Sociology Methods in Context

Primary data

Info collected by sociologists themselves for their own purpose

Social surveys

Involve asking people Qs in written questionnaires or an interview - type of way to obtain primary data

Participant Observation

Sociologist joins in with activities of group he or she is studying - type of way to obtain primary data

Secondary data

Info that has been collected or created by someone else for their own purpose, but which the sociologist can use

Official statistics

Produced by government on wide range of issues such as education, crime, divorce and unemployment, as well as other statistics produced by charities, businesses, churches and other organisations - type of secondary data

Public documents

Produces by organisation such as governments, schools, media, etc. These include Acts of Parliament, school prospectuses, newspaper articles etc. -type of secondary data

Personal documents

Created by individuals and often provide first person accounts of evens and experiences. They include diaries, letters, autobiographies, etc - type of secondary data

Quantitative Data

Refers to information in numerical form, e.g.; official stats on how many girls passed five or more GCSEs

Qualitative Data

Info, usually expressed in words about people's thoughts, feelings, motivations, attitudes, values, etc. - obtained through (e.g.) participant observations

Positivism

Belief that society is made up of 'social facts' that can be studies scientifically to discover laws of cause and effect

Interpretivists

Focus on how we construct our social worlds through the meanings we create and attach to events, actions and situations

Subjectivity

Bias, lack of objectivity, where individuals own view point influences their perception or judgement - interpretivists believe sociology is inevitably subjective, since it involves understanding other humans by seeing the world through their own eyes

Objectivity

Absence of his or preconceived ideas that have an influence on how we see things - positivists believe sociology can achieve objectivity by modelling itself on natural sciences, using methods that keep sociologists detached from their research methods

Hypothesis

Possible explanation that can be tested by collecting evidence to prove it true or false - positivists favour, as it is starting point for research and finding cause and effect relationship

Operationalisation

Process of turning a sociological concept or theory into something measurable e.g.: sociologist studying effect of "social class" on education achievement might use parental occupation to measure the concept "social class"

Pilot study

A small-scale trial run, usually of a social survey, conducted before the main study. It's purpose is to iron out any problems, clarify questions and their wording, giver interviewers practice, etc, so that any necessary changes can be made before the main study I carried out

Triangulation

Use of two or more different methods or sources of data so that they complement each other; the strengths of one countering the weakness of the other and vice versa. E.g.: using both qualitative method such participant observations and quantitative method such as structured interviews

Sample

Smaller group selected for larger survey population to take part in a study - too costly and time consuming to study whole population

Sampling

Process of selecting a sample. Aim of sampling is usually to select a sample that is representative of wider survey population, so as to allow the study's findings to be generalised

Sampling frame

List of people from which a sample for a social survey is selected e.g.: school role could be sampling frame for survey of pupils - must be complete and accurate as possible, as well as up to date and without duplications, otherwise the sample chosen from it may not be truly representative of the pop

Random sampling

Simplest technique, where sample selected purely by chance e.g.: names drawn out of a hat - everyone has equal chance and large enough sample should reflect characteristics (class, gender, etc) of whole research pop

Systematic sampling (or Quasi-random)

Where nth person in sampling frame is selected

Stratified random sampling

Research first stratifies (breaks down) pop in sampling frame by age, class, gender, etc. Sample then created in same proportions, e.g. If 20% of pop are under 18, then 20%of sample also have to be under 18

Quota sampling

Pop is stratified (broken down) by age, gender, etc. And then each interviewer is given quota of (e.g.:) 20 females and 20 males, which they have to fill with respondents who fit these characteristics. Interviewer keeps at this task until quota is filled

Snowball sampling

Involves collecting sample by contacting number of key indiv who asked to suggest others who might be interviewed and so on until enough data has been collected

Opportunity sampling

Involves choosing from those indiv who easiest to access e.g.: passersby on strewn or captive audience like a class

Hawthorne Effect

Where subjects of research study know they are being studied and begin to behave differently as a result, thereby undermining study's validity

Field experiments

Takes place in subjects natural surroundings rather than in artificial lab environment and those involved are generally not aware that they are subjects of an experiments, so no Hawthorne Effect

Rosenthal and Jacobson - field experiment

Example of field experiment - IQ test in school

Comparative method

Carried out only in mind of sociologist - "thought experiment" and doesn't involve researcher actually experimenting on real people at all. It works as follows:

-identify two groups of people that are alike in all major respects except for one variable we are interested in

-then compare two groups to see if this one difference between them has any effect

Describe Harvey and Slatin (1976) experiment

Lab experiment to see if teachers had preconceived ideas about pupils of different social classes - pics of 18 children from different social class backgrounds - to control other variables pics equally divided by gender and ethnicity and teachers asked to rate on performance, parental attitudes to education, aspirations and so on (no real children)

Harvey and Slatin's Conclusion

Lower class children were rated less favourably, esp by experienced teachers - based rating off similarities between pupil in pic and pupils they've taught - this, indicated teachers label pupils from different social classes and use these labels to pre-judge pupils potential (no real children)

Charkin et al (1975) - lab experiment

Used sample of 48 uni students to teach lesson to 10 y/o. 1/3=told boy highly motivated and intelligent (high expectancy group). 1/3=told boy poorly motivated with low IQ (low expectancy group) and 1/3=no info. Conc: high expectancy group=mods eye contact and more encouraging body language than low expectancy

