Computer-Aided Qualitative Data Analysis: Theory Methods Crime Survey (BCS) (a random household and Practice. London: Sage. survey of criminal victimization) as a base Lewins, A. (2001) ‘CAQDAS: Computer line, it is estimated that only 45 per cent of Assisted Qualitative Data Analysis’, in the crimes recorded by BCS are reported to N. Gilbert, Researching Social Life, 2nd the police, only 24 per cent are recorded as edn. London, Sage. pp. 302–23. crimes, 2 per cent result in a conviction and Weitzman, E. and Miles, M. (1995) A Software 0.3 per cent result in a prison sentence Source Book: Computer Programs for (Home Office, 1999). Thus there is a consid- Qualitative Data Analysis. Thousand Oaks, erable loss of data between the base source CA: Sage. and subsequent measures. Evaluation The problem of attrition is not merely the ATTRITION problem of a reduction in sample size but, more importantly, it raises the possibility of Definition bias. Those who ‘drop out’ may have particu- lar characteristics relevant to the research The ‘wearing away’ or progressive loss of aims (for examples, see Sapsford and Jupp, data in research. 1996: 9). The same point applies to the crime statistics example: certain crimes may be less Distinctive Features likely to be reported, recorded or result in Attrition occurs when cases are lost from a a conviction. For example, Kelly (2000), sample over time or over a series of sequen- reviewing studies of attrition in rape cases, tial processes. One form of sample attrition found that fewer than 1 per cent of these cases 12A-Jupp-3373.qxd 3/8/2006 10:41 AM Page 13 AUDITING result in conviction. Researchers thus need to Also, a form of evaluation called programme be aware of the problem of attrition and the and performance auditing is routinely threats to representativeness that can result. performed at state and national levels. As defined by the Comptroller General of the United States (US GAO, 1994), a performance Maggie Sumner audit is ‘an objective and systematic examina- tion of evidence … of the performance of a gov- Associated Concepts: bias, cohort study, ernment organization, program, activity, or longitudinal study, official statistics function in order to provide information to improve public accountability and facilitate Key Readings decision-making’. A programme audit is a sub- category of performance auditing in which a Capaldi, D. and Patterson, G. R. (1987) ‘An key objective is to determine whether pro- approach to the problem of recruitment and gramme results or benefits established by the retention rates for longitudinal research’, legislature or other authorizing bodies are being Behavioural Assessment, 9: 169–77. achieved. Evaluators of social programmes and Home Office (1999) Digest 4: Information on performance auditors share a professional the Criminal Justice System in England and interest in establishing their independence and Wales. London: Home Office. warranting the credibility of their judgements. Kelly, L. (2000) ‘A war of attrition: recent Second, an auditing procedure has been research on rape’, Trouble and Strife, 40 suggested as a means to verify the depend- (http://www.cwasu.org/warattrition1.htm). ability and confirmability of claims made in a Nathan, G. (1999) ‘A review of sample attri- tion and representativeness in three longi- qualitative study (Lincoln and Guba, 1985; tudinal surveys’, Government Statistical Schwandt and Halpern, 1988). A researcher is Service Methodology, Series No. 13. advised to maintain an audit trail of evidence London: Government Statistical Service. documenting the data, processes and product Sapsford, R. and Jupp, V. (1996) Data (claims) of the inquiry. A third-party inquirer Collection and Analysis. London: Sage. then examines that the audit trail can attest to the appropriateness, integrity and depend- ability of the inquiry process and also the extent to which claims made are reasonably grounded in the data. AUDITING Third, auditing has entered the scene of social science theory. At issue is the prolifer- Definition ation of audit practices in all spheres of human activity – management, education, A procedure whereby an independent third social services, healthcare and so forth – party systematically examines the evidence influenced largely by the ideology of New of adherence of some practice to a set of Public Management (NPM). NPM empha- norms or standards for that practice and sizes a programmatic restructuring of organi- issues a professional opinion. zational life and a rationality based on performance standards, accountability and Distinctive Features monitoring. By being submitted to formal The concept and practice of auditing is mani- audit procedures the work of organizations is fest in several ways in the social sciences. First, held to be more transparent and accountable. in social programme evaluation, the general idea of auditing has informed the process of Evaluation meta-evaluation – a third-party evaluator exam- ines the quality of a completed evaluation Whether auditing and evaluation are (or against some set of standards for evaluation. ought to be) comfortable bedfellows can be 13A-Jupp-3373.qxd 3/8/2006 10:41 AM Page 14 THE SAGE DICTIONARY OF SOCIAL RESEARCH METHODS debated. Some observers have argued that measurement that strongly shapes conceptions evaluation and performance auditing differ in of knowledge, politics and ethics. For example, the ways each conceives of and accomplishes some scholars argue that auditing (and associ- the aim of assessing value. Some of the dif- ated practices such as total quality manage- ferences noted between the two practices ment, performance indicators, league tables, include the following. First, auditors address results-oriented management, monitoring sys- normative questions (questions of what is, in tems) is not simply a set of techniques but a light of what should be) while evaluators are system of values and goals that becomes more concerned with descriptive and impact inscribed in social practices thereby influenc- questions. Second, auditors work more inde- ing the self-understanding of a practice and its pendently of the auditee than evaluators do