Sociology A2 Methodology

Set of revision cards for A2 sociological research methods, enjoy :)

HideShow resource information

Key definitions (1)

-Access: Being able to get in touch with the group/population that is to be studied, sometimes difficult when the group is closed or "deviant"

-Case Study: The detailed study of a social group/event/insitutution

-Casual Links: Way in which one thing may be associated with another e.g rising unemployment with boys' underachievement at school

-Casual relationship: The way one variable may result from another e.g boys' educational under-achievement may be partly the result of rising unemployment

-Cohort Study: Research on people who share a certain characteristic-often age; National Child Development Study was based on all children born in one week in March 1958

-Comparisons: To compare data or material from different studies

-Correlation: Tendency for one thing to be found with another e.g children who are in poverty do less well in education

1 of 72

Key definitions (2)

-Covert observations: Observation where the researcher does not reveal that they are a researcher

-Ethics: The moral rights and wrongs in relation to the conduct of research e.g those being studied should be assured of confedentiality, anoymity and the right to withdraw at any stage

-Ethnography: The study of people in their natural environment using qualtitive methods, particuarly participant observation

-Expressive documents: Diaries, letters, paintings and other secondary data that reveal the personal side of a persons' life

-Generalization: The ability to apply the findings of some research to the whole of the population being studied

-Go native: This occurs when a researcher becomes so involved with thr group being studied that he/she loses sight of their role as a researcher; associated with ethnographic research

2 of 72

Key definitions (3)

-Hawthorne Effect: When those being studied behave differently as a result of knowing that they are being researched

-Hypothesis: A statement of what the research will test

-Interviewer effect: The ways in which the social characteristics of an interviewer may affect the response of the interviewee

-Methdological pluralism: An approach that is usually in two stages, the method used being of equal status; a strategy often used  by realists who want to study different aspects of the research question

-Mixed methods: Using a combination of methods

-Objectivity: A scientific  approach to to the study of society free from bias and values of the researchers

-Observation: Studying people/groups in their natural environments: primarily used by in interpretivists but observation matrixes are occasionally used by positivists

3 of 72

Key definitions (4)

-Operationalize: To measure abstract concepts by defining them in research, for example by writing questionnaire questions

-Overt observation: When being observed are aware of the researcher's presence

-Participant observation: When a researcher blends into the life of those being researched e.g working on a building site to observe how construction workers behave

-Patterns: A consistent relationship between 2 or more variables e.g sociologists try to explain the patterns of educational achievement in relation to social class, ethnicity and gender

-Pilot study: A test run of a study to check for potential problems

-Positivism: A theory which suggests that the social world can be studied using the methods of the natural sciences; the behaviour of individuals and groups can be objectively measured

-Qualitative data: Data that can be expressed in ther form of words

-Quantative data: Data that can be expressed numerically

4 of 72

Key definitions (5)

-Random sampling: Sampling technique which gives everyone the same chance of being picked

-Rapport: Development of a relationship between the interviewer and the interviewee

-Reflexity: The willingness of the researcher to consider the implications of the methods they are using: the extent to which their values are affecting the research and what implications the research might have on those being research-a strategy used by feminists

-Reliability: The extent to which the research could be repeated and the same or similar results are obtained

-Representativeness/representative: The extent to which a sample is a fair reflection of the target population

-Researcher imposition: The way in which a researcher may impose his/her values on the research process; often used in relation to interpretivists and the anaylsis of data but can also be applied to positivists at the point of devising a questionnaire or interview schedule

5 of 72

Key definitions (6)

-Respondent validation: Most commonly used by interpretivists who ask those they are researching to look at their findings and give feedback on them in terms of whether they hace accurately portrayed the lives of the respondents

-Sampling frame: A list of the survey/target population e.g school register

-Semi-structured interviews: Interviews that combine elements of structured/unstructured interviews

-Social survey: A way to collect large amounts of quantative data in a logical and scientific way, which is considered to be objective

-Stratified random sampling: A sampling technique which, by dividing the target population into groups which the researcher is itnerested in studying, attempts to ensure that the sample if a fair reflection of the target population

-Structured interviews: Formal interviews based on pre-set questionnaires

-Target population: The group that the researcher is interested in studying and from whom a sample may be drawn

6 of 72

Key definitions (7)

-Trend: A tendencey or a direction, usually over time e.g there is a trend for more households to have access to the internet

-Triangulation: The use of more than one method, usually one of more that collects quantative data and one or more that collects qualatitive data

-Typical: Applied to a sampling unit where it could be described as similar tom or the same as, the majority of the target population

-Unstructured interviews: informal interviews, sometimes reffered to as structured conversations, in which the researcher is able to prove the interviewee for additional information and the interviewee can ask for clarification

-Validity: The extent to which the research is a true picture of the social reality of those being studied

-Value freedom: The view that sociology/sociological research should be as objective as possible and that the views of the researcher should not influence the research in any way

-Values: General principles/guidelines on behaviour; the things that members of society hold to be important e.g honesty, loyalty and personal achievement

-Verstehen: A term used by Weber meaning gaining an understanding of how those being researched feel, or seeing the world through their eyes; ethnographers aspire to achieve this

7 of 72

Methodology (1)

Key values in sociological research (1):

1. Validity: extent to which data is a true picture of the social reality of those being studied so that the research does what it intends to, important to see the meanings/motives they attack to their actions/actions of others, qualatative data is rich and detailed, need to have rapport with the group so they can achieve verstehen, studies that aim for vaidity are often ethnographic, and use observation/unstructured interviews in order to study groups like youths

2. Reliability: extent to which research can replicated by other sociologists and the same/similar results are obtained, factors relating to his include the sampling technique being transparent (difficult to replicate research if new sample pop. is very different from the origina sample pop.), whether the measuring instrument is standerdized (e.g questionnaires) and the way questions are in a structured interview are asked (must be the same), important for positivists who want to identify patterns/trends and to find correltations/make comparisons with other groups, positivsts argue that the research must be carried out logically and systematically in an unbiased and objective way

