Section 1 - Sociological Methods - complete

  • Created by: scarlett
  • Created on: 09-01-20 19:46

Collecting and Using Data

- sociologists try to make their research reliable and valid
- reliable research can be repeated to get the same results and another researchers would be able to get the same results using the same methods
- sociological research isn't generally as reliable as research in the natural sciences
- valid data is a true picture of what the researcher is trying to measure
- reliable data isn't always valid; for example: you could use unemployment stats to measure how many people don't work but this wouldn't give a true picture because these stats don't include students who don't work or people who are unable to work
- there are several reasons why research may not give a true picture:
- respondents in an interview may forget things, exaggerate or lie
- asking people about their attitudes to an event a long time afterwards often isn't valid as people change their views over time, and may alter their description of the past in the light of their current beliefs
- you can't research the whole population, so you have to take a sample
- sociologists try to make their samples representative (reflective of the population as a whole) and to do this, its needs similar proportions of different ages, genders, classes and ethnic groups
- if a sample is representative then sociologists can generalise
- sociologists aim to be objective and avoid bias

1 of 52

Primary data

- researcher collects primary data first hand
- primary data is obtained from first-hand research
- it doesn't rely on another sociologist's research and you can carefully choose your method to make your data as valid and reliable as possible
- primary data is always new and up to date
- some methods of getting primary data can be expensive and time-consuming
- some methods may put the researcher in a dangerous situation
- some methods may be unethical if you don't get informed consent
- the researcher's own values may mess with the research process - bias
- you can't always get access to the group you want to study

2 of 52

Secondary Data

- official stats, diaries, letters, memoirs, emails, TV documentaries, newspapers etc
- generated by other people
- quick and easy to collect
- easily use to compare different societies
- able to study past events and societies (able to comapre past and present)
- don't have to worry about informed consent
- may not be valid or reliable
- may not be authentic, representative or credible - official stats can be biased
- the data you need might not exist
- your values can affect how you analyse the data
- researcher's values might ruin the validity of the orignial research

3 of 52

Quantitative Data

- reliable but not very valid
- numbers and stats, can easily be put into a graph or chart
- able to test hypothesis and look for cause and effect relationships
- compare stats against existng stats and look for trends over time and between societies
- easy to analyse tables, charts and graohs
- repeat questionnaires and structured interviews to test reliability
- allow large samples so findings can represent the general population
- stats can hide reality
- categories in interviews or questionnaires can distort the truth
- stats don't tell you anything about the meanings, motive and reasons behind behaviour (not much depth and insight into social interaction)
- stats can politically biased as the method may have been chosen to get the 'right' data

4 of 52

Qualitative Data

- valid but not reliable
- detailed picture of what people do, think and feel but hard to turn into a graph or chart
- subjective: opinions, meanings and interpretations 
- gives insight into social interaction
- lets you find out the meanings and motives behind behaviour
- don't have to force people into artificial categories like in questionnaires
- let you build up trust and research sensitive topics
- qualitative investigations are difficult to repeat so they aren't very reliable
- the research is often on a small scale so the findings might not represent the whole population
- positivists say qualitative results lack credibility because they're subjective and open to interpreration
- the researcher can get the wrong end of the stick and misinterpret the group or individual they're studying

5 of 52

Theoretical Issues and Positvists

- positvism looks at the institutions in society and it's called macrosociology
- positivists say behaviour is influence by external social factors
- they think sociology should be scientific and analyse socialf facts
- social facts are things that affect behaviour and can be easily measured. they're external things like laws, not internal things like people's opinions
- they uses quantitative data
- they use stats to measure the relationships between different factors
- they're interested in cause and effect relationships e.g. the factors that cause underachievement in schools
- they use sources like questionnaired and official statistics as they are (generally) objective and reliable

6 of 52

Theoretical Issues and Interpretivists

- believe that you can only really understand human behaviour using empathy
- think that it is important to uncover and understand the meaning individuals give to their actions and to the actions of others
- use methods that let them discover the meanings, motives and reasons behind human behaviour and social interaction
- think that the scientific methods used in positivist research don't tell you much about how individual poeple act in society
- don't believe in the existence of 'social facts' and think that the findings of research are always subjective
- say you can't count meanings and opinions and turn them into charts
- think sociology isn't scientific because humans can't be measured like ants in an ant farm
- people don't always understand questions in questionnaires and they don't always tell the truth
- use methods that produce qualitative data
- try to understand human behaviour from the point of view of the individual person
- methods such as participant observation and unstructured interviews to build up a rapport with people so they can produce a valid and detailed picture of what they think

