The 3 evaluation criteria of sociological research
Reliability - it is reliable if someone else is using this method or on another occasion and would produce the same findings. A survey using a well-designed questionnaire is reliable. The same results would be gained, regardless of who is asking the questions. Research that involves someone working alone and relying on their own interpretation - like participant observation - must be suspect as to its reliability.
Validity - it is valid when the data is a 'true' picture of what is being studied. A questionnaire may be well designed and produce reliable data but if the data is invalid, it is no use to the researcher. The data is a product of the research method, rather than an account of what is being studied. Participant observation - when done well, it produces valid data that reflects the reality of the situation, unaffected by the research method that is used.
Representativeness - how far the individual, group or situation being studied is typical of the rest of the population. Researchers who conduct quantitative surveys have various ways of selecting a sample and there are no sophisticated tools to assess how far a sample is representative of the whole population. Small-scale research must always be questionnable as to its representativeness.
Getting the balance right - Good sociological research often involves a trade-off between reliability, validity and representativeness. This is done by collecting data in different ways. Thus Barker (1984) in her study of the 'Moonies' used written questionaires and face-to-face interviews, joined a group (at their invitation) in their everyday life and for religious meetings and ceremonies and read all she could about their beliefs. She also compared the attitudes of the Moonies she worked with to the attitudes of non-moonies.
Sociology and values
Bias & objectivity - science methods should be objective, as well as its findings - not be influenced by the personal interests or bias of the researcher. It should discover facts and be value free.
Sociologists have argued for years about whether sociology is value free, can be value free or should be value free.
Values and bias effect:
- what to research
- who pays for research
- what research methods to use
- which questions to ask and to leave out
- who, what, where and how to observe
- what secondary data to study
- what data or info to record (from questioning, observation or reading documents)
- what to include in the research report
- whether the report is published
Sociologists today - sociology cannot be completely value free
Some argue - values will effect what topics a sociologist decides to study, but the research methods should be as objective and value free as possible. personal bias and opinions should not affect questions being asked.
some agree - some degree of bias and value is inevitable and that value freedom is a myth. the researcher should be open and honest about their values so the reader can take into account when assessing their work.
others agree - sociologists should take a committed approach to their work. defend the interests of the poor and the powerless, and to challenge the authority and power of dominant patriarchy.
postmodernists - there are no facts or objective truths and there is no way of deciding whether one account is more accurate/better than another.
ethics in sociological research
ethics is the study of what is morally right or wrong. the main ethical principles of social research are:
- no one should suffer any harm as a result of the research
- researchers should be honest and open about what they are doing
- participants rights to privacy and confidentiality should be protected
- no laws should be broken
Harm - there is a risk of harming someone emotionally (insensitive questions) or socially (damaging reputation or exposing someone to ridicule or punishment). ethical research is designed to avoid theses risks.
Honesty and openness - ethical researchers seek the informed consent of participants, ensuring that they have agreed to take part in the research and they they know:
- that the research is going on
- who is doing it
- why it is being done
- how the results will be used
it is not always straightforward to gain informed concent. e.g. young children or people with learning disabilities may not fully understand what the researcher is doing.
Privacy and confidentiality - all research participants have a right to their privacy. they have a right to know what it is about and to refuse to take part in it or to answer particular questions and is confidential if they do. ethical researchers are careful to disguse identity of individual participants when they write up their research. this is easy in a survey as all individuals may be anonymous in the first place and where individual repsonses are merged into totals. this is more difficult in a small group of people through participant observation.
Legality - researchers should not become involved in illegal activity. this can sometiems be a problem for a researcher studing criminal behaviour when asked to help with or take part in the activity.