8 of 72

Methodology (2)

Key values in sociological research (2):

3. Representativeness: extent to which sample being researched are a fair reflection of the target population and are typical of those in the target population, positivist sociologists are keen to obtain a representative sample so they can generalize their findings, to obtain a represenatative sample researchers must use a random sampling technique such as stratified random sampling and ensure they avoid a skewed sample, good example is the ELSA study which has 12,000 respondents

4. Generalizability: extent to which it is safe to apply the findings from the research to the target population, in order to do this the sample needs to be both large enough and typical of the target population, positivists seek to make generalizations, but interpretivists are less concerned about this because they want to obtain an insight into the lives of those in the study

5. Ethics: BSA has guidelines that they recommend all researchers follow, however some ethnographers such as Patrick, argue that to do so would be to reduce or destroy their chances of gaining an insight into the lives of groups they are studying

9 of 72

Questionnaires (1)

-Way of carrying out a social survey, used by positivists to collect large amounts of quantative data in a logical/scientific way that can be considered objective

-Questionnaire is a set of standerdized questions, disrubted by hand/mail/internet, designed for self-completition and eveyone is asked the same question in the same order

Designing a questionnaire:

-Aim of research needs to be clear so that questions collect the data required

-Need for consistency in the use of terms, meaning should be the same for everyone

-Designers shoulds create clearly worded questions which are not ambigous, do not lead the respondent into a response and are not "double-barreled" (2/3 questions in 1)

-They need to avoid technical language

10 of 72

Questionnaires (2)

Types of questions:

-Closed: responses are pre-set

-Open: tend to ask how people feel about something, or why they did something

-Attitudinal scale: ask respondents to rank their answers on a numerical/rating scale, the critical issue is the problem of phrases such as strongly agree/disagree when the meaning of "strong" will vary from person to person

Pilot studies: used before questionnaire research, allow researcher to identify and resolve potential problems regarding the sample/access, method of distribution, wording of the questions, layout of questionnaire, language used, extent to which data obtained is what is required

-Researcher can ask those who took part for feedback as another way to "iron out" problems, it  saves time/money by sorting out problems before the main survey is conducted and can sometimes even help refine the research objectives

11 of 72

Questionnaires (3)

Advantages of Questionnaires in general:

-Relatively cheap/quick /easy                 -Pre-coded are easy to anaylse by computer

- Completed by resondents themselves, little work invovled

-People can complete in own time          -Statistical data allow for comparisons in patterns/trends

-Anonymous, so people may divulge sensitive info

-Avoids the "interviewer effect"

-Can be translated                                    -Can establish casual links

-Can be replicated                                    -No danger to the researchers

-Likely the same/similar results be obtained, data reliable

-Can be used in conjunction with other methods

12 of 72

Questionnaires (4)

Disadvantages of questionnaires in general:

-Socially constructed  so reflect the values/interests of the researcher

-Sample may end up skewed if the response rate is low, my mean results are not representative of the target population and it then becomes unsafe to make generilizations

-Respondent cannot ask for a question to be explained or expand on their response

-Researched unable to dig deep; often what someone says they do differs from the reality of what they actually do, so data is likley to be low on validity.

13 of 72

Questionnaires (5)

Advantages and disadvantages of Mailed questionnaires:

Advantages: Can be distributed across a wide geogrpahic area (e.g the British Household Panel Survey collects data from 5500 households, or 10300 individuals from 250 areas across the UK)


-Needs carefully-worded cover letter explaining purpose of questionnaire and guranteeing anoymity and confidentiality

-Resercher can never be sure who has completed it

-Response rate is often low, 40% is considered to be reasonable but it can be much less

-Respondents may be self-selected group, not typical of the target population

14 of 72

Interviews (1)

4 kinds of interviews (most done face-to-face, first 3 can be over telephone as well):

-Stuctured interviews: Quantative data, used by positivists who prefer a more structured and scientific approach

-Unstructured interviews: Qualitative data, used by interpretivists (find out about experiences/motives attached to people's actions) and feminists (minimize the hierarchal relationship between the researcher and the interviewee and empower them as far as possible)

-Semi-structured interviews: Quantative and qualitative data, no particular perspective

-Group/focus group: Qualitative data, interpretivist and feminist

15 of 72

Interviews (2)

1. Structured interviews: involves a set of pre-determined questions, all are closed, although occasionally there is a "catch-all" open question at the end where interviewees are asked if they have anything else they wish to see

Advantages of structured interviews:                                                                     

-Interviewer has control        -Data collected is likely reliable   -Less refusals form face-to-face

-Questions/procedures are pre-determined  -Answers are easy to record/anaylse

-Interviewer may be permitted to explain a question if it is unclear, this is not always allowed                                          

-Face-to-face allows researcher to see body language

-Interview questions are standerdized and therefore can be replicated

-If sample is collected systematically, able to make comparisons with other/same groups and identify patterns/trends     -If sample is large enough, generilizations can be made   

16 of 72

Interviews (3)

Disadvantages of structured interviews:

-Interviewer effect can occur (interviewee changes their answer in response to interviewers characteristics such as age/ethnicity/gender)

-If there is a team of researchers, all use different body language/tone likely to comprimise the study

-If more than one researcher is required, there will be a cost for training

-Time consuming

-Some interviewees may be put off by pre-coded questions and answers, may not have the answer they want to put/want to have some different questions

-Oppurtunity to misinterpret the questions/not able to expand on answers, will reduce the validity

-Refusals to answer may lead to a skewed sample

17 of 72

Interviews (4)

2. Unstructured interviews: is an open-ended informal interview, the interviwer will know what they want to cover in terms of topics but will allow a  "conversation" to develop and give the interviewee some control of the situation

Advantages (1)

