Experiments - Lab Experiments

Experiments are more likely to be used by psychologists but are occasionally used by sociologists. They would be favoured by Positivists who want to take a scientific approach to studying society. There are various types of experiments:

A) Lab Experiments

B) Alternatives to Lab Experiments - field experiments, comparative method.


Lab Experiments:
An experiment is a method that will try and establish 'cause and effect' between two variables I.e. Plant growth (the dependant variable that is measured) and water (the independent variable which is controlled). The results of an experiment should enable a scientist to predict what is going to happen in the future.

Milgram - "Some conditions of Obedience and disobedience to authority"
Milgram was interested in the effects of authority on human behaviour. He set up an experiment, in laboratory conditions where volunteers were asked to co-operate in a learning experiment. The "teacher" was shown an apparatus which could deliver electric shocks to the learner. If the learner made mistakes with the linked words, they were to be punished by delivery an increased shock each time. The teacher was told to regard silence as wrong and to punish it accordingly. The teacher did not know that the whole experiment was a set up. Milgram found that 65% of the volunteers were prepared to administer the maximum shock to a stranger in these circumstances. The conditions were varied e.g. sometimes the subject was in a different room. But even when the teacher had to hold the learner's hand on the electric shock plate, 30% of the volunteers were prepared to do this.


Harvey and Slatin - "A lab Experiment into teacher expectations"
They wanted to look at whether teachers had preconceived ideas about pupils of different social classes. They used a sample of 96 teachers and each was shown a picture of children from different social class backgrounds. To control other variables the photos were equally divided in terms of ethnicity, gender etc. The teachers were then asked to rate the children on performance, parental attitudes to education, aspirations etc. Lower class children were rated less favourably. It was thought that teachers based their ratings on similarities between the children in the photos and children they had taught and these labels were then used to pre-judge the children in the photos. 


Advantages - 

• It accurately measures the effect of one variable on another i.e. authority on obedience. 
• It is easy to control variables in a lab situation. This makes sociology more scientific as it is using exactly the same method as the natural sciences. 
• Data is collected objectively and the sociologist's personal values do not influence the research. This makes the research easy to replicate, thus making it high in reliability. 
• Results can be easily quantified to produce hard, scientific facts, making this a method favoured by positivists.
• Because experiments measure cause and effect, social facts can be established which again make sociology more scientific. 

Disadvantages -

Practical

• Some situations cannot be repeated in a lab for example family situations.
• Participants may behave differently when in a lab (Hawthorn Effect).
• Can only do experiments on a limited number of people which makes it an unrepresentative method. 
• As we are dealing with people, it may not be possible to control all the variables in lab conditions. 

Theoretical

• Some Sociologists, in particular Interpretivists would argue that it is impossible to discover the meanings and motives of an individual when you are observing them in a lab.
• Sociology is not a science and it is impossible to treat individuals as if they are lab specimens. 
• If behaviour is being affected due to being observed then you could argue that lab experiments lack validity.

Ethical

• There are moral concerns about studying Human Behaviour in a lab i.e. the Milgram experiment.

Participant Observation

Participant observation is one if the main research methods used in ethnography. Ethnography is the study of the way of life of a group of people - their culture and the structure of their society. Some would say that what participant observation lacks in reliability it makes up for in validity.

Sociologists face several problems when conducting a participant observation study.

Getting in:

Getting in may depend on persons skills, having the right connections or chance. Polsky was a good pool player and this allowed him to gain entry into a gang of pool hustlers. Patrick was able to join a Glasgow gang because he looked quite young and knew one of it's members who he had taught at an approved school. Thornton's study of the British Rave Scene found that she had to befriend a girl called Kate to gain acceptance. Thornton found that her age and nationality were a barrier: "I began my research when I was 23 and slowly aged out of the peer group I was studying. Also, as a Canadian investigating British clubs and raves, I was quite literally a stranger to the scene."

Once a researcher is in a group they have to make a decision about what type of role to adopt. They should aim to disrupt the group's normal behaviour as little as possible.

Staying in:

Once in a group, the researcher has to be involved enough to understand the group fully but not too involved that they lose a sense of objectivity. One danger of staying in a group too long is that the researcher may become too involved and "go native". This happens if you stop being an objective observer and simply become a member of the group.
Punch in his study of the Amsterdam police found that in striving to be accepted he even started acting as a "policeman" himself, chasing and holding suspects etc...
The longer someone stays with a group, the less strange their ways will appear. As Whyte put it, "I started as a non participating observer and ended as a non observing participator.

Getting out:

In practical terms, getting out of the group poses fewer problems than getting in. If the worst comes to worst, the researcher can simply call a halt and leave. This happened in Patrick's study of a Glasgow gang. Sickened by the violence, he left abruptly.
Whyte found it difficult it communicate with academics back at Harvard university after so long in a different world. The researcher may also find that loyalty prevents them disclosing everything that they have found out in case if criminal prosecutions etc... Concealment of data will reduce the validity of the study.

