Research Methods, Experiments, Questionnaires and Interviews

Primary data is info gathered first hand by sociologists for their own purposes. These may be to obtain a first hand picture of a group or society. 

These are the methods used to gather primary data: 

• Social surveys: These involve asking people questions in a written questionnaire or an interview.
• Participant observation: The sociologist joins in with the activities he or she is studying 
• Experiments: This can be in a labatory form or field experiment. 

The main advantage for using primary data is that the sociologist can find out precisely the information he or she needs. However this is often time consuming.

Secondary data is information that has been collected or created by someone else for their own purposes, but which the sociologist can then use for their own study.

Secondary data is a quick and cheap way of obtaining data. However secondary data is less likely to be subjective and may not answer the questions which the sociologist is looking for.

Quantitative data refers to data in its numerical form, an example of quantative data could be sales of a certain product in a shop, or attendance for the average student at a school. 

They frequently come in the forms of opinion polls or market research. For example the score for the presidential debate between Donald Trump and Hillary Clinton are a form of quantitative data. Or how many people take holidays abroad etc.

Qualitative data. By contrast, gives a feel for what something is like, for example what it feels like to get good GCSE results or for one's marriage to end in divorce. Evidence gathered by using participant observation aims to give us a sense of what it feels like to be a member of a group.

Qualitative data is mostly opinions or statements, unlike quantatitive data it is not expressed numerically. It can provide rich descriptions of peoples experience etc. 

Sociologists often aim to produce generalisations that apply to all cases of the topic that they are interested in. For example, if we were interested in educational achivement, we would ideally want a theory which explains results for all pupils and not just the ones who were being tested. But we do not have time or money to include every single student in the UK is this leads to us sampling.

Sampling is a smaller sub-group drawn from the wider group which we are interested in.

The basic purpose of sampling is usually to ensure that those people we have chosen to include in the study are representative, this means including the relevent amount of ethnicities the correct gender ratio, etc.

By making the sample representative it allows us to make generalisations. Which is attractive for postivists who wish to make general, law like statements around society.

A Sampling Frame is a list of all members of the population we are interested in studying. For example Young and Wilmott used the electoral register.

Sociologists use various sampling techniques to obtain a representative sample:

• Random sampling is the simplest technique, where the sample is selected purely by chance. So this could be names draw out of a hat. An advantage to this is that there is no researcher or self selection bias. A downside to this is that it is unlikely to get a representative sample.
• Quasi-random or systematic Is where ecery nth person in the sampling frame is included, Young and Wilmott used every 36th person off the electoral register.
• Stratified random sampling is where the researcher first breaks down the population by class, age gender etc then proportions are made and used in the sample, for example if 20% of the population are under 18. 20% of the sample will be under 18.
• Quota sampling The population is stratified as above, and then each interviewer is given a quota of which they have to meet.

There are also reasons to do non-representative sampling this could be that the social characteristics are unknown, may be impossible to find a frame or refusing to participate.

Snowball sampling is where individuals are collected and then from these respondents the researcher asks for another potential individual who will take part and so on.

In education we can look at five main groups which make it easy or difficult to study. Pupils, Teachers, Parents, Classrooms and Schools. Each one of these raise particular problems and opportunities for the sociologist to choose a suitable method. 

Researching Pupils 

In education, many of the people that sociologists study are children and young people, pupils and students . Malcolm Hill 2005 suggests there are 3 big differences in studying young people and adults. Power and status, ability and understanding and vulnerability.

Power and status. Children and young people have less power than adults, this makes it mroe difficult for them to express their opinions openly especially when challenging adults. This is definitely true in schools as they are hierarchial institutions that give teachers higher profile than kids. Formal interviews or questionnaires also tend to reinforce this as it is the researcher who is determining the questions and how the answers should be formulated. Sociologists need to overcome this power difference. One way they do this is by using group interviews as opposed to 1 to 1s.

Ability and understanding 

Pupils vocab, powers of self-expression, thinking skills and confidence all play a role when a study is being  carried out. When abstract concepts are a central part of a study major problems are faced for researchers. They will have to simply word their questions so the respondents have a good understanding of the question.

It also makes it difficult to gain informed consent. This is because the sociologist may not be able to explain the nature of the experiment appropriately to the children. Young peoples memory is also less developed than adults making it hard to recall events which the researcher needs to know.

Vulnerability

Young people are more vulnerable to physical and psychological harm than adults. This raises ethical issues for the researcher. 

Teachers often feel overworked and may be less than fully co-operative even when they want to be helpful. This may mean that questionnaires or other research methods are kept short. On the other hand as professionals, teachers are likely to be sympathetic to educatinal research.