-Can develop a relattionship with the interviewee based on trust, they may reveal more infromation making the data more valid (For this relationship to develop the interviewer needs good personal skills)

-Interviewer may achieve verstehen (Max Weber, an understanding of how the respondent feels)

-Interviewer may develop a rapport, making communication more relaxed and increasing the possibility of valid data (gaining rapport particularly helps when the subject matter is sensitive/personal)

-Gaining a relationship with interviewees means they are more likely to agree to attend times 

-Interviewee can ask for clarification on questions that are asked

18 of 72

Interview (5)

Advantages (2):

-Interviewer can probe for more detailed information, perhaps something they would not have thought of asking about

-Interviewer may obtain data not possible to acquire in other ways about people's feelings, for example on domestic violence

-Interviewer can sometimes learn about related issues from the interviewee

-Once a rapport has been developed and a relationship of trust secured, the researcher can go back to interviewees after data has been analysed to check they are portraying what has been said accurately

-Becker (1970) suggests that an agressive interviewing technique is useful for some topics, for example, hidden racist feelings may be revealed through confrontation

19 of 72

Interview (6)


-Interviewer needs to keep the interviewee focused without disrupting conversation flow/imposing own views on them

-Cannot be repeated; they are unique encounters, which means the data will be low in reliability

-"Conversation" will be guided by the researcher

-Gender/class/ethnicity of interviewer could lead to an interviewer effect, Labrov (1973) found that the race of the interviewer affected young black people in speech tests, interviewers might consciously or unconsciously lead respondents to preferred answers

-Sample size is usually small and this means the sample may not be representative

-If sample is not representative then it is not safe to make generalizations

-Volue of data collected takes a long time to analyse

-Risk at the analysis stage that researcher imposition will occur (i.e the views and values of the researcher determin the data they choose to use)

20 of 72

Interviews (7)

Group interviews/focus groups: A group interview usually covers quite a wide range of topics, whereas a focus group is usually concerned with a specific topic


-More people can be involved

-Members of the group can bounces ideas off each other

-Group can empower people, so Feminists could be supportive of this approach

-Researcher can obtain in-depth information

-Participants can reflect/rethink what they say, potentially making the data more valid

-Bryamn (2001) they emphasize the "joint construction of meaning", as focus groups can lead to greater probing of why people do things and they allow the researcher to observe how people construct meaning in groups, and in that sense Interactionists are supportive of this approach

21 of 72

Interviews (8)


-Can be difficlt to record

-Dominant participants may "take over" the conversation, giving a skewed idea of the the target population coming from one person

-Interviewer affect can still occur

-Participants may say what they think the researcher (or their peers) want them to hear

-Feminists argue group interviews reduce the hierarchial relationship between the researcher and thos being researched

22 of 72

Interviews (9)

Semi-structured interviews: 

-Semi-structures interviews obtain both qualitatuive and quantative data

-They are often used when resarchers wish to maximize the advantage of structured and unstructured interviews as well as minimizing their disadvantages

-Pope (2008) decided to use semi-structured interviews in order to combine a positivist approach with an interpretivist one

-Pope's research was an oppurtunity to find out about the backgrounds and experiences of femal football fans, and she was able to build a rapport with her respondents because she was young, female and a football supporter

-Using semi-structured interviews enabled her to "stay in control", and she also chose the method in order to be able to draw comparisons between respondents as well as enabling them to expand on their answers

23 of 72

Observation (1)

-Observation is carried out in a natural environment of those being studied, and this type of research is usually used by interpretivist sociologists who want to collect qualitative data from the point of view of those they are studying

-The researcher may take a role as a participant in the group ior stay on the outside for non-participant observation

-A key variation with participant observation is whether the group knows they are being observed (overt observation) or is not aware, as is the case when the sociologist either infiltrates the group or takes on an undercover role to enable the observation to take place (covert observation)


-Researcher may observe the group in their natural environment and gain detailed, rich data

-May gain insight into the group and the meanings they attach to activities

-May obtain valid data as a result of the depth of the data

-Mya gain ideas for further researcger by observing and becoming involved in the group

24 of 72

Observation (2)


-Need to gain access coudl take time, some sociologists gain access through a person already involved in the group or by an introduction from a "gatekeeper"

-Time and place of the observation will be fixed

-Research may take a long time and costs may therefore be high

-Researcher may get too involved and lost their ability to stand back-"go native"

-The Hawthorne effect, Henry A. Landsberger (1950): is where those who are being studied behave differently as a result of knowing they are being researched

-Reliability of the data will be low since it is very difficult to repeat observation

-Researcher imposition may occur when the data is analysed

-Sample will not necessarily be representative or typical of anyone but themselves

-Generalization will be unsafe

25 of 72

Observation (3)

Participant observation: In order to carry out participant observation, the social characteristics of the researcher (gender/age/ethnicity/class) need to be close to those of the group, especially if the observation is to be covert

-Potential ethical issues need to be considered, Hobbs (1988) decided to be involved with illegal activities in order to maintain his insider status with some of the groups in the East End

-The question of how involed to get and for how long then needs to be though through, the longer the involvement the more likley the researcher will "go native" or influence the behaviour of the group

-Whyte (1943) affected the street corner society he studied, they knew he was there and Doc, the leader, said that he now had to think about what Whyte would want to see and how he would explain it rather than just doing things like he used to

Structured observation: This is a systematic observation which generates quantative data and is usually used when the researcher knows what they are looking for and wants tor ecord the frequency of what is being studied

26 of 72

Observation (4)

Overt participant observation:


-Researcher will be able to ask questions when appropriate, Barker (1984) did this when she studied the religious sect known as the moonies


-Hawthorne effect may change the behaviour of the group

Overt non-participation observation:


-Consent can be obtained


-Lack of involvement may mean the researcher misunderstands what is happening

27 of 72

Observation (5)

Covert participant observation (1):


