The research methods used in Sociology are little different to those used in other social and political sciences. Common modes of enquiry include the use of social surveys: experiments, interviews and observation. Descriptions of their various strengths and weaknesses can be found in any introductory School or College text and Themes and Perspectives by Michael J. Haralambos is probably the most comprehensive of these. Yet what is more interesting is why particular methods are used or not used.
What is your theoretical position?
Positivist sociologists like Marx, Comte and Durkheim attempted to explain society by discovering universal laws. They sought to emulate the methods used in the natural sciences and by doing so they hoped to raise the status of the subject and to convince others about the worthiness of their political views. Those that follow this research tradition are referred to as macro theorists and they use methods that necessarily involve the generation and collation of quantitative data taken from large, representative samples.
Additionally, man is viewed as relatively passive. Powerful institutions and structures like the mass media, family, education and social classes, shape our identity and behaviour. This is important because without subscribing to a passive view of man, the quest to discover scientific laws is doomed to failure. If this belief was mistaken and we are all directed by our own consciousness, behaviour could not be generalised and sociology would remain very much to poor relation to natural science. Therefore, one tradition within the subject is positivist, macro and structural.
These factors dictate that the research methods used must be imbued with objectivity in order to collect data on social facts. The survey method can be implemented on a large scale and it has a kudos that makes it attractive to national governments to help them with social planning and policy reform. Questions are predetermined and possible answers are limited and fixed. Research can be replicated in order to ensure that it is reliable and surveys can be implemented at intervals to provide data on longitudinal patterns. This standardised method can also be used for comparative approaches like Durkheim's study of suicide or it might be used as part of a unique case study.
Anti Positivists like Goulder, Becker, or Mead argue that universal laws of social behaviour don't exist. So the pursuit of them, no matter how "scientific" you are in going about it, is a folly. Natural science and social science deal with different subjects. One concerns itself with matter and one concerns itself with people. One always acts in the same way in the same set of circumstances and one doesn't. Micro theorists argue that social actors have agency, self determination. As people have a large degree of control over their own behaviour, social actors are unpredictable, fickle creatures who defy generalisation. Social behaviour defies neat typologies and universal laws only apply to the natural world. The focus of sociology should be to explain the social actors first order constructs; why we act in the ways we do and how we make sense of the social world. The best way to achieve this explanatory and essentially subjective aim is to apply qualitative techniques like observation or unstructured interviews. So the search for meaning dictates the methodology.
What is the nature of your enquiry?
If you are conducting research on a social problem like crime, some form of positivist methodology would set you on the right path. It gives you an overview and paints a broad picture of your topic. It might give you data on national, regional or local trends, offender profiles and it should tell you if the problem is getting better or worse. If your enquiry is descriptive in nature, then quantitative approaches make sense. However, if you want to go deeper and delve into the reasons for offender behaviour, then a qualitative approach like interviews or covert participant observation might be used. Whether that technique is ethical in the circumstances is another matter. That said, many contemporary researchers triangulate and use varying combinations of positivist and anti positivist methods. This helps to inject greater rigor into their work and deters critics from trivialising the research because of its methodological exclusivity.
What is state of your finances?
If your coffers are full, you may be able to afford the indulgence of taking several months or years to complete your research. As a general rule, anti positive research tends to be more expensive than positivist methods because it is far more time consuming. Merely identifying your research subjects can take time, access to group has to be negotiated and once the primary research has been carried out, the data may take months to transcribe or collate before you have even had a chance to write a meaningful sentence worthy of appearing in a learned journal. If your coffers are empty, it may be prudent to opt for a method that generates statistics and make best use of SPSS, Statistical Package for the Social Sciences. Also, statistical methods often convey a certain seriousness and competence about an individual. So less senior researchers who don't enjoy the financial benefits of a professorial Chair might do well to invest in the requisite computer software and training.
What are the ethical considerations?
The British Sociological Association is actually quite non committal regarding what one should and shouldn't do as a researcher. Although they have ethical guidelines, they are quite broadly written and essentially say that one shouldn't do harm and that researchers shouldn't conduct research they are neither qualified nor competent to do. Certain subjects like child abuse, domestic violence and other forms of crime or deviance are prone to a number of ethical problems or dilemmas. Merely speaking to a researcher about some things may do harm in certain cases and it might not be ethical to pry into someone's painful or embarrassing experiences just to satisfy ones professional curiosity or career aspirations. Laud Humphreys Tea Room Trade, which investigated homosexual relations in public toilets was not only criticised for its subject matter but also for its methodology. Humphreys was complicit in the criminality because he acted a lookout and he was also criticised for deceiving his subjects and invading their privacy (what privacy one might expect in a public toilet is another issue). In short, positivist methods avoid many of the ethical problems inherent in studying social behaviour but it lacks depth and humanism. Researchers sometimes excuse the inexcusable by saying that "the ends justify the means."