7 of 52

Practical Issues

- some methods need more time such as covert participant observation as the researcher has to get into the group they're studying and win their trust before starting the actual research
- a social survery doesn't need the researcher to participate all the time and the workload can be shared in a team
- affects the length ans method of the research as money is needed to pay the researcher, for transportation to interviews, and to pay for resources
- large scale social surveys are expensive (2011 census cost £480 million) whereas small focus groups will cost a lot less
Characteristics and Skills of the Researcher
- difficult for a female researcher to be involved in a participant observation of monks in a monastery
- some researchers may be OK with dangerous situations and others may prefer to stay at their desk and do detailed analysis of stats
Access and Opportunity
- if researchers don't have access to certain groups to carry out interviews or observations then they have to turn to seconday sources of data

8 of 52

Ethical Issues

- all participants must have openly agreed to take part
Avoidance of Deception
- researchers should be open and honest about the study and its implications
- the details of all participants and their actions must remain confidential and private
Avoidance of Harm
- participants should not be physically or psychologically harmed by the research process

9 of 52

Ethical Issues and Covert Studies

- the researcher should get consent before they start and should be open and honest about the work they wish to carry out
- poeple with learning difficulties may not fully understand what participation would entail which is problematic as it could be argued that uninformed consent is not actually consent
- consent can be difficult to obtain from secretive groups or when the research is about a sensitive topic
- covert methods involve not telling the groups being studied that they actually are being studied
- they're often criticised for their lack of honesty and the absence of true informed consent
- consent participant observers argue that to negotiate access into sensitive or dangerous groups such as criminals, the researcher often had to either pretend to be part of the group, or not inform the group of the true purpose of the study

10 of 52

Laud Humphreys' "Tearoom Trade" (1970)

- Humphrey wished to study men who engaged in gay activities in public places (e.g. toilets)
- they were secretive about their sexuality as homosexuality was taboo in mainstream society, sexual activity in public is against the law, and some of the men may have been married men leading a "secret life"
- he probably wouldn't have gained access to this group if he'd openly and honestly informed them about the nature of his research and then sought their permission
- even if he did get consent, it is likely that they would have acted very differently if ther were aware that they were being observed
- therefore he posed as someone who watches homosexual acts for a sexual thrill which enabled him to gain the trust of the group and observe genuine actions
- other sociologusts argue that work like this shouldn't be conducted, even if it would provide valuable information

11 of 52

Milgram (1974) - Experiments on Obedience (1)

- Milgram conducted a series of experiments in which volunteers were told to administer electric shocks to another person (who was actually an actor) on the other side of a glass screen, when that person failed to give the correct answers in a memory test
- many volunteers kept on giving punsihment shocks until the actor pretended to pass out
- Milgram lied about the purpose of the experiment as he told the volunteers that they were doing an experiment about memory
- the electric shocks weren't real and the person who the volunteers where "shocking" was pretending
- the results of the experiment were very useful 
- the experiment showed how people are ready to obey authority figures without question
- this helped people understand how ordinary people take part in war crimes and genocide

12 of 52

Milgram (1974) - Experiments on Obedience (2)

- the experiment wouldn't have worked if the volunteers knew the real purpose
- if they knew their obedience was being tested, they might have deliberately been less obedient
- if they knew the shocks weren't real, they probably woudln't have behaved in the same way
- many of Milgram's original participants showed signs of distress during the experiment, and some of them were disturbed by how easy they were manipulated
- however, Milgram did debrief all of his participants afterwards to they all understood the study, and did follow-up work to check on their psychological state
- he found that some participants saw the experiment as a valuable learning experience

13 of 52


- all respondents taking part in a piece of research must have their basic right to privacy valued and upheld
- the data gathered from them and thier personal details must not be distributed to anyone outside the research process
- when the report is finally produced, respondents must be made anonymous
- any descriptions of people, geographical locations and institutions have to be written in a way that prevents reader from easily recognising the participants
- false names may be used in which case the researcher should clearly state that false names have been used, in case someone who shares the name is mistakenly identified as having taken part in the research
- if a researcher breaches trust and confidentiality, potential participants will be put off taking part in fufture studies
- research participants must feel they can trust the researcher, especially if the research is of a sensitive nature e.g. self-reported crime stufy or a sexual health study