-Avoids suspicion by the group if it is engaged in illegal activities

-Data will be valid because the group will treat the researcher as one of them

-Hawthorne effect will be minimalized

-Can gain Verstehen

-Can develop a rapport with the members of the group

28 of 72

Observation (6)

Covert participant observation (2):

-Risk of going native

-Little if any chance of repeating the research, therefore reliability is low

-Getting out may be problematic

-Researcher will be unable to get informed consent

-It is unethical and may place the researcher in danger

-Recording field notes will be a problem

29 of 72

Content analysis (1)

Method of collecting quantative and qualitative data, it is primarily used to analyse the mass media and historical documents

1. Formal content analysis: Statisical excercise in which previously identified categories are counted and "ticked off" on a pre-coded grid, sociologists have used this to look at sex roles in children books

2. Thematic analysis: With this approach an area of reporting is identified and the number of times th theme appears is counted, Beharrell (1995) used this in his study of HIV/AIDS coverage in newspapers, and he was able to identify the different type of coverage in tabloid press compared to the broadsheets and suggests this reflected the biases of different types of papers at the time

3. Textual analysis: Involves a texual analysis in terms of the use of language that leads readers into particular interpretations of news and stories, it has similarities to semiology, which studies signs/symbols in order to decode and understand messages contained in the text, Glasgow University Media Group (GUMG) have used this in a number of their studies, including their analysis of the 1976 Longbridge car strike where they highlighted the different adjectives/descriptive words used to describe the strikers compared to the management

30 of 72

Content analysis (2)

4. Audience analysis: way of finding out the audience response to media content, and it enables the researcher to check their own interpretations against those of the audience, and doing this overcomes the other approaches that the audience are passive recipients of media messages

Research design decisions:

-There are a number of decisions that need to be made specific to content analysis, once the researcher had decided what strategy of the above four to use, they need to choose what media to include

-They should be careful that this does not simply reflect the values of the researcher

-If formal content analysis is to be used, the research questions must be clearly defined to ensure that the coding schedule does not miss revelant categories, and once identified these categories need to be formally set out

31 of 72

Content analysis (3)


-It is cheap; news coverage can be accessed at little or no cost from TV broadcasts, the internet and the newspapers/internet

-Allows researchers to make comparisons, by quantifying results they can compare news reporting or represenation across channels, newspapers, media, over time and between societies

-The coding system can be applied by anyone, this means it can be repeated and if quantiative data is being collected on a pre-coded grid, the same or similar results will be found, making it more reliable

-More than one of the approaches can be used in order to check the validity of the data collected

-Can also be used with other methods, such as focus groups

32 of 72

Content analysis (4)


-Coding categories are socially constructed; they are subjective so may be affected by the researcher's political biases which will influence findings

-Possible to misinterpret the meaning

-Problem of understanding what messages readers are recieving from a text

-Formal content analysis on its own reveals nothing about meanings, merely counts how often something appears in a programme

-Semiological analysis is entirely subjective and therefore very low in reliability

33 of 72

Secondary data (1)

Official Statistics: generally derived from comuplsory registration such as births and deaths, or large-scale government surveys such as the Labour Force Survey, and they usually cover all aspects of social life including education, employment, crime and health

-Positivists use official statistics because they provide "hard data", they argue that statistics are reliable because the research can be repeated and the same results found

Advantages (1):

-Easy to access from free government websites

-Up to date, many are collected annually or biennially

-Frequency of collection enables comparisons to be made year on year

-Comparisons can be made between social groups

-Patterns and trends can be identified

-The effect of government legislation can be tracked

34 of 72

Secondary data (2)

Advantages (2):

-Usually based on standardized measuring instruments such as questionnaires, therefore can be replicated

-Sample size is usually large, making it likely the sample will be representative, which means it is possible to make generalizations

-Often used as a starting point for social research

-Can be used to inform or support social policies such as health/crime initiatives

-Can allow a sociologist to reach groups whom they might not otherwise have access to

-Can be used to test a hypothesis (Durkheim did this with his study on suicide)

35 of 72

Secondary data (3)


-Based on questionnaires designed by the government; consequently questions may not be those a sociologist would choose to ask

-Concepts have been operationalized by the government, so may reflect their priorities

-No chance to prompt and probe respondents so validity may be low

-Can be manipulated, sata in some areas (employment/crime) can be collected in a way that makes the government appear in a better light

-Do not give a full picture (e.g crime statistics only include reported crime)

36 of 72

Secondary data (4)

Public/Personal documents: Either come from government sources (Ofsted/Hills report) or from pressure groups such as the Child Poverty Action Group, personal documents can come in many forms including diaries, letters, pictures, photos and audio recordings


-Usually cheap

-Easy to use, accessible

-Authentic true accounts so high in validity

-Provide detailed material, giving an insight into the lives of people who created them

-Historical documents are able to provide material that would not otherwise be available to sociologists

-Can be checked by other reseacrhers because they already exist in a written/visual/audio form

-Sociologists sometimes ask the people they are researching to keep diaries/make drawings

37 of 72

Secondary data (5)


-Many are one-off (e.g diaries) so there is nothing to cross-check against them

-Lack representativeness (e.g a historical diary will have been written by someone who is not literate, which was not typical of their time)

-Interpretation of documents is subjective, particularly if the written meaning is unclear

-Material may be partial, simply the view of the individual writers

-Writers may not be typical of other people

-Lack reliability, cannot be repeated

38 of 72

Secondary data (6)

Scott's (1990) four-step criteria to assess documents:

1. Authenticity: whether a document is written by the person it syays is is, is it genuine?

2. Credibility: whether the documents are free from error or whether they are subject to biases and distortions of the authors. e.g political memoirs of politicians like Tony Blair and Gordon Brown may offer very different interpretations of the same events

3. Representativeness: the extent to which documents have survived and are accessible, also if they are typical of the particular document being used

4. Meaning: the ability of the researcher to interpret the document

39 of 72

Mixed Methods (1)

