Research Methods

Positivists believe that society can be studied using the method of natural sciences. They favour quantitative data that allows them to measure behaviour. They want their data to be high in reliability so that comparisons can be made and trends identified. By doing this, causal relationships can be identified (i.e. where one factor causes another), and correlations (a link between factors) can be made

Interpretivists favour qualitative data that allows them to find out why people behave in particular ways and to understand the meanings people give to their actions. Interpretivists are aiming for verstehen: an empathetic understanding of the world from the viewpoint of the respondents.

They want their data to be high in validity.

• Personal interest/experience

• Values of researcher e.g. topics they believe are important

• Knowledge gaps: if the topic has not been researched before

• Funding body: the organisation that funds the research will influence the choice of topic: e.g. the government will only fund research that is of interest or relevance to them

• Topical issues/current affairs/news events

• Practical issues: will it be possible to get any data

Data: information gathered as part of a research project

Primary data: Data that did not exist before the research began

Secondary data: Data that already exists

Quantitative data: Data that is expressed numerically

Qualitative data: data not expressed numerically (words + pictures)

Validity: data that is true or correct; it measures what it sets out to measure

Reliability: The method can be carried out again and get similar results

Representativeness: The data can be said to reflect the characteristics of the research population as a whole, meaning generalisations can be made

A survey is the systematic collection of data from a research population. The national census is carried out every 10 years and allows for trends over time and correlations between social groups to be identified.

The census is carried out by the government who have the time, money, power and resources to ensure every household completes it. Most sociologists lack these, so take a sample of their research population. They aim to get a representative sample which reflects the characteristics of the research population, allowing generalisations to be made.

Sample: small group of research population

Sampling frame: A list of names from which the sample is selected

Sampling unit: one member of the survey population

It is important to present all respondents with the same question because it results in comparable data, it is easy to replicate and is therefore is high in reliability.

A questionnaire should be kept as simple and as clear as possible because it should be easily understood by the respondent and should be easy to follow and therefore complete, this will result in high response rate and validity.

A pilot survey is where they test the survey out and see the flaws in the survey before they send out the real thing. 

 It might be useful for the researcher to keep the questionnaire anonymous because the respondents are less likely to conform to what they think the researcher wants them to be like, therefore making them seem better than they really are. This will improve validity and response rate. (Social desirability bias)

A leading question is, for example, “Why do you think sex before marriage is wrong?”. This encourages people to accept that sex before marriage is wrong because of the way the question is worded. This will decrease validity.

Advantages

High in reliability if questions are kept the same = can quantify results and measure behaviour. Therefore favoured by positivists as they can compare the data

Quick and cheap to do, compared to other research methods and get large amounts of data

Possible to reach a large sample over wide geographical area = more representative

No real ethical issues, especially if anonymous

Disadvantages:

Cannot be sure if it has been completed by the intended respondent = less valid

Leading questions could influence respondent

Absence of researcher for postal questionnaire may mean researcher cannot influence respondent’s answers

Low response rate = less representative data. Those who respond may not be typical at all of the target audience

 Can be time consuming waiting for response

 Incentives such as prizes can improve response rate but increase cost

 Respondents might lie = reduces validity

 Interpretivists criticise because the answers lack detail and verstehen(although possible with open questions)

Imposition problem = researcher chooses the questions and the answers that can be given which restricts the respondent

No researcher present – so respondent cannot clarify questions/misunderstandings = less valid

Participant observation: a method where the researcher gets involved with the activities of the group. It can be done covertly or overtly. 

Gatekeeper: a person who can allow or deny the researcher access to the group

Key informant: A member of the group with whom the researcher builds a close relationship and can give the researcher useful information

Going native: the researcher becomes so strongly attached to the group that they are no longer able to view the group objectively

• Interpretivists would favour participant observation, as the data is valid and can produce verstehen.
• It might be hard to gain access into the group, because of theresearcher’s own personal characteristics. Football hooligans are deemed to be racist, so if a person of colour wanted to research this group, they would not be welcome. Finally, it may prove hard to even get into the group, because you cannot just ask to join.
•  The data would be recorded qualitatively when the researcher gets home, which may cause problems, as the researcher would have to relyheavily on their memory – which might not be practical.

1. Simon, a detective in the CID acted as a gatekeeper between Hobbs and the police

2. Hobbs referred to the relationship as a ‘trading relationship’ because they both gave something to the relationship, where Hobbs would coach Simon’s son, and Simon would give Hobbs access to the CID.

3. Without this contact with Simon, Hobbs would have found it very difficult to gain entry into the world of detectives, as their business is strictly confidential and the information cannot be given to the public without reason.