14 of 52

Avoidance of Harm

- emotional and physical harm is never acceptable in sociological research, and work is actively criticised and rejected if it has allowed harm to come to those involved
- researchers studying topics such as mental health or geriatric care may have contact with vulnerable groups of people, or witness situations and experiences that cause individuals harm e.g. inappropriate living conditions, or abuse by carers
- there is an ethical question as to whether they should stop or suspend the research in order to remove the individual from the dangerous situation
- some topics that are discussed may be traumantic for the respondents and they would need to be informed of the possible temporary mental and emotional harm before starting the study

15 of 52

Bending or Breaking Ethical Rules

- many sociologists can justify breaking or slightly bending some of the ethical rules if the data that they'll gather is likely to make a beneficial contribution to society
- this justification becomes even stroneger if potential ethical problems are minimised e.g. if there's minimal harm and full confidentiality

- for example, Nigel Fielding (1981), in a study of the National Front (an extreme right-wing political parrty with a secretive hierarchy) argues that he needed to conduct covert research otherwise he wouldn't have been able to gain access to the group and gather information

- "James Patrick" (1973) was a false name given to a researcher conducting a study on violent gangs in Glasgow to ensure his own safety and protection

- Roy Wallis (1977) wasn't entirely honest when researching scientology
- he didn't say we was a sociologist when he signed up to a Scientology course
- if he had been honest, they may have told him to go away
- also he was forced to name some of his sources, during a legal battle between the Church of Scientology and another researcher 
- this broke the rule of privary and anonymity, but in this case Wallis had no choice

16 of 52

Picking a Topic based on Preference and Knowledge

- sociologists often specialise in different fields of sociology and will often choose a topic they have experience or knowledge of
- they try to pick a topic that they think they'll find enjoyable and interesting as if not it may lead to a poorly constructed report (i.e. flawed or boring)
- certain topics become popular at different times e.g. in the mid-1900s, it often focused on stratification and the class system but nowadays its more about world sociology
- to gain prestige, funding and public or academic interest, sociologists are more likely to focus their research on topics that are currently in vougue
- sociologists and other academics who want to make a change in society prefer research that could help develop solutions to social problems
- sociologists may feel that a particular issue is neglected so they'll research the issue to try and 'plug the gap' and encourage others to embrace the issue as well

17 of 52

Funding and Cooperation for Research

- there are a wide range of potential sources of funding: charities (Joseph Rowntree Foundation), industry, government etc.
- a lot of quantitative studies are done directly by government agencies
- the organisation which funds the research can sometimes be known as the gatekeeper because they often have the final say in the choice of topic, the way it is researched or whether a topic gets researched at all
- government agencies often do research into areas covered by current or proposed government policy
- industrial giant providers tend to fund research that gives their industry some practical benefit
- a researcher also needs to decide whether or not they will be able to get the cooperation of the groups they'll be studying if they choose a particular topic
- if potential subjects refuse to give their help for the research, then the topic may not be viable

18 of 52

Career in Sociology

- researchers would jump at the chance to conduct a study that improves their employability
- interesting, original or popular topics that are well researched, with good clear results, improve an academic's chance of having their work published
- getting work published, particularly in a big sociological journal, really improves a researcher's standing in academia
- a quick way for a sociologist to progress in their career is to respond to another sociollogist's work
- the aim can be to prove or disprove their theory or to add something to their research
- practically speaking, this could mean investigating the same topic, but using slightly different methods, or investigating a different group of people
- this can mean that particular social groups are researched a lot e.g. routine office workers are frequently researched in order to test out theories of stratification as some identify them as w/c and some as m/c
- each sociologist who wants to disprove or add to earlier research on classification has to reseach yet another bunch of routine office workers

19 of 52

Reviewing the Field

- reviewing and critiquing existing data and literature is an important feature in any sociological report
- it requires the researcher to spend time reading articles, publications and other sources of information alredy produced on the subject
- the researcher then analyses this material to help clarify the issues around the subject
- reviewing the field gives the researcher useful information on the types of methodology used in previous studies
- they can see whether specific methods, e.g. structured interviews, worked in the past
- they can see if research samples were big enough, and form ideas about how big their own sample should be