-Positivists/Interpretivists have different views as to the best type of data to collect, but different methods are based on different sets of assumptions and there is no one true method that can be said to be the most "advantageous" in identifying the "truth", in pratice using a mixed methods approach is quite common (debate over what the source of ture knowledge is known as the epistemological debate)

-Mixed methods, multi-methods or multi-strategy research is used to improve the quality and validity of the data, according to Bryman (2001) this includes facilitation, triangulation and methodological pluralism

1. Facilitation: use of a method at the start of research to help formulate the research question or hypothesis which the rest of the research will then be based on

2. Triangulation: use of more than one method, usually one or more that collects the quantitative/qualitative data to: Cross check data for accuracy, re-examine the data, cross check the data for consistency, counter the bias of the data of a single method and increased confidence in the findings

-Can also happen naturally e.g Newson's study on parental discipline had mother who in interviews did not agree with smacking, yet observation showed them hitting the children

40 of 72

Mixed Methods (2)


-Increases the accuracy of the research data

-Improves validity (if someone has been economical with the truth or even lied in response to a question, an unstructured observation or interview can reveal what they really think)


-Assumes there is a "truth" to be found

-Different results could emerge from the different methods being used

-Researchers may interpret results of the methods used in different ways

-Even if the combination of methods being used seem to come to similar conclusions, another method could uncover different findings

41 of 72

Mixed Methods (3)

Methdological Pluralism: refers to a mixed methods approach where methods are combined to provide a fuller and more detailed picture of a topic under investigation, often one method is used first and the results of this inform the way other methods are used, it is a strategy often used by realists who want to study different aspects of the research question


-Can provide a fuller, more detailed picture

-Each method produces a "different slice of reality" (Denzin), therefore allows for a wider range of issues to be addressed

-Advantages and disadvantages of the methods can counter-balance each other

-Possible to identify from the questionnaire who to follow up with an interview

-Qualitative methods can help to explain the broader picture captured by quantative data

-Qualitative data can provide an examination of the small-scale phenomena that lie beneath the large-scale ones identified by macro quantitative data collection

42 of 72

Mixed Methods (4)


-Collecting and analysing data can be time consuming

-Analysis can be quite complex

-Lack of depth as individual methods may not be fully explored

43 of 72

Sampling (1)

Sampling: is the selction of a group to study from the target/survey population, when quantitative data is being collected a smaple should be chosen on the basis of being representative, and a large sample will mean the findings of the research will be able to be generalized

-Ten yearly census carried out for the government by the Office of National Statistics is based on the UK population as whole, it is carried out on the first year of every decade and all other research involves the sociologist choosing a sample of individuals or social groups they want to use for their research

-Various factors need to be considered by the researcher when selecting a sample, including: 

1. Identification of the target population

2. Identifying what sampling frames, if any, are available

3. Finding out how to access the sampling frame/target population

4. Understanding who is included in the sampling units

5. Deciding what size the sample should be and the sampling strategy to use

44 of 72

Sampling (2)

Sampling frames:

-A sampling frame is a list of the target population, for example the Electoral Register, the Postcode Address File (PAF), school/university registers and payroll lists

-However, there can be problems with sampling frames, which include:

1. May be out of date, e.g when people move out it may take them a while to update their address details

2. They may not include everyone in the target population, for example not all school-children will be on the school register because some are home-schooled

3. They may not identify ethnicity, gender or age

4. Many files may not be available due to confidentiality reasons

5. There are no lists of groups such as homeless

-In order to access sampling frames, a gatekeeper is sometimes required, this is someone like a head teacher who controls access to the sampling frame, in this case school register

45 of 72

Sampling (3)

Sampling techniques: Random/probability sampling: means everyone in the target population has the same chance/probability of being selected, seen as objective and scientific and used by positivists who want a representative sample

-Involves selecting a certain number of entries from a list (usually generated), systematic random sampling involves the selection of every nth name on the list

-Stratified random sampling favoured by researchers who want to obtain a truly representative sample, target population divided into social characteristics the researchers is interested in, the sample is then randomly selected from a list in that category, and the sample then shpould be a fair reflection of the target population, should be representative

-Cluster sampling: target population sub-divided, random sample selected from these sub-dvisions, then a further sample from those samples until the final sample of the required size is achieved

-Evaluation: Can be repeated and a similar group will emerge making it reliable method of sample selection, sample is large enough it will also be generalizable, however selecting sampling frames time-consuming, and if sampling frames fails to provide revelant info like age, stratified smapling is not possible, meaning sometimes skewed/biased samples will be created

46 of 72

Sampling (4)

Non-random sampling: This is more likely to be used by interpretivists who are more interested in finding specific types of individuals/groups than ensuring that their target is representative

-Quota sampling: this is a useful method if the demographics of the target population are known, people are selected by the researcher according to the category they fit into, and the number required from each category

-Purposeful sampling: used when a certain sample profile is required (Jackson in her study of lads and ladettes chose her sample carefully to get a mix of social class, race and ethnicity)

-Snowball sampling: used when a group is difficult to access, involves a contact introducing the researcher to other people

-Oppurtunity sampling: involves the researcher asking people who happen to be about and "fit" the criteria to take part

-Volunteer sampling:often involves advertising for people who fit a certain criteria to get in touch

-Evaluation: Useful when there is no sampling frame/when a specific type of person is wanted or group is difficult to contact, but it is unlikley to representative (people "fitting" critieria is subjective)

47 of 72

Longitudinal studies (1)

Longitudinal studies: carried out over a period of years in order to study changes/developments over time, can be used to establish correlations and casual relationships, can take various forms like panel studies and cohort studies

-Panel Studies: data is collected from a sample selected from sampling frames such as the PAF, sample unit may be individuals/households/organizations, example of this is the British Household Panel Study (BHPS) launched in 1991 carried out anually with a sample of 5500 households, interviewed in the first year and then interviewed every subsuquent year