4. Humphreys became the ‘watchqueen’ of the group he was studying, where he was the lookout for police and strangers, whilst the men engaged in sexual acts. This gained their trust, as he looked out for them.

5. Humphreys used covert research, because he could not overtly take part in what they were doing, as they would not trust him, as the acts the men were committing were illegal, secretive and personal.

6. One ethical problem with Humphrey’s study was that people being observed could not give their full informed consent, as they were being observed covertly and did not know the true identity of the watchqueen. Humphrey’s also deceived the people being watched, so that he could find out information.

7. An advantage for Barker in her overt research was that several Moonies felt that they could confide in her because she was organisationally and emotionally uninvolved. This could lead to valid data, as they are unlikely to lie.

Hobbs was concerned about being obtrusive because the police officers would’ve seen the obvious difference between Hobbs and themselves, and suddenly become more aware of the disparity between them and the reasons for why he was there, which would induce the Hawthorne effect.

He had to remind himself that he was an observer because he was not there to enjoy himself, but to conduct an academic inquiry. (could 'go native'

Getting out of covert observation with groups such as gangs, may prove dangerous, as they could track the person down. Moreover, in any situation, you would have to just disappear.

Practial Issues: 

 Time

Cost

Characteristics of researcher e.g. an adult male could not do covert observation on a teenage girl gang. Also, some researchers may be good at questioning or listening, making interviews a good method for them.

Choice of topic: researching on issues in the past often means only secondary methods can be used

Ethics are morals about what is right or wrong. There are five ethical issues to consider when doing research:

• Informed consent: the participants must be able to give their informed consent to take part in the research, being fully aware of the aims of the research, what they will be doing and the consequences, and how the data will be used once the research has been completed.

• Deception – information is deliberately withheld from the participants or they are deliberately misled about the aims of the research. If the participants are deceived, they cannot give their informed consent.

• Privacy: sociological research does invade people’s privacy: this is ethical if the participants know how far their privacy will be invaded.

• Protection from harm: the participants must not be harmed emotionally or physically, both during and after the research.

• Confidentiality: the identity of the participants is kept secret

Theoretical Perspectives: A positivist would rather use quantitative methods; whereas an interpretivist would rather use qualitative methods.

Sensitivity of Topic: Like Humphreys, he could only use covert observation, as the acts he was witnessing were illegal and personal.

Strengths:

• Interpretivists favour because PO allows verstehen as researchers see life from respondent’s viewpoint.

• Highly valid and detailed data

• May be only method for researching some groups, especially those involved in illegal activities

• Researcher can explore new ideas that emerge during observation which they hadn’t previously thought of

• Can see how group develops and changes over time

Weaknesses:

·         Covert PO = difficult to maintain covert role and to not reveal identity

·         Covert – researcher may have to take part in illegal activities to gain trust – could be dangerous

·         Going native

·         Covert PO – ethical problems of lack of informed consent due to deception; invasion of privacy, risk of harm etc.

·         Overt PO – could cause Hawthorne effect

·         Low in reliability (due to it not being standardised) so criticised by positivists. Very difficult to identify correlations

·        

Time consuming and expensive

Low in representativeness – only one group

Problem of recording data, especially covertly.

Hard to get out of the group

Hard to get in – have to get access through the gatekeeper

An unstructured interview is like a conversation. The interviewer's questions are influenced by the answers given by the respondent. The interviewer can probe the respondent further to get more detail and information. The researcher tries to build a rapport with the respondent: a friendly, relaxed relationship to encourage the respondent to trust the interviewer. This should mean the respondent opens up more and provides honest detailed answers, making this an excellent method for researching sensitive topics.

(LIKE PIERS MORGAN SHOW :))

Strengths

Favoured by interpretivists because answers can be very detailed, valid and can gain verstehen.

Fairly ethical

Interviewer is able to explore new ideas created by respondent’s answers

Very useful for sensitive topics

Interviewer can probe for more detail

No imposition problem – respondent decides to answer questions, rather than selecting from a list of options chosen by researcher

Questions and answers can be clarified to avoid misunderstandings

Weaknesses:

Interviewer bias could reduce validity

Interviewers need to be trained = expensive

Time consuming – small sample = low representativeness

Social desirability bias

Low reliability – questions determined by respondent’s answers and no structure: therefore, criticised by positivists

Possible ethical issues of invasion of privacy

Strengths

Respondent may feel more comfortable than being interviewed alone

The ideas of other group members could stimulate new ideas

Larger sample

Weaknesses:

Peer pressure to conform could influence answers - less valid

Researcher can observe group’s behaviour

 One or two people could determine the interview - others remain quiet

Documents include a range of sources which are made up of writing or some form of visual image. These can be of considerable use to a sociologist. Documents can be:

1.       Personal documents are first person accounts of social events, produced by the person who has experienced them e.g. diary, letter, blogs/social networking, text, photos, etc. These can provide a great insight into a person’s thoughts and emotions, but can be difficult to access, could invade privacy and are usually only one account, so not representative.