20 of 52

Research Questions

- researchers need to narrow down the focus of their research so they don't spread their work out too thinly and end up with not enough detail
- they do this by coming up with a single research question that their research aims to answer
- a good question should focus on one part of the topic, and it should be clear and easy to research
- questions should be unbiased and should not suggest potential social changes
- for example, "should governments provide vocational education to 14-year olds?" is not a good question because it asks for a value judgement on social policy
- "what are the attitudes of employers, parents and teachers towards vocational education for 14-year olds?" is better

21 of 52


- a hypothesis is a statement that makes a prediction
- it acts as a starting point for research
- the research will aim to either show that the hypothesis is true or false
- it states a relationship between two factors e.g. "sociology teachers wear corduroy trousers" or "material deprivation causes educational underachievement"

22 of 52

Terms like "Democracy" need to be Operationalised

- sociology often gives names to concepts and ideas that aren't easily explained or measured e.g. it's tricky to measure things like 'democracy', 'development' and 'culture'
- you end up measuring these concepts by measuring something else that's linked to the tricky concept which sociologists call an indicator
- this is called operationalising a concept and it means making it operational, or workable, by finding a way to measure it
- researchers do this every time they conduct a piece of research, because you can't research something it you can't measure it
- each difficult concept needs an indicator, e.g. electoral participation or diversity of electoral results for democracy
- researchers need to be able to justify how they operationalised their concepts in their final report
- this is often a subjective process and the way a researcher operationalises may be criticised by other sociologists

23 of 52


- when sociologists try to combine different methods or data to get the best out of all of them
- triangulation gives a more detailed picture than when you only use one method, so it's more valid
- when you triangulate, you can check different sets of data against each other 
- triangulation combines strength and weaknesses of different types of data
- it can be expensive and time-consuming to do the same research by lots of different methods
- sometimes it's not possible to use triangulation e.g. when there's only one viable method to get the data

24 of 52

Pilot Study

- a pilot study lets you test the accuracy of your questions, and check if there are any technical problems in your research design
- you can use them to make studies more valid and reliable, test how long the research will take, and train your interviewers
- though they can be time consuming, expensive and create lots of work, they show that the project is feasible, and can help you secure your research funding

25 of 52

Social Surveys

- social surveys collect information about large target populations (group of people being studied e.g. women over 50), using questionnaires or interviews
- tend to be used by positivitists as a primary source of quantitative data which can be analysed to discover overall patterns and trends
- reliable, so they're used by government agencies and research companies

26 of 52


- too expensive and time-consuming for sociologists to survey the whole target population so they select a sample
- if the characteristics of the sample reflect the characteristics of the target population (with similar proportions of people in terms of age, class, ethnicity and gender) then the sample can be said to be representative of that target population
- the extent to which a sample represents the target population is known as its representativeness
- if the sample is sufficiently large and representative, then it should be possible to make generalisations from it about the wider target population
- the extemt to which you can accurately do this is the sample's generalisability

27 of 52

Representative Sampling

- involves picking names out of a 'sampling frame' (a complete list of the population being sampled, which needs to be accurate, complete and without any duplicate entries)
Simple Random Sampling
- names are taken completely at random, e.g. randomly selected from a list by a person or a computer, so each member of the population has an equal chance of being selected
Systematic Sampling
- involves choosing a random starting point in the sampling frame and selecting every nth value but there may be bias is there's an underlying pattern in the sampling frame
Multi-Stage Sampling
- selecting a sample from within another sample and it's often used to select samples for opinion polls to measure voting intention
- first, a selection of constituencies is chosen to represent the whole country, then postcodes within that constituency are selected, the houses from those postcodes

28 of 52

Representative Sampling (2)

Stratified Random Sampling
- population is put into segments called 'strata' based on things like age, gender or income, and names are selected at random from within each segment
Quota Sampling
- similar to stratified random but its not random
- the selection is made by the interviewer, who'll have a quota to meet e.g. interview 20 women aged between 25 and 34
- interviewers tend to pick people who look 'nice', which introduces bias (however it is quick and useful)