Mason (1973) - lab experiment

Looked at whether neg or pod expectations had greater effect. Teachers given pos, neg or neutral reports on pupil - then teachers observed recordings of pupils taking test, watching to see if any errors were made - finally asked to predict pupils end of year attainment. Neg reports had greater impact than pos ones on teachers expectations

Rutter (1979) -questionnaire

Used questionnaires to collect large quantities of data from 12 inner London secondary schools - from this, could correlate achievement, attendance and behaviour with variables such as school size, class size and number of staff. (Diff to do with methods like interviewing or observation) -no explanations for correlations

Powney and Watts (1987) - interview or questionnaires

Note young children tend to be more literal minded and often pay attention to unexpected details in questions and may use diff logic from adult interviewers, therefore training needs to be more thorough (costly) -verbal skills better of young people so interviews are better

Di Bentley (1987) - interviews

Began interview with young people with jokey image of her fooling around with her daughter to maintain relaxed atmosphere as well as by nodding, smiling and making eye contact (not easily standardised so unreliable)

Field (1987) - access rate

Tried to do study of pupils experience of sex and health education in school - high refusal rate of 29%, mainly because parents withheld consent - if did gain consent could conduct interviews

Bell (1981) - interviews

May see interviewer as "teacher in disguise" in interviews

Greene and Hogan (2005) - interviews

Argued interviews should use open ended questions, not interrupt children's answers, tolerate long pauses to allow children to think about what they want to say, recognise that children are more suggestible and so it is particularly important to avoid asking leading Qs and avoids repeated Qs since makes children change first answer as they think it's wrong (to improve validity of interviews with pupils)

Labov (1973) - interviews

When using formal interview technique to study Lang of black American children, appeared to be tongue-tied and 'linguistically deprived', however when adopting more relaxed, informal style - interviewer sitting on the floor and children allowed to have friends present - they opened up and spoke freely, showing they were competent speakers (unstructured favourable)

Eggleston (1976) - observations

Needed over three months just to set up his cover role for his observations (time consuming)

Wright (1992) - personal characteristics and observations

Found her African Caribbean ethnicity prod antagonistic reactions from so white teachers, on other hand found that many black pupils held her in high esteem and would ask for her support

Delamont - practical and ethical issue of observations

'Guilty knowledge' -sees something or hears something that could get children into trouble (e.g.: stealing from school) - may involve law- both ethical - obliged to report wrongdoing- and practical problem - breach trust pupils have placed in researcher and may mean pupils no longer confide or cooperate with research. Harm could also be done to pupil, teachers and school - protect identity. Issue in marketisation of school- image

King (1984) - Hawthorne effect

Tried to blend into background of infant school by initially spending short periods of time in classroom so children Familiar with presence. So not seen as teacher, avoided eye contact and politely refused their request for help. In attempt to be unobtrusive, used classrooms Wendy house as 'hide' - diff to reduce Hawthorne

Willis (1977) - observations and rep

Small scale study of anti school subculture of 12 boys - took time to become familiar with setting, gain trust of teachers and pupils and carry out actual observations. Limited scale of typical observational study and size of edu system makes observing school interaction unlikely to prod Rep data.

Gillborn (1995) - secondary sources and practicality

In study of racism and schooling, able to access wide range of school docs, inc school policy statements, local authority guidelines on anti-racism, and minutes of staff netting a and working parties -gave official image of what happening in terms of racism and anti racism, compared with data from interviews and observations

Gerwitz et al (1995) - secondary sources and practicality

In study of marketisation and education found school brochures and prospectuses were useful free source of info about how school present themselves in edu marketplace

Valarie Hey (1997) - secondary sources and practicality

Personal docs - made use of notes girls passed to each other in class to understand friendship patterns, however not always easy to obtain as girls experts at hiding them from teachers (ethical issue: consent given - girl gave her notes freely - also not given- taken from wastebasket. Rep- unlikely as unsystematic way in which she came by notes)

Non participant observation

Primary research method where observer records events without taking part in them

Covert participant observation

Sociologist Identity and purpose kept secret when sociologists studies group by taking role within it and participating in activities

Overt participant observations

Participants aware of researcher true identity and motive when sociologist studies group by taking role with it and participating in activities

Questionnaires

List of questions - written or self-completion questionnaires are widely used in large-scale social surveys where they may be sent out and returned by post - mainly close ended Qs with pre-coded answers for ease of analysis

Response Rate

Proportion of those people included in social survey who actually reply or respond to the Qs asked.

Stratification

Division of society into hierarchy of unequal groups. The inequalities may be of wealth, power and/or status (based on age, gender,etc)

Representative sample

When participants in research have same characteristics as those in target pop

Soft statistics

Harder to quantify and define or measure, changes over time might be artificial, e.g.: crime stats are only made up of those reported, rape victims may begin to report the crime as result of campaign

Longitudinal studies

Followed same sample or group over extended period of time

Example of longitudinal studies

-Jwb douglas' study of over 5000 from first week born in 1946 through schooling

-Lacey's four year participant observation study of hightown grammar school