-Cohort Studies: based on people with the same social characteristic often their age, and it may be a whole group or sample of a group, first cohort study was the National Survey of Health and Development was carried out on all babies born on the first week of March 1946

-Methods used to collect: usually collected by questionnaires or interviews, although some researchers have used ethnographic as well as mixed methods, and some participant observation takes 3/4 years and could therefore be described as a longnitudinal study

-Decisions of research design: Lenght/scope of study means cost is high, so selection of intial sample is highly important, decisions about how it will be financed/how data will be collected also needs to made, attrition must be considered (attrition=participants dropping out)

48 of 72

Longitudinal studies (2)


-Changes in opinion/attitudes can be tracked over time

-Correlations can be identified (such as links between social characteristics)

-Possible to make comparisons with data from other studies

-Data from a study can be used to inform policy making

-Data can be quantative/qualitative or both

-In-depth data can be obtained, even though some studies are questionnaires after a period of time the data becomes increasingly rich and detailed and validity is obtained

-A relationship of trust can be established as the respondents feel at ease with the researchers (even if the study is conducted by post)

49 of 72

Longitudinal studies (3)


-Time consuming and expensive to run

-Access to an appropriate sample is sometimes problematic

-Attrition will occur, people die, drop out or simply move, and this can lead to a distorted sample and skewed results

-If attrition of the sample is high representativeness may be lowed and generalizability will be unsafe

-Researchers could "go native" or become too close to their sample and comprimise findings

-Hawthorne effect can occur as respondents build a relationship with the researchers, they might know what might be expected and either change their behaviour or answers as expected

-Attrition of researchers, some people change their jobs or simply retire

50 of 72

Case Study (1)

Case studies:

-Detailed in-depth studies of a group, an individual, an organisation, an incident or an event or indeed any single example of what the researcher wants to investigate

-Case studies may stand alone or be part of bigger research

-One characteristic of a case study is that the individual, the group or the event maintain some individuality, they are not "lost" as part of larger research

-Case studies use a variety of methods, many are based on the collection of qualitative data such as observation or unstructured interviews, others the combination of quantative and qualitative data to gain as wide-ranging an insight as possible of the "case" being studied

51 of 72

Case Study (2)


-Can act as a start to more wide-ranging research, allowing the researcher insight into the area being studied and obtain ideas for further research

-Might throw up facts that need further research, can be used to generate a hypothesis

-Provide an insight into a  "case" which others might find difficult to access

-Can be abducted by a single researcher (in some cases only one researcher can gain access)

-Tend to be high in validity, range of methods used means data can be cross-checked and time spent with a group means in-depth data can be collected

-Researcher can develop a rapport with those they are studying

-Possible to gain verstehen, researcher is likley to spend a long time with the group/individuals and likley develop a sensitivity to their situation

-Can be useful to do a case study within a large piece of research, gives increased insights

52 of 72

Case Study (3)


-Foucs on particular groups and do not claim to be representative, however they can be used as a starting point for other research

-Even though case studies are unlikley to be based on a representative sample, the sample may be typical of others

-Since the sample is not representative it is not safe to make generalizations

-Platt (1993) argues there is adanger that even experienced researchers begin to make generalized comments when it is not safe to do so

-Given the close relationship usually developed between the researcher and the researched it would be difficult to repeat the case study

-Risk of the over-involvement of the researcher, of them "going native"

-Problem of researcher imposition, although this might be solved by respondent validation

53 of 72

Values and sociological research (1)

-Debate between positivism and interpretivism between whether sociology can or cannot be scientific, underpins the debate about values and value freedom in other words the extent to which sociology can be objective, two main alternative positions:

1. Sociology can be value-free: Positivists argue for sociology to have credibilty it must be scientifically objective and value-free, many 19th century founders of sociology agreed with this including Comte, Durkheim and Marx all believed that this was possible, Durkheim argued that social facts could be investigated and measured


-However, these sociologists were informed by their views of the world, Marx bourgeosie and proletariat, Durkheim believed that inherited wealth should be abolished and Weber argued that large bureaucracies were not helpful in terms of people's freedom and creativity

-Bowling (2004) argues that it is "naive" to assume objectivity is achieved in any field of social science

54 of 72

Values and sociological research (2)

2. All research is value-laden: advocates of this position argue that values are present from the start of the research process, and values can be said to have an impact at each stage of the research process:

1. Intial idea: Researchers choose to research areas of social life of personal importance to them

2. Choice of method: Quantative methods are sociall consructed and informed by the views/values of the researcher who decides what questions to ask and who to exclude, however positivists argue that equally interpretivists selectively choose what to write about based on their values/views

3. Control of research by a funding body: Gouldner argues that if someone is paying for research they had a reason to want it done

4. Career aspirations of researcher: Gouldner argues that academics need to get published in order to develop their carrers and therefore they operate within constraints of the dominant view

5. Analysis of results: Issue here is the researcher imposition when analsying results

6. Publication of results: Journal/Publisher/Organization willing to public the results tends to share the same values as that of the researcher

55 of 72

Values and sociological research (3)


-Even those who argue research is value-laden point out that it is possible to be honest/transparent about the views of the researcher, and Gouldner argued this was vital if the research as to have credibility

-Reflexity is a key factor, relates to willingess of researcher to consider implication of the methods they use, extent their values are affecting the research and what implications the research might have on thos being researched, feminists are particulalry concerned that this should be the case

-Weber acknowledge the intial idea might be based on the values of the researcher but argued that thereafter it is possible to carry out studies in an objective and value-free manner, he argued the intial idea would probably have "value revelance"

-Also possible to use a narrative approach whereby those being researched ban have their voices heard without being dominated by the researcher, both Oakley and McDowell did this

56 of 72

Values and sociological research (4)

Committed sociology:

-Argue that sociology should be more than academic research; it should attempt to change the world

-Becker (1967): "The question is not whether we should take sides, since we inevitably will, but rather whose side are we on?"

