Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
SOCIOLOGICAL RESEARCH METHODS ADVANTAGES/ DISADVANTAGES/ EXAMPLES– USES LIMITATIONSTABLE EDUCATION SUMMARY METHOD DEFINITION Manip IV measure DV – control.- Most controlled setting Likely overt SOCIAL SURVEYS (SELF REPORT) (P.180-205) EXPERIMENTS (P172-179) LAB Generates quantitative data – easy to analyse Best at identifying causal relhips Most control of confounding variables Replication – reliability Experimenter =detached – less involved. Artificial – low ecological validity – mundane realism (cp. exp realism) Demand chics & evaluation apprehension – Hawthorne Effect Ethics – informed consent, deceit, welfare Small-scale – inappropriate large scale social phenomena & unrepresentative. Interpretivists –inappropriate for studying people – freewill. T/ P Relationships – Labelling Harvey & Slatin (1976) photos Charkin et al (1975) – students & 10yr old Mason ((1973) – Reports & video MiC ISSUES Pupils = special ethical issues – age, understanding, welfare. Difficult to control variables in schools – e.g. other factors affecting labelling Hard to study effect of large scale changes e.g. Gov policy in lab setting Artificial may not apply to real schools Replication difficult because schools differ and change rapidly over time. Special ethical issues in schools FIELD Poss causal reliships Some control Less Hawthorne effect External validity – more realistic More difficult to control – lack internal validity Ethics – harder to inform/debrief, often involves deceit & may affect welfare Rosenthal & Jacobson ‘Spurters’ (1968) NATURAL - Avoids artificiality Can be used to study past events No ethical issues re harming others/deceit et. Very little control therefore can’t clearly establish cause and effect. Brunsma &. Rockquemore (1992) US study -Effects of school uniform Schools are very complex institutions – difficult to control confounding variables eg tendency to insist on uniforms likely to be related to range of factors Favoured by positivisits Easy replication Comparable data Generalisation Inexpensive Large scale - representative Closed Qs – easy to quantify No interviewer bias Postal – geographically diverse sample Anonymity/confidentiality – good for sensitive issues e.g. bullying. Fewer ethical issues Interpretivists reject – imposed ‘meaning’ – thus lowvalidity Postal – low response rate - unrepresentative Qaire design – ambiguity, leading Qs. Understanding Qs Incomplete, illegible, incomprehensible responses Closed Qs – limit responses Not taken seriously - socially desirable / acquiescent response set. Rutter (1979) – London school factors in achievement. Chubb & Moe (1990) US schools, students & parents factors affecting ed achievement Conner & Dewson(2001) Postal Q aire – W/C students university choice Sullivan (2001) – cultural capital & ed achievement McBeath et al (2001) School effectiveness. Smith & Tomlinson (1989) – Pupil Qaire ethnic variation. Age group – language, understanding, attention span Easy access to sampling frames – e.g. registers, tutor groups, departments. Time constraints e.g. teachers & working parents –low response or unrepresentative sample Authority of head/school may increase response rate. Informal communication may increase demand characteristics. May get treated like a ‘test’/associated with authority. Natural setting e.g. school Likely covert Comparitive Method Compares 2 groups who differ in only one respect QUESTIONNAIRES INTERVIEWS Differ from questionnaires in that they involve some form of social interaction, depending on type: Self-completed, postal, interview (structured – see below). Qs closed/open. STRUCTURED Similar to Qaire but Qs (usually closed/pre-coded filled in by interviewer SEMISTRUCTURED UN-STRUCTURED Informal open-ended. May have a genera topicdriven interview schedule but more like a conversation. GROUP Typically 8-10. Asked Qs in turn or focus group – guided discussions. Preferred by positivists Easy to train interviewers Fairly quick & cheap – larger scale. Higher response rate Opportunity for clarification Can include responses from those who can’t read and write. Comparable/quantifiable data Formal enough to reduce interviewer bias / effects Easy to replicate Similar to structured approach More opportunities to probe, clarify, develop etc More depth, detail. Preferred by interpretivists Rapport – more in-depth info esp on sensitive topics. Interviewer free to take direction from interviewee. More flexible Fresh insights/unfamiliar topics More opportunities to clarify More like real world – most opinions manifested in gp contexts Richer data if more informal. Restores power imbalance between interviewees and interviewer. Ideas stimulated by interaction Opportunities to observe group dynamics. More expensive / time-consuming than Qaires Interviewer bias – effects of age, gender, personality Demand characteristics – social desirability, screw you effect. Closed Qs - limited opportunities for original responses – less valid. Interpretivist/feminist critique – meanings imposed rather than understanding subjective experience in a broader social context. Beron et al (2001) US Study – vocabulary tests for 3-14 yrs Loss of standardisation Loss of comparability Becker (1977) _Teacher perceptions of the ‘Ideal Pupil Mirza (1992) with black girls Sue Sharpe (1976 & 1990) Time-consuming/expensive More training & skills Smaller scale – less representative Difficult