Overt Participant Observation
This could be used if the researcher wants to avoid the ethical issues often found when doing covert participant observation such as lying to the target group.

Covert Participant Observation

This might be used if the target group is difficult to access for example criminal gangs would be extremely difficult to access. 

The observer might require that the target group doesn't know they are being observed so as to prevent the Hawthorne or Halo effect. 

Advantages of Participant Observation:
Validity - gains information from true, everyday life of the people being studied.
Insight - by actually living with a group, we get an insight into their way of life and viewpoints. (valid again)
Flexibility - the observer can keep an open mind and vary their own methods and in some cases their topic, judged entirely upon what they see. Whyte: "I learned answers to questions that I would not have had the sense to ask if I had been using interviews"
Practical Advantages :
- Might be the only method available of studying certain groups for example crime.
- Participant Observation allows the sociologist to build trust with the group they are studying.


Disadvantages of Participant Observation:
Practical disadvantages:
- Time consuming
- The researcher has to be well trained
- Stressful and demanding
- Age, Gender etc can restrict the groups as Downes and Rock say "Not everyone would pass uneventfully into the world of punk robbers and Hells Angels"
- Powerful groups such as government and mafia could potentially resist being studied. Therefore, observation can only really work in lower level groups of criminals.
Ethical problems - deceiving people and participating in illegal or immoral acts are serious ethical problems, people must consider.
Representativeness - In participant observation, the group studied is often small and selected haphazardly, this does not provide a good basis for making generalisations from the research.
Reliability - observations cannot physically be replicated by other researchers because they may not be able to get into the same group or may not ask the same questions to the group.
Bias and lack of objectivity - issues regarding concealing information or going native lead participant observation to be criticised as too subjective a method.
Validity - positivists argue that it isn't a valid method since rather than "telling it like it is" they tell it as they see it. Criticism also regarding Hawthorne effect.
Lack of structure -  Marxists and functionalists would criticise observation for concentrating on a micro approach and ignoring structural issues such as class or race inequalities.

Non-Participant Observation

One of the most famous studies of non-participant observation led to what is known as the Hawthorn effect. This was a factory where the managers were trying to increase the productivity of their workers. They brought in experts to look at how they worked and to try different methods to ultimately increase profits such as increasing and decreasing heat in the factory. It was found that whatever variables were controlled that productivity increased. It was concluded that just the fact that the workers were being observed was enough to increase the work-rate.


Some of the Advantages of Non-Participant Observations include:

• It is a simple and cheap method to use.

• If people are unaware they are being observed they will act as they normally do.

• It can be used to analyse social interaction in a wide variety of contexts.

• It can be a very valid research method given the correct circumstances. 

The Disadvantages of Non-Participant Observations:

• It relies heavily on the researchers interpretation of what is going on. 

• The researcher selects the actions they deem significant.

• The researcher may interpret things from their own ethnic/gender/class standpoint which can affect the validity of the research. 

• If people know they are being observed they will act differently than usual. 

Observation

There are several different types of observation:

Non-participant observation - The researcher observes the group without taking part. For example, using a two-way mirror to watch children playing.

Participant Observation - The researcher takes part on the event or everyday life of the group while observing it.

Further, separate distinction can be made between Overt Observation which is when the researcher makes their true identity and purpose known to those being studied. The sociologist is open about what they are doing.

Covert Observation - The study is carried out "under cover" The researcher's real identity and purpose are kept concealed from the group being studied. The researcher takes on a false identity and role, usually posing as a genuine member of the group.

It is not always possible for sociologists to fit neatly into each category. Whyte's study "Street Corner Society" involved him being semi-overt. He revealed his identity and purpose to a key member of the group, "Doc", but not to other members.

In sociology the most popular type of participant observation is unstructured and usually used by Interpretivists. However very occasionally Positivists use structured observation, usually non-participant, to systematically categorise what is happening e.g. how often a boy or girl in a reception class plays with a particular toy. This can then be used to produce quantitative data.

Interviews - The interview process

All interviews, whether structured or unstructured involve a social interaction between the interviewer and the respondent. The danger is that the respondent may be responding not to the questions but to the social situation in which they are interviewed.

Interviewer bias - This is where the interviewer may ask leading questions in order to prove a hypothesis, more than usual. Another source of bias is when the interviewer identifies too closely with the respondent.
Artificiality - Even the most relaxed unstructured interview is still an interview and as such, the conditions are artificial. It is doubtful in such circumstances that truthful answers can be obtained.
Cultural Differences - Different meanings may be given to different words. Or the culture gap may mean that the interviewer cannot tell when they are being lied to.
Status and Power Inequalities - In general, the bigger the status difference, the less valid the data.
Social Desirability - In interaction, people often seek to win approval. This may be even truer in an interview when respondents are on their best behaviour and give answers that present them in a favourable light. They may wish not to appear ignorant and may, if they don't understand the question, offer any answer rather than none at all. Unstructured interviews are more likely to have this effect.