role in society. Thus, to be audited, an organi- with their clients. Third, auditors are more zation (or practice like teaching or providing exclusively focused on management objec- mental health care) must transform itself tives, performance and controls than evalua- into an auditable commodity – auditing thus tors. Fourth, auditors work with techniques reshapes in its own image those organizations for generating evidence and analysing data and practices which are monitored for perfor- that make it possible to provide quick feed- mance (Power, 1997). Others argue that audit back to auditees, while evaluations often culture or society promotes the normative ideal (though not always) have a longer time frame. that monitoring systems and accountability Fifth, although both auditors and evaluators ought to replace the complex social-political base their judgements in evidence, not on processes entailed in the design and delivery of impressions, and both rely on an extensive kit social and educational services and the of tools and techniques for generating evi- inevitably messy give-and-take of human inter- dence, they often make use of those tools in actions. Still others contend that the growing different ways. Finally, auditors operate influence of an audit culture contributes to under statutory authority while evaluators the disappearance of the idea of publicness as work as fee-for-service consultants or as traditional public service norms of citizenship, university-based researchers. Other observers representation equality, accountability, impar- have argued that the practices of auditing and tiality, openness, responsiveness and justice are evaluation, although they often exist inde- being marginalized or replaced by business pendently of one another, are being blended norms like competitiveness, efficiency, produc- together as a resource pool for decision makers tivity, profitability and consumer satisfaction. and managers in both public and private organizations. In this circumstance, an amal- Thomas A. Schwandt gamated picture is emerging of professional objectives (for example, placing high value on independence; strict attention to documenta- Associated Concepts: applied research, tion of evidence), purpose (for example, com- cost–benefit analysis, critical research, evalu- bining normative, descriptive and impact ation research, meta-analysis, performance questions) and methodologies (for example, indicator, politics and social research, process making use of a wide range of techniques). mapping, secondary analysis, social indicators A variety of criticisms based in empirical and conceptual investigations are directed at the audit society, audit culture, or the culture of Key Readings accountability as the latest manifestation of the infiltration of technological, means–end and Lincoln, Y. S. and Guba, E. G. (1985) instrumental rationality into the forms of every- Naturalistic Inquiry. Beverly Hills, CA: Sage. day life. Auditing is viewed as an example of a Power, M. (1997) The Audit Society: Rituals of key characteristic of modernity – that is, the Verification. Oxford: Oxford University drive for efficiency, perfection, completion and Press. 14A-Jupp-3373.qxd 3/8/2006 10:41 AM Page 15 AUTOETHNOGRAPHY Schwandt, T. A. and Halpern, E. S. (1988) their own subjectivity and life experiences Linking Auditing and Metaevaluation. (usually within the context of fieldwork). For Newbury Park, CA: Sage. literary critic Mary Louise Pratt (1992), US General Accounting Office (1994) autoethnography is a mode of self-and group Government Auditing Standards. representation on the part of colonial or post- Washington, DC: US General Accounting colonial subjects that is informed by repre- Office. sentations of them by others who are more Wisler, C. (ed.) (1996) ‘Evaluation and audit- dominant. In literary theory, autoethnogra- ing: prospects for convergence’,New phy is frequently viewed as a form of Directions for Evaluation, No. 71. San counter-narrative. Francisco, CA: Jossey–Bass. Memoirs, life histories and other forms of self-representation are analysed as autoethno- graphies when they deal with topics such as transnationalism, biculturalism or other forms of border crossing, and local cultural practices. AUTOETHNOGRAPHY Ethnographers make use of such texts to explore these issues in the context of life expe- Definition riences and cultural constructions of these. A form of self-narrative that places the self Ethnographers have also adopted the term within a social context. It includes methods autoethnography to label forms of self- of research and writing that combine autobio- reflexivity. For example, Arthur Bochner and graphy and ethnography. The term has a dual Carolyn Ellis (2002) are sociologists who advo- sense and can refer either to the ethnographic cate an ‘emotional sociology’ that incorporates study of one’s own group(s) or to autobio- personal narrative as a method to avoid the graphical reflections that include ethno- objectification of more scientific methods of graphic observations and analysis. research by erasing boundaries between the self of the researcher and that of the researched. Anthropologists such as Reed- Distinctive Features Danahay (1997) use the concept of autoethno- In autoethnography there is both self- graphy to analyse the uses of self-writing reference and reference to culture. It is a among anthropologists and among ‘natives’. method that combines features of life history and ethnography. The term autoethnography Evaluation has been used both by qualitative researchers Autoethnography, as a method of research and in the social sciences, who emphasize the writing, requires reflective and critical connections between ethnography and approaches to understandings of social and autobiography, and by literary critics who are cultural life and the relationship between the mainly concerned with the voices of ethnic self and the social. Autoethnographic texts are autobiographers. Autoethnography can be most compelling when they synthesize objec- associated with forms of the following: first, tive (outsider) and subjective (insider) points of native anthropology or self-ethnography – in view, rather than privileging the latter. which those who previously were the objects Because there is such a wide range of work of anthropological inquiry come to undertake currently labelled ‘autoethnography’, ranging ethnographic research themselves on their from creative nonfiction to ethnic autobiogra- own ethnic or cultural group; second, ethnic phy to the testimonials of postcolonial sub- autobiography – in which autobiographers jects, the task of evaluation is challenging and emphasize their ethnic identity and ethnic depends in large part on the intent of the origins in their life narrative; and third, auto- writer/researcher. Questions are frequently biographical ethnography – a reflexive raised about authenticity and voice in such approach in which ethnographers analyse 15A-Jupp-3373.qxd 3/8/2006 10:41 AM Page 16 THE SAGE DICTIONARY OF SOCIAL RESEARCH METHODS texts, regarding the authority of the speaker. context, this form of research and writing is Personal accounts of fieldwork are, for example, critically evaluated on the basis of how well often used to evoke the ‘ethnographic author- it synthesizes the subjective experience of ity’ (Clifford, 1983) of the ethnographer. participant(s) in social and cultural life and Autoethnography calls attention to issues of the structural conditions in which their lives power, and can be most effective when it fore- take place. grounds the complex and nuanced relation- ships between researcher and researched, Deborah Reed-Danahay dominant and subordinate, in the context of individual experience and socio-cultural struc- Associated Concepts: cultural research, tures of beliefs and control. ethnography, intellectual craftsmanship, life As a method, autoethnography is frequently history interviewing, oral history, reflexivity contrasted to ‘objective’ scientific or positivist methods that call for the researcher to set him- or herself apart from the object(s) of research. In ethnography, this would require a stance of Key Readings distance from the self, from those who are Bochner, A. P. and Ellis, C. (eds) (2002) being studied and from the social contexts of Ethnographically Speaking: Autoethnography, the research situation. Those who advocate an Literature, and Aesthetics. Walnut Creek, autoethnographic approach argue that reflex- CA: Alta Mira Press. ivity about oneself and about the research sit- Clifford, James (1983) ‘On ethnographic uation, that is, being aware of one’s position in authority’, Representations, 2: 118–46. the context of research rather than denying it, Pratt, M. L. (1992) Imperial Eyes: Travel is vital to a full understanding and is not com- Writing and Transculturation. London: pletely at odds with forms of ‘truth’ or valid- Routledge. ity. Although there are variations in the degree Reed-Danahay, D. (ed.) (1997) Auto/ to which autoethnographers emphasize their Ethnography: Rewriting the Self and the own experience or that of the ethnographic Social. Oxford: Berg. 16B-Jupp-3373.qxd 3/8/2006 10:41 AM Page 17 B uncontested central component of social BIAS science methodology. In recent years, however, especially under the influence of construction- Definition ism and postmodernism, there has been grow- ing debate, especially among qualitative Generally regarded as a negative feature of researchers, about the meaning and usefulness research, as something that can and should of these terms (see, for example, Lather, 1986; be avoided; occasionally the term is used in a Kvale, 1989; Harding, 1992; Altheide and neutral or even a positive sense, referring Johnson, 1994). In part, this reflects the fact simply to the fact that the researcher has that they had previously often been interpreted adopted a particular angle of vision. in ways that depended on a form of positivism that is now largely discredited. The latter pre- Distinctive Features sented research, when properly executed, as producing conclusions whose validity follows Even in its negative sense, there are broader automatically from the ‘givenness’ of the data and narrower interpretations of the term. on which they are based. Sometimes it refers to any systematic deviation On this view, the course that inquiry should from the truth, or to some deformation of take is clearly defined and, as a result, devia- research practice that produces such deviation. tion from it – whether caused by prior com- Thus, quantitative researchers refer to ‘mea- mitments or by some other source of error – is surement bias’ and to ‘sampling bias’, by which also straightforwardly identifiable. What is they mean systematic failure in measurement required to avoid bias is for researchers to be or sampling procedures that produces erro- objective; in other words, they must pursue neous results. The contrast here is with random research in the way that ‘anyone’ would pursue (or haphazard) error. However, another influen- it who was committed to discovering the truth, tial usage of the term ‘bias’ refers to a particu- whatever their personal characteristics or lar source of systematic error: a tendency on social position, appealing only to data that are the part of researchers to collect data, and/or to observable by ‘anyone’. interpret and present these data in such a way The influence of positivism meant that a as to favour false results that are in line with clear distinction was not always drawn their presuppositions and/or their political and between, on the one hand, a researcher having practical commitments. This may consist of a potentially biasing commitments, for example positive tendency towards a particular, but particular political views, and, on the other, false, conclusion; or it may mean the exclusion these commitments impacting negatively on from consideration of some set of possible con- the research process. In other words, clusions that happens to include the truth. researchers were (and sometimes still are) described as biased simply because they have Evaluation commitments pertaining to the topic on ‘Bias’ is part of a set of terms – ‘validity’ and which research is being carried out. This fol- ‘objectivity’ are others – that were once an lows from the false assumption that, in order