29 of 52

Non-Representative Sampling

- some target populations may be difficult to access (e.g. criminals, very old/young people) or the characteristics of the population may be unknown
Snowball Sampling
- finding initial contacts and getting them to give you more names for your research
Purposive Sampling
- when researchers select non-representative samples, often in order to falsify a hypothesis e.g. feminist sociologists trying to disprove the idea that gender roles are determined by biological differences deliberately looked for samples where women's roles weren't different from men's roles, or weren't traditionally feminine
Opportunity Sampling
- used when researchers need to select a non-representative sample quickly and easily
- researchers can use captive audiences which are groups of people who are gathered together for another reason (e.g. a group of school children or office workers)
- researchers also go to public areas and select people who are nearby

30 of 52

Questionnaires (1)

- mainly used closed questions and multiple-choice answers
- some have open-ended questions
- the reliability and validity of a questionnaire depends on how it's designed:
- closed questions give you quantitative data, which positivitists like
- open-ended questions can give you some insight into meanings and motives (qualitative data, inpretivists)
- use clear, simple questions which are easy to understand
- clear instructions and make it easy for the respondent
- have a nice layout that doesn't intimidate people
- give a range of options on multiple-choice questions
- measure what you want to measure
- ask embarrassing, threatening or complex questions
- ask 2 questions instead of 1
- be too long
- use terms that only a few people will understand
- lead the respondent to answer a question in a particular way

31 of 52

Questionnaires (2)

- can be used to investigate topics such as TV viewing habits, purchasing habits, voting behaviour and experiences of crime
- the Crime Survey for England and Wales (CSEW) is a questionnaire that is carried out continually by the British government
- They survey about 38,000 people a year and publish new results annually
- the British Social Attitudes Survey is carried out annually by the National Centre for Social Research
- each year they select around 3000 British adults at random and send them a questionnaire

32 of 52

Advantages of Questionnaires

- easy to administer, and they can collect a lot of data in a short time
- closed questions provide quantitative data which can be quickly analysed too
- they are reliable
- are anonymous and don't require the respondent to sit face-to-face with an interviewer, which makes them suitable for sensitive topics
- for example, the National Survey of Sexual Attitudes and Lifestyles was a postal questionnaire rather than a face-to-face structured interview
- a large sample can be given a questionnaire, so if the sample is representative the questionnaires should produce represntative data that can be used to make generalisations

33 of 52

Limitations of Questionnaires

- respondents may not tell the truth - they may lie or they may be mistaken
- questions may be misleading or mean different things to different people
- this means they may not actually measure what you want to measure
- respondents can't give any extra information, even if it would be really helpful to the researcher
- because the respondent fills in the questionnaire on their own, there's no one there to explain the questions if the respondent doesn't understand them
- postal questionnaires have a low response rate and if it's too low it won't be a representative sample

34 of 52

Structured Interviews

- Structured interviews are like questionnaires given to individuals or groups, except an interviewer is present to ask the questions
- structured interviews ask the same closed, multiple choice questions
- they give quantitative data, they're very reliable and are used in large-scale social surveys
- the main advantage over a postal questionnaire is that the intervieweer can explain and clarify the questions
- most structured interviews get a much higher response rate than questionnaires and people tend to agree to be interviewed (unless the research topic is sensitive or taboo)
- However, they're more expensive than questionnaires as you have to pay for the interviewer
- the interviewer has to follow the list of questions so they can't ask for more detail if the respondent says something particularly interesting

35 of 52

Unstructured Interviews

- informal, with no rigid structure
- flexible as they can be used to find out facts or attitudes
- good for researching sesitive issues where the interviewer has to gain the respondents trust
- they use open ended quesitons and give qualitative data (they're quite valid)
- the interviewer needs to have skill so they can probe to find out more detail about the interviewer's opinions
- they're used with smaller samples, which means they're not very representative
- it takes a long time to write up an unstructured interview as you have to write down a whole conversation, not just the answers to particular multiple choice questions

- pilot studies allow the researcher to find out what kind of question gets a substantial response
- they also tell the researcher whether they need to warm up with a gentle chat to gain rapport with the respondent before asking more meaty questions