2.       Historical Documents are documents created in the past. These include many of the personal documents and also propaganda, paintings, advertisements, flyers, posters, poems, newspapers, radio broadcasts, diaries, maps, text of a speech, music, tapestries, books, novels, etc.

1. Authenticity: Whether the documents are genuine or forgeries (like Hitler’s diaries)

2. Credibility: Whether the author is sincere in what they say or write, or whether they have distorted the information to deliberately mislead the audience.

3. Representativeness: Many documents have been lost or damaged. Those that remain may not be representative of all the original documents.

4. Meaning:  Words change meaning or go out of use. The researcher may not be able to understand the meanings that the original author had intended.

Content analysis is a method of analysing a qualitative document by quantifying it. This is done by sorting the document into categories and counting how often each category appears. E.g. Ferguson used content analysis to study women’s magazines. She used categories such as:

• Self-help: overcoming misfortune

• Getting and keeping your man

• The happy family

• Heart vs. Head

She then counted up how often each category appeared, producing a quantitative measurement.

This method is fairly quick and easy and produces reliable data, but we do not learn how the audience understands the document.

In a structured interview a list of questions are read out by the interviewer, who then records the respondent’s answers. The interviewers must keep to the same sequence of questions for every interview. They cannot ask the respondent for further detail, but can give prompts (i.e. a further explanation of the question) if the respondent does not understand the question. The interviewer is trained to carry out each interview in the same way and must keep an ‘emotional and personal distance’ between themselves and the respondent to keep each interview the same.

1.       An interview schedule is similar to questionnaires, but the interviewer asks the questions and records the responses. It is a plan of the questions, so they can be read out in order of which the researcher wants. These are important as it makes the interview and the answers reliable because questions are kept the same for every respondent.

2.       It is important that the researcher and respondent have an ‘emotional and personal distance’ between themselves so that both parties remain objective to the matter at hand, which will heighten the validity as emotions do not affect the questions.

3.       Prompts must be carefully planned in advance to keep the interview the same with all the respondents and does not give any additional information that probes the responder’s answers further.

4.       The presence of the interviewer may affect the respondent as they might feel they cannot be completely honest and therefore give invalid answers.

5. 

6.

      The structured interview is high in reliability because the researcher has an immutable interview schedule, which can be carried out again and again, and result in similar answers.  Moreover, as the prompts are planned carefully beforehand, this increases the reliability, as no more details are being given to the respondent.  However, the interview is low in validity, as the presence of the researcher may affect the truthfulness of the respondent.

Strengths

Interviewer is present to explain the question if the respondent does not understand – unlike self completion questionnaire

Highly reliable: standardised interview gives all respondents the same experience, so able to compare results. (positivists would support this)

A trained interviewer can avoid leading questions, making answers more valid

Higher response rate than questionnaires which can improve representativeness

Favoured by positivists for quantitative data to measure behaviour

Fairly cheap and quick to do & fairly ethical

Interviewer bias is a problem with any type of interview. This occurs when the personal characteristics of the interviewer influence the answers given by the respondent. These characteristics include the interviewer’s age, gender, style of dress, tone of voice, etc. Interviewer bias can significantly reduce the validity of the data.

A white person being asked by a black interviewer about their prejudices will lie to the interviewer if they do in fact have the aforementioned prejudices, as they risk offending the interviewer and being arrested or being hit.

  A truant pupil being asked about their absences by an adult may lie, as they don’t want to be in trouble, as adults have power over children.

.

Weaknesses:

Interviewer bias – low validity as respondents may lie to please the interviewer

Not practical for sensitive issues as the respondent is having a one-to-one conversation in person and the sensitive topic may cause the respondent distress and they might not want to talk about it

Interpretivists criticise because answers lack detail so no verstehen

Inflexible: interviewer cannot ask for more detail from respondent if an interesting answer is given

Cannot get such a large sample or cover such a large geographical area to questionnaires

Not as cheap/quick as a questionnaire as interviewers need to be paid and trained

Non-P.O is a method where the researcher watches and observes the group but does not get involved. The researcher usually has a checklist of things they wish to look for; this produces quantitative data to measure behaviour which should be reliable to allow comparisons to be made, therefore, favoured by positivists. Interpretivists would criticise non-P.O for lacking verstehen: as the researcher is not involved with the activities of the group they cannot understand behaviour from the participant’s view.  