Power and status  As we have already established, power relationships are not equal in the school. Teachers have more power and status due to their age, experience and responsibility in the school. They also have a legal responsibility and owe a duty of care to the choldren. The nature of the classroom reinforces the power of the teacher and a researcher may be seen as a tresspasser. However, teachers are not fully independent, even in their classroom. Heads, governors parents and pupils all restrain what teachers can do. A researcher will have to develop a cover if they intend to carry out covert investigations and this may mean representing themselves as a supply teacher.

Impression management Evring Goffman looks at the fake impression which teachers put on when they are viewed by an outside source. Goffman also looks at how, as social actors  we behave differently when we are acting out a role on the front stage. Some researchers look at back stage such as the staffroom. However the staffroom is a very small social space and other teachers will be wary of what they say as one critical comment could affect their career.

The classroom is an unusual social setting with clear physical and social boundries. Although not as closed as a prison or psychiatric ward, for example, the classroom is less open than many other settings such as leisure centres or shops.

It's also highly controlled, activites, language dress and noise levels are all controlled and monitored. As a result this classroom behaviour may not accurately reflect what those involved really think and feel. However they can be regarded as easy to analyse as there is only two roles, the teacher and the student.

Gatekeepers These are people who control that social setting. These include head teachers, teachers and child protection laws. Making it difficult for researchers.

Peer groups Young people may be insecure about their identity and status. Therefore, when in school-based groups such as classes and friendship groups they may be more sensitive to conform and give different answers. 

There are tens of thousands of schools of many different kinds in the UK. If the sociologist uses observational methods, they are unlikely to have the time to investigate more than a few making it unrepresentative. Using large-scale surveys may overcome this problem as data can be gathered from more schools making it more representative. However they may lose detail which is gained when observing a single school. The researcher studying schools would need only a few minutes to identify their research population.

Schools own data  Schools are data rich places, they have many sources of secondary data, such as exam results truancy figures attendence rates etc. However it may be hard to obtain this data as it is usually classed as confidential. Also schools with high truancy rates may falsify their statistics to make the school look better, this would make a researchers findings invalid. 

The law in effect requires young people to attend school in order to be educated. They are legally required to attend. Just like the prison system. This means there is a "captive population" which enjoys both advantages and disadvantages for the researcher. One advantage woiuld be that the researcher will know where everyone should be at a given time. A disadvantage is that by researching in schools you are jepordising the schools main objective.

To understand education we must also look at parents. Parents can influence what goes on in education. For example 

• How they bring up their children 
• By their involvement in school through parent-teacher contacts, parent governors, parent evenings etc.
• Marketisiation policies encourage parents to see themselves as consumers, for example in choice of school.

However they aren't easy to study, they are not a single homogeneous group. Their class, gender and ethnicity may all effect how willing they are to participate in research. Like middle-class parents are more likely to return questionnaires compared to working-class parents. Making findings unrepresentative. Parents may engage in impression management, presenting themselves in a better light by trying to be involved more in their childs education.

There are two groups found in experiments:

• The experimental group With this group, we will change variables to try and examine a change. For example looking at plant growth, we might vary the amount of nutrients given to them to see a change
• The control group This is where all variables are kept the same and aren't changed such as the nutrients, this occurs as sociologists can then measure a difference. 

In layman's terms, a lab experiment involves changing the indpendent variable to see a change in the dependent variable and then measuring that change against a control group. Whilst controlling all other variables which may have an impact. 

Reliability The lab experiment is highly reliable, as every little detail can be replicated and be carried out by a other sociologist. There are two reasons for this, the experimentor can specify exactly what has been used and the researchers opinions or values have very little effect.

Practical problems Society is a very complex phenomenon. You can't use a lab for childs educational achievement. Another practical problem is that you can't look at past events. In addition lab experiments have small samples, making it difficult for large scale problems.

Ethical problems 

• Lack of informed consent
• Deception, we saw this with Stanley Milgram's obedience to authority 1974
• Harm, Again Stanley Milgram's test subjects suffered from real psychological harm after this took place

Hawthorne effect 

A labatory is not a normal or natural environment. It is thus unlikely that the test subjects will act usually as they know they are being studied.

Free will 

Inteprevists say as conscious humans our behaviour cannot be explained through a cause and effect relationship, they prefer field experiments which let us see humans as they are and not manipulating variables.

A field experiment has two main distinguishing features from a lab experiment. 

• Takes place in a natural surrounding
• Those involved are generally not aware they are being studied, this means no Hawthorne effect.

The researcher manipulates one or more of the variables in the situation to see what effect it has on unwitting subjects. We saw this with Rosenhans pseudopatient experiment. Where people were admitted to mental hospitals who weren't actually ill and were still treated like a mentally ill person. 

The comparative method This is where two groups are gathered who are alike apart from the one variable we are interested in and then the two groupsnare compared to see the difference. A main example of this would be Durkheim's study of suicide.