36 of 52

Interviewer Bias

- resopndents in interviews may give the sort of answer they think the interviewer wants to hear (or the exact opposite, if they're feeling uncooperative)
- interviewers can give subtle direction towards certain responses, often without realising they're doing it
- Becker (1970) suggested that an aggressive interview style could actually uncover more honest responses that a participant might otherwise have kept to themselves
- these are known as 'interviewer effects' or 'researcher effects' which make the data less valid

37 of 52

Experiments (1)

- experiments are used by natural scientists (e.g. biologists, chemists etc.)
- the researcher starts with a hypothesis and they use the experiment to test it out
- all the variables are kept constant apart from the one you are interested in, the independent variable
- scientists change the independent variable and observe the effect on the dependent variable
- for example, if you were testing the effects of temperature on electrical resistance, temperature would be the independent variable which you control and electrical resistance would be the dependent variable which you measure
- the results are turned into numbers, the scientists looks for patterns and cause-and-effects relationships
- this method has been developed and used by social scientists to look for social causes and effects

38 of 52

Two Kinds of Experiments

- lab experiments are done in a controlled environment where the researcher changes the independent variable, and observes the effect on the dependent variable
- the researcher usually uses a control group, which is left alone to see what happens if you don't do anything to the independent variable (this method is often used by psychologists)
- field experiments are a response to the criticisms of lab experiments
- field experiments take place outside of the lab in real social settings, and those involved are often unaware (this method is used by interpretivist sociologists)

39 of 52

Strengths and Limitations of Lab Experiments

- the researcher has control over the experiment
- you get quantitative data
- you can replicate the research

- it's hard to reproduce real social situations in a lab (lab experiments are artificial)
- it is difficult to isolate single variables (social behaviour is influenced by many factors)
- there are often moral and ethical issues in lab experiments
- people may feel intimidated or act differently in the lab

40 of 52

Strengths and Limitations of Field Experiments

- they're done in natural social settings and are more like real life
- they can show the hidden meanings of everyday social interaction

- you can't control the variables like you can in lab experiments
- if people know they're being studied they may change their behaviour
- there's an ethical problem in carrying out experiments when the subjects aren't aware that they are taking part in an experiment

41 of 52

Hawthorne Effect

- when people are more interested in something, they try harder
- they may try harder at what they're doing because they know they're being observed and want to appear in a good light (Hawthorne effect)
- people usually have an idea of what kind of response the researchers want
- people often either give the researchers the response they think they want or the exact opposite depending on whether they want to please the researchers or whether they want to be stubborn
- these effects mean data from experiments may not be valid

42 of 52


- in covert observation, the researcher doesn't tell the group they're being observed
- the British Sociological Association (BSA) advise that you should only use covert participant observation when there's no other way of obtaining the data
- for example, Nigel Fielding (1981) used covert observation when researching the National Front (a far right-wing political party) because he believed he would encounter hostility if they knew he was a sociologist
- overt observation (direct observation) is when the group is aware of the research and they know who the researcher is
- for example, Beverly Skeggs (1991) used overt observation when studying female sexuality among students at a college
- particiapnt observation is when the researcher actively involves themselves in the group
- non-participant observation is when the researcher observes the group but isn't actively part of the group
- interpretivists prefer observation because the researcher can get to the action
- it tends to produce qualitiative data that's more valid than data from questionnaires

43 of 52

Pros and Cons of Participant Observation

- participant observation gets the researcher right to where the action is so they can check out the dynamics of a group from close up
- participant observation allows you to research the workings of deviant groups
- the researcher gets first-hand insight of people in natural real-life settings
- if it's covert, people can't mislead the researcher

- the researcher may get too involved and find it hard to be objective
- overt research may influence the behaviour of the group
- covert observation may lead to illegal acts in they're in a deviant group
- you can't repeat the research so it's not reliable
- they may find it difficult to remember all the events and accurately record them
- there are ethical and practical problems in getting in, staying in and getting out of the group
- the research usually includes a small group so it's not representative of the population
- it is hard work, time-consuming and expensive

44 of 52

Pros and Cons of Non-Participant Observation

- in non-participant observation, the researcher isn't drawn into the group so they can be more objective about the group's behaviour
- if you want to observe deviant groups, you have to be very inconspicuous

- observing from the outside stops you from getting to where the action is 
- overt research may influence the behaviour of the group