Non-P.O can be:

• Covert – the researcher identity is kept secret = less Hawthorne effect = more valid but unethical

• Overt – the researcher’s identity is known = more ethical, but Hawthorne effect likely = less valid

The problem Walford experienced was that he didn’t want to be treated as a visitor, as he wanted to see the true nature of the students at the college. Because of his overt observation, he induced the Hawthorne effect, which made the students act differently than they normally would, lowering the validity of the data.

This was not a long-term problem as four weeks later, the students simply forgot he was there and acted how they normally would.

Walford used non-participant observation as he could not simply join in, as he was much older than them.

Official statistics are produced by the government. They cover a range of issues such as: births, marriages, crime, education, health, unemployment, deaths, etc. They are a quick and cheap source of data because they are produced by the government; they are done on a large scale and therefore are representative. They are collected on a regular basis, allowing comparisons over time to be made, and between different social groups.

However, official statistics must be treated with caution. The Government can manipulate statistics to ensure the Government looks good. For example, between 1979 and 1997, the Government was able to show a decrease in unemployment. However, this was because the definitions of unemployment changed 30 times during that period. This significantly reduces the validity & reliability.

Official statistics may lack validity 

Crime statistics are a social construction based on a series of decisions made by various people involved.

For example:

Is the victim aware of the crime? No.

Does the victim report the crime? No.  

Do the police record the crime? No.

If the answer to any of these statements is ‘no’, the crime is not included in crime statistics. Therefore, crime statistics lack validity because they are not a true picture of crime – they ignore the dark figure of hidden crime.

Advantages

Large scale due to Government resources – representative

Easy to access and cheap

Not time consuming & no ethical issues

Used by positivists to compare/measure behaviour

Reliable if definitions are kept the same

Disadvantages:

May not be valid

Can be manipulated to show Government positively (political bias)

Interprevists would criticise because of the lack of detail

Definitions can change reducing reliability

No ethical issues

A longitudinal study is a study of the same group of people over a long period of time. This allows the researcher to see how the group has changed over time, but is expensive to do. Also, sample attrition will reduce the sample size, reducing the representativeness of the data.

1.       The sample of longitudinal studies become smaller as time goes on (sample attrition) this lowers the representativeness of the study.  Sample attrition may occur because people might move away, be too busy, lose interest, or die.

Advantages

Can see how group changes over time

Can identify causes of changes

Should be fairly valid as there is previous research to refer back to, reducing the need to reply on the respondent’s memory

Disadvantages:

Hawthorne effect could reduce validity

Sample attrition – reduces representativeness

Time consuming, so expensive and difficult to get funding

Triangulation combines qualitative and quantative methods to improve the validity and reliability of the research data.

Covert P.O = (Qualitative) – High validity low reliability

+

 Unstructured interviews (qualitative) High validity, low reliability

+

Questionnaires (Quantiative) Low validity high reliability

Lab experiments:

An experiment tests a hypothesis: a statement the researcher thinks may be true but has not yet tested.

The researcher sets up a controlled environment. The researcher uses two groups that are alike in every way: a control group and an experimental group. The researcher will change an IV in the experimental group to see if the DV being tested changed.

Any changes between the two groups can be seen as a result of changes to the IV because, beforehand, the groups were exactly the same.

This allows correlations between two variables to be identified.

The Hawthorne Effect:

The Hawthorne Effect occurs when participants change their behaviour because they know they are taking part in research. This can significantly reduce the validity of the data. (Sociologists may try to prevent the Hawthorne effect by not telling the participants about the research: however, this creates an ethical problem due to deception and lack of informed consent.)

Strengths

Controlled environment makes data high in reliability

Can identify correlations between variables

Favoured by positivists – realise quantitative data to measure and compare behaviour

Weaknesses

Some researchers may try to avoid the Hawthorne Effect by deceiving them – but this is unethical

The artificial environment may cause the Hawthorne effect, where participants change their behaviour, reducing the validity of the data

Gaining access to a laboratory and equipment can be difficult and expensive.

Small scale – low representativeness

Interpretivists criticise lab experiments because cannot understand why people behave and cannot get verstehen.

Field experiments are carried out in a natural setting. This should increase the validity of the data as people are in their normal setting but it reduces the reliability because it is not possible to control all the variables.

The difference between a lab and field experiment is that in a lab experiment, there is complete control over all variables. However, in a field experiment, you cannot control all of the variables.

Lab experiments are high in reliability, but low in validity, where field experiments are high in validity and low in reliability.