to replicate Difficult to quantify, analyse, compare, establish causal relationships More interviewer bias / effects Demand characteristics – social desirability. Group may be dominated by a few eg most confident/articulate. Difficult to keep focus Effects of peer group pressure affect validity. Harder to analyse data Gerwitz et al (1995) – Parental interviews on school transfer Labov – Black kids informal cpd formal setting Lacey (1970) Informal conversations with staff & pupils Willis (1976)– Anti-school subculture Age group – verbal/linguistic skill (tho usually better verbally than literacy), attention spa, focus ,suggestibility – importance of body language – kids more alert to nonverbal cues and give time to think – can consider alternatives e.g. using pictures Interviewers need special training and skills. School ’grapevine’ – increase demand characteristics & undermine validity Dangers of formality – associated with teachers & authority – reduce validity. Schools are hierarchical – influences interaction and validity Gatekeepers – permission from e.g. head LEA, parents etc – may increase or decrease access, depending on topic OBSERVATION (P. 206-217) PARTICIPANT OBSERVATION Favoured by interpretivists –rich qualitative ’real world’ data – based on subjective insights, ‘verstehen’ (empathy). Covert – natural behaviour – valid esp sensitive issues Overt – reduced moral & ethical issues, easier to record, can question openly Flexible - New insights Access to ‘hidden’ worlds e.g. delinquent groups. NON-PARTICPANT OBSERVATION Overt/covert – as above Structured – preferred by positivists –Quick, cheap, less training, generate quantitative data – easier to replicate, analyse, compare Unstructured – Preferred by interpretivists – More valid, qualitative data, giving access to meanings. Unstructured. Natural situation, observer participates – usually covert – enter ‘meaning world’ May be covert/overt May be structured SECONDARY SOURCES / unstructured Rejected by positivists – unreliable (can’t replicate) & unrepresentative. Also ignores structural context. Overt – Observer effects Covert – note taking & going native Getting in, staying in, getting out esp covert Time, money, involvement, stress Highly trained / skilled observer with appropriate characteristics Objectivity - observer bias – overinvolvement, selective reporting Changing behaviour – researcher effects Small-sample. Can’t generalise Ethical issues – deception, consent, confidentiality, privacy, welfare – legal & moral duties. Overt/covert – as above Structured – Rejected by interpretivists – lacks validity, categories controlled, ignoring meanings, Unstructured –Rejected by positivists More time consuming, more training required, harder to analyse & compare data. Difficult to replicate Willis (1977) Anti-school subculture. Ball (1981) – Overt participant observation of a comprehensive school Wright(1992) – Ethnic stereotypes in four primary schools Hargreaves ((1967) As a teacher observed effects of streaming and labelling. Fielding – Students acted as researchers Only a limited number of roles – resticted e.g. teachers pupils, caretaker – many have restrictions – physical – age, gender, ethnicity and educational qualifications educational Difficult to get time / privacy for recording Ethics – special issues – related to age group and context – potential harm if info sensitive info revealed. Also need to protect school in marketised economy. Flanders (1970) - US classroom analysis using Flanders Interaction Analysis Categories FIAC Keddie (1991) – Observed staff meetings and classes. Gilborn (1995) – school policies, LEA guidelines, minutes of staff meetings Gerwirtz et al (1995) – school brochures & prospectuses Hey (1997) Used notes passed in class to understand friendship patterns Lobban (1974) – Content analysis of gender roles in reading schemes. Schools are structured environments – easier to develop effective behavioural categories e.g. FIAC Teachers are practised performers – e.g. Ofsted Children may be particularly vulnerable to observer effects Gatekeepers – e.g. teachers, heads may control who is observed. Schools unique Accountability means official data accessible e.g. league tables. Issues of equality & achievement shared by Gov & Sociologists Official definitions e.g. league tables – vary over time Errors & Bias – e.g. present schools, gov, individual in a good light ref. marketisiation Access QUANTITATIVE Preferred by positivistis – ‘social facts’ Free Official - Large amounts of data – representative, allow comparisons over time, standardised collection & measures Rejected by interpretivists – socially constructed – lack validity. Marxists – Official statistics represent the interests of Capitalism Not specific to aims – collected for different purpose No control over collection QUALITATIVE Preferred by interpretivists – more authentic, rich qualitative data esp personal. Positivists can analyse using content analysis Historical can allow study of patterns over time Rejected by positivists –esp personal can’t generalise, unrepresentative, open to interpretation. Problems of assessing authenticity esp historical Validity – may be written for a specific purpose 1.Statistics – Official Data gathered Gov & official bodies and nonofficial. 2. Existing sociological research 1. Documents – public, personal & historical 2. Existing sociological research ALSO REF: Case Studies & Longitudinal Research