Interviews - Unstructured (or Informal) interviews

These interviews are more like a guided conversation. The interviewer has freedom to vary the questions, their wording, order etc. from one interview to the next. Follow-up questions can be asked as needed such as Dobash and Dobash - "Violence Against Wives" who conducted 109 unstructured interviews with women who had experienced battering. All the women had lived in refuges and no-one asked had refused to take part. The interviews were conducted by two female researchers and varied in length between 2 and 12 hours. The women were asked to talk freely about the first, the worse and the last experience of violence. All the interviews were taped and when the women did not talk freely, they were gently probed by the interviewer. Dobash and Dobash were able to provide detailed accounts of the women's experiences and they also gained some understanding of the meanings and their experiences.

The advantages and disadvantages for unstructured interviews are shown below. Each of these has been previously covered.

Advantages of Unstructured Interviews:
Rapport and Sensitivity
The Interviewer's View
Checking and Understanding
Flexibility
Exploring Unfamiliar Topics


Disadvantages of Unstructured Interviews:
Practical Problems
Representativeness
Reliability
Quantification
Validity



Semi-structured interviews:
In between structured and unstructured interviews, lies the aptly named semi-structured interview. Each interview has the same set of questions in common but the interviewer can also probe for more information. Cicourel and Kitsuse always followed up their questions with "How do you mean?" or "Could you explain further?" as a way of gaining more information.


Group Interviews and Focus Groups:
These can involve up to a dozen or so people being interviewed at the same time. They will be asked to discuss a particular topic and their views will be recorded. These types of interviews have certain strengths and weaknesses compared to one on one interviews. Firstly, people may be more likely to open up in front of a group. It can stimulate each other's thinking and the interviewer can observe group dynamics. In terms of disadvantages, one or two individuals may dominate a group session, it can be hard to keep the group focussed on the discussion. Peer group pressure may lead some people not to voice their own opinions and finally data gathered is difficult to record and analyse what happens in these focus groups. One example is Willis whose study "Learning to Labour" used group interviews as part of his research into working class lads and their negative attitudes towards school.

Interviews - Structured (or Formal) interviews

These are very similar to structured questionnaires. The interviewer is given strict instructions on how to ask the questions. The interview is conducted in exactly the same way each time and the questions are pre-coded. Wilmott and Young used structured interviews in their research into the extended family in East London. Their sample consisted of 933 people. They could not do all the interviewing themselves so they employed other interviewers to help them. The interviews were formal and standardises and the questions, precise and factual. The interviewer's task was to ring the appropriate code number opposite the answer they received or, at a few points in the interview, write in a fairly short, simple reply. Each interview took between 10 minutes and half an hour.

Structured interviews are like questionnaires and involve asking a set of prepared questions that are closed-ended. These are then read out and the answers filled in by a trained interviewer. Due to the similarity to questionnaires they share many of the same advantages and disadvantages, the main differences come from the fact that structured interviews involve more interaction between the researcher and the interviewee.


Advantages:
Practical Issues - Training interviewers is relatively straightforward and inexpensive, however it is more costly than postal questionnaires. They can cover large numbers of people due to being cheap and easy to administer (though again, not as many as postal questionnaires). They are suitable for getting straightforward information about things such as job, age etc. The results are easily quantified which makes them good for testing hypothesis.
Response Rate - The large numbers you can survey increase the chances of a representative sample. People find it hard to turn down a face to face request unlike a postal questionnaire which is easy to ignore.
Reliability - If a method is reliable it can be repeated and get similar results. Structured interviews are seen as reliable as the questions are standardised and controlled by the researcher, they can be asked in the same order, with the same wording.

Disadvantages:
Type of Respondent - Those with the time and willingness to be interviewed may be untypical, this makes for unrepresentative data and undermines the validity of the finding.
Validity - A valid method provides an authentic and true picture of what is being studied. Structured questions restrict respondents, they may wish to say more but are not given the opportunity. The interviewer cannot explain questions or clarify misunderstandings. People may lie or exaggerate which may produce false data.
Inflexibility - The interviewer draws up the questions in advance to find out what they think is important. This may not coincide with what the interviewee thinks is important. This can affect validity. Like questionnaires, structured interviews only provide a "snapshot" of someone's life, it doesn't capture how people's views and opinions may change over time.
Feminist Criticisms - Graham argues that structured questionnaires and interviews are patriarchal and give an untrue view of women's experience. The researcher controlling an interview reflect the control of women by men in wider society. By not allowing the respondent to express their views, the structured interview doesn't allow women to express themselves. Feminists believe observation would be a better method to use.