45 of 52

Case Studies

- detailed investigations of a specific thing (e.g. one person, one group, one institution or one event)
- one particular kind of case study is the life history, which studies one person's whole life
- examples of case studies include Willis's (1977) study of one group of boys in a school and Venkatesh's (2008) study on the organisation and impact of one criminal gang
- interpretivists like case studies because they can provide very detailed data, and they can give the researcher great insight into the subject under investigation
- positivists dislike case studies as they aren't representative of wider populations, and so they can't be used to make accurate generalisations because of the small sample size

46 of 52

Focus Groups

- a focus group is a small sample, perhaps fewer than ten people
- the sample are put in a room together, and asked to talk about a particular issue or to try to answer a specific set of questions
- the discussion is observed by a researcher
- because this is more like a natural conversation, the subjects may feel more able to express themselves than if they were speaking directly to an interviewer
- sometimes the focus group is left alone and a video camera or audio recorder is used to record the discussion for later analysis 
- sometimes researchers stay with the group and take part in the discussion
- they use the focus group to conduct a group interview

47 of 52

Longitudinal Studies

- studies done at regular intervals over a long period of time
- they're often large-scale quantitative surveys and they tend to be used by positivists
- however, some studies like the TV programme Seven Up are more qualitative 

- you can analyse changes and make comparisons over time
- you can study how the attitudes of the sample change with time

- it's hard to recruit a committed sample who'll want to stay with the study
- it's hard to keep contact with the sample, which may make the study less valid
- you need long-term funding and you need to keep the research team together
- longitudinal studies rely on interviews and questionnaires which might not be valid or reliable

48 of 52

Ethnography Studies

- ethnography is the scientific description of a specific culture by someone with first-hand experience of observing that culture
- it was first used by anthropologists to study traditional societies 
- they joined the community, learnt the language, and noted their observations
- it is based on small-scale field work that tends to produce qualitative data
- it is valid because you can study behaviour in natural settings 
- you can use ethnography to see what a whole community get up to, or to find out just one individual's life history
- you can use all sorts of methods to get primary data, including case studies, focus groups, longitudinal surveys and obervations
- researchers may also analyse documents such as diaries and letters, which are secondary data
- ethnography is in-depth research which gives inside knowledge about a community
- you get a valid picture from ethnography, but it relies on the researcher's interpretations of what people do and say
- it's difficult to make generalisations from small-scale research, and it may not be reliable (it's difficult to reproduce)

49 of 52


- official statisics are a source of secondary data, and they're produced by local governments, central government and government agencies
- hard stats are objective as politicians can't fiddle with them (statistics on births and marriages are hard stats)
- soft stats are more objective as politicians can fiddle with them (stats on crime, poverty and unemployment are soft statistics. in the 1980s and 1990s, the government changed the method used to measure unemployment over 20 times)
- social trends was a collection of government surveys published annually until 2012, a great source of secondary data
- the UK census is a survey of every household every 10 years, required by law
- the crime survey for england and wales looks at victims of crime
- non-official stats are stats collected by organisations other than the government
- examples include TV ratings collected by the British Audience Research Bureau and surveys carried out by special interest groups, such as charities, or by other sociologists

50 of 52

Documents and Mass Media

- documents can be either on paper or in a digital format e.g. online text
- documents can be personal such as letters, diaries, autobiographies and memoirs
- they can also be official, like school records, health records, church records and social work records
- public documents are produced by charities, businesses and local government
- documents can be expensive (more to do with meanings, like a diary)
- documents can be formal, like official documents
- interpretivists prefer expressive documents because they're a big source of qualitative data
- content analysis is a method of systematically analysing a communication (e.g. a speech, film or letter) to understand its meanings
- it is often used to study the mass media
- documents can be difficult to understand if they're old, they may be fakes, or they may contain lies (especially personal documents)

51 of 52

Comparing Secondary Documents

- sociologists look for similarities and differences between secondary documents
- they can compare different times, different cultures and different groups within society by looking at secondary data
- researchers can analyse real social behaviour and make comparisons without having to set up artificial experiments
- Durkheim used this comparative method in his famous 1897 study of suicide
- he looked at the rates of suicide in different European societies
- he found that the suicide rate was consistent over time, but varied between societies and varied for different groups within society

52 of 52


No comments have yet been made

Similar Sociology resources:

See all Sociology resources »See all Sociological research methods resources »