-Goulder argued that this approach by Becker led to the study of the "underdog" rather than those with power in society

-Feminists argue that research should empower the women being researched, whilst those working in the field of race and race discrimination argue that research should highlight the reality of racism for people and attempt to influence what is done about it

-Evaluation: Committed sociology raises questions about ethics, as individuals like Hobbs noted that he had to engage in illegal activities in order to remain an "insider", apart from this, there is the likelihood of over-involement of the researcher wit those they are researching "going native" and potential collusion with unacceptable views

57 of 72

Ethical issues (1)

-Ethics are concerned with the moral issues sociologists need to consider when embarking on research, guidelines issued by the British Sociological Association (BSA) in their published a Statement of Ethical Practice to guide it's members, and government bodies such as the NHS also have their own set of ethical guidelines

1. Informed consent: participants must be fully informed of the precise nature of the research and then willingly agree to take part, should always be told what the research is for, if they are under-age then parental consent is required, although this could mean children are taking part in research that they would much prefer not to

Issues with informed consent: 

-When they are fully informed Hawthorne effect could take place, comprimising research

-May be inapproriate to ask for full consent if research is ethnographic

-Some groups like the elderly are vunerable and may agree to participate without realizing the full implications

-It may not be possible to gain parental consent if children have run away because of abuse

58 of 72

Ethical issues (2)

2. Confidentiality: the participants should be assured of confedentiality, ensures relationship of trust is not broken and rapport can be built

3. Avoiding harm or danger of any sort: researcher should ensure that the participants will not be harmed by the research, they should also be careful not to put themselves in danger, and in this way covert participant observation can be particuarly problematic from the researcher's POV

Issues with confidentiality/avoiding harm:

-Researchers have a dilemma when harming themselves e.g taking drugs

-There is a further issue when the participants intended to harm others

-The extent to which research may harm others may not intially be apparent, Israel and Hay found that study of domestic violence in New Mexico was counter productive in that the male abusers became even more abusive when they discovered their wives had taken part in the research

59 of 72

Ethical issues (3)

4. Anoymity and privacy: this can be difficult when there are only a limited number of people in a particular role, e.g Hargreaves research when talking about certain members of staff was obvious because only so few staff members occupied such specialist roles

5. The right to withdraw: participants should be told of their right to withdraw at any stage

6. The need to report illegal activities: although most codes of ethics include this, there is also a section on the professional integrity of the researcher

7. Respondent validation: participants right to read what the reaseacher has writtenr before published can be problematic

-Extent to which codes of ethics should be strictly followed is contested:

1. Interpretivists argue that their research would be comprimised by the Hawthrone effect if they followed the BSA guidelines on informed consent

2. Strict adherence would also mean that the research of some deviant groups might never take, and understanding of these groups would therefore be lost

60 of 72

Positivism (1)

-Roots in Comte's 19th-century work in that he argued it was possible to use natural science methods to study the social world, Positivists believe there are social facts that can be observed and from these casual relationships can be sought

-Social facts are external to individuals, Durkheim looked at how the personal act of sucide was an objective social fact, positivists argue individuals are subject to external social forces like capitalism or competion in the job market

-Positivists also argued that human behaviour was predictable, and that it was possible to seek laws that govern people's behaviour, and that research questions/hypotheses can be tested

-Social scientists who favour this approach now tend to opt for a hypothetic-deductive or deductive approach, starting with a theory and testing it against evidence derived from the research

-Positivists through research want to identify patterns/trends, look for cause and effect, identify what is predictable, make comparisons over time from one group to another, look for correlations, conduct research that can be re-tested in order for it to be reliable, have objective and value-free research, aim for consistency and conduct large-scale data collection so it is representative and generalizable

61 of 72

Positivism (2)

-Positivists make a number of decisions before undertaking research; choice of topic (depends on funding source and this may not be a free choice), operalization of key ideas so they are measurable, establish a hypothesis/research question, identify an appropriate sampling technique to obtain a representative sample, identify a method that will obtain quantative data, deciding whether or not a pilot study should take place (most do due to research size), identify the locations these methods will be used, how to record/analyse the data

-Positivists seek quantative data, so use methods to produce and replicate this data like large-scale surveys, experiments, content analysis, comparative methods and official statistics

-Characteristics of methods of collecting quantative data: produce numerical data, graphs/tables/charts can be derived from data, making comparisons/idetification of patterns and trends straightfoward, tend to be high in reliability (so long as the measuring instrument is standerdized for the next researcher), can be large scale and if the right sampling method is used it will be representative of the target population, and this makes it generalizable data

62 of 72

Positivism (3)


-Argued that science itself it not value free/objective

-Humans have a free will and cannot be put in a laboratory

-All variables in the social world cannot be controlled

-Positivism ignores meanings and interpretations that people put on their actions, making it low in validity

-It can identify where two facts are related but these facts may not be the cause and effect

-Questionnaires are socially constructed, so are not objective of value-free

63 of 72

Interpretivism (1)

-Sometimes referred to as Anti-Positivism, Weber (19th-20th century) was one of the main influence on it's development, and Interpretivists believe that human beings are not inanimate and therefore cannot be studied in the same way as phenomena in the natural world, and that it is impossible to identify cause and effect because human behaviour is unpredictable 

-Social world is about meanings and human agency rather than social facts, people's have a variety of subjective view in society which positivism can't appropriately handle, findings are highly likely to be influenced by researcher values; it is impossible to be value free and the social world and the researcher affect each other 

-Also argue social world does not exist outside social realities people construct, so researcher's task is to uncover meanings/motives people attach to the actions of others, important to gain rapport in order to increase data validity, idea of "verstehen" comes from Weber and at a basic level means having an empathetic understanding of people while researching them

-Interpretivists want to move from intial impressions to deep understanding,also aim to be reflective (help to reduce issues like the Hawthorne Effect) and reflexity is also about the researcher standing back and considering the research from their own perspective and as far as possible from the perspective of those being researched

64 of 72

Interpretivism (2)

-Interpretivist research process: need to identify the group/issue to research, need to work with gatekeeper or use some other way to gain access to groups and inviduals, and there is the need for a method to obtain qualitative data where the research will take place, also need to consider ethics of their research

-Methods they wil use include participants observation and unstructured interviews (sometimes seen as better than observation because they can ask probing questions) which will increase the validity, and personal expressive documents may also be used

-Characteristics if methods used to collect qualitative data: data is in the form of words, give an insight into the meanings people attach to actions, allow people to speak for themselves, enable researcher to do research in the natural environment of the people they are researching without disturbing their day-to-day behaviour too much

65 of 72

Interpretivism (3)


-Those being observed may be affect by the presence of the researcher (Hawthorne Effect)

-Social characteristics of a researcher may or will affect the research

-Important to know exactly how the research affects those they are interacting with

-Hard work physically and emotionally; often research vunerable emotionally-needy groups

-Collecting qualitative data can be time-consuming

-Possibly the problem of researcher imposition and what data to select, which could affect validity

-Results will be low in relaibility, no means of replicating the research effectively

-Often a need for respondent validation, either with the individuals themselves or via other researchers

66 of 72

Realism (1)

-View that arguably bridge the gap between positivism and interpretivism, they argue that there are underlying social structures and these can explain observable events, and that it should be possible to study the meanings people attach to actions even if they are phenomena that cannot directly be observed (for example class consciousness)

-Believe sociology can be scientific because of the difference between open (variables cannot be controlled making predictions more difficult) and closed (variables are controlled e.g chemicals in a laboratory) systems as objects of study

-Realists believe that sociology is scientific if it follows an open system approach, sociology must deal with what is observable as well as the meaings people attach to actions, as there are social structures that lie beneath society's surface and these underpin the events that can be observed

-In carrying out research, realists aim to understand the structural mechanisms that cannot be seen (ideologies, false class consciousness and the reproduction of the class system), be as objective, systematic and logical as possible within the constraint that the social world is not always based on logical and systematic principles, and uncover underlying casual mechanisms that lead to observable events

67 of 72

Realism (2)

-Realist research: believe there is no such thing as theory-free data, believe that theories and observable phenomena cannot be seperated and that research should be based on comparison and evaluation of theoretical ideas

-Also believe that the purpose of research is to obtain data to test one theory against another theory, some aspects of society are very difficult to quantify because they are happening at a deep level

-Realist research process: decides on a problem, identifies the most likely theories to explain it and uses a range of methods to compare the theories with each other, often use a comparative approach, favouring mixed-methods research which breaks down division between the two data, also argue that research cannot be reliable because it is difficult to repeat, and Churton also argues this for scientific research (criticism against Positivism)


-Is realism just a new name for what a lot of sociologists have been doing for a while?

-There is no way of testing realist theories because they argue that sociological structures are like open systems in science; they cannot be observed

68 of 72

Feminism (1)

-Believe that most sociological research reflects a patriarchal society and benefits men, it is called the "malestream", and they have challenged the malestream for a number of reasons: research was focused on males/male activities and ignored/excluded women, aspects of social life such as domestic labour and childcare were ignored, malestream sociological methods focused on so-called rational and scientific approach, and this excluded female experiences

-There are a number of feminist methodologies:interview techniques break down hierachical nature of interviewer and interviewee and put an end to the power of the interviewer leading to a more collabarative approach and this will lead to more validity but means the researcher will have to answer questions asked by the interviewee

-Reflexity is a concept used by Roberts, wherby researchers are more open and honest with their interviewees and also think about their own experiences when conducting research, use of methods that collect qualitative data, argument is that quantative data is inconsistent with feminist values and objective of researching women's experiences

-However, shift of opinion by Oakley and others, and now suggest there is a willingness to use the "powerful" quantative methods of data collection, which some feminists argue is necessary in using methods that suit the purpose

69 of 72

Feminism (2)

Evaluation (1):

-Pawson argues there is no new method; feminists use the usual range of methods and there are no particular innovations

-Few researchers ultimately hand over power to the interviewees

-Reflexity is open to accusations of a lack of objectivity

-Some researchers become over-involved with those being researched, comprimises validity as they collude with the comments the people being resarched made

-Oakley did break down the hierarchial nature during her research with pregnant women, however (although she maintained contact with the women) she may have decived other into thinking she had become a real friend/confidante

-The relationship in ther interview may be affected by class and ethnicity

70 of 72

Feminism (3)

Feminist rearch topic:

-Should by "On, by and for women" (Stanley and Wise, 1983), decided before any decisions are made about how to conduct the research

-Revelant and sympathetic to women's experiences, able to contribute to the exposure of women's oppresion and help end it (e.g gendered domestic roles), also able to help at an individual level

-Political, meaning it should raise awareness of issues pertinent to women, such as child sex abuse and domestic violence, and ithe reseach topic should also be able to challenge male power and domination

Theoretical research issues:

-Theory must come from the research and not determine it

-Research should be more interpretive than positivist

-Should always raise new questions rather than viewed as an end in itself

71 of 72

Feminism (4)

Evalaution (2):

-Could be argued that the researcher has a priveleged ideologically-correct position, which may not be the view the interviewee wishes to take up

-Cain and other argue that men should not be excluded, as Stanley and Wise imply

-Many of the strategies feminists use are also used by male sociologists like Blackman and MacDonald, who have all done research with young people, including women

-Feminist research is likley to have high validity but be low in reliability, because the nature of the relationship between the researcher and the researched will not be the one that cna be replicated

-The people studied will be small in number and therefore may not be representative, making it difficult to generalize

72 of 72


No comments have yet been made

Similar Sociology resources:

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