Choice of method & the research process

Reliability

Reliability

Emphasised by positivists; they like reliable, scientific methods

→ Replicable; repeatable by other researchers to obtain same result

→ Standardised measurement; data can be used to re-test hypotheses

Representativeness

Emphasised by positivists to discover wider social behaviour

→ ==Characteristics of sample ==should be same as those of the wider group, to make generalisations

→ Cant study whole population youre interested in, so study simply

Validity

Emphasised by interpretivists as structured techniques dont reveal what social actors really think

→ Authentic/true data; verstehen

→ Want to get close to reality of a situation

Primary vs secondary data

→ Primary is collected firsthand e.g. questionnaires/observation

→ Secondary is already collected by someone else e.g. official stats, documents

Positivism

Behaviour is result of social forces which shape our actions, we should aim to discover causes of our behaviour

→ Standardised, reliable, representative quantitative data

→ Use data to identity cause/effect relationships + patterns

Interpretivism

Social actors give subjective meanings to events rather than there being an objective social reality; we should try to uncover their meanings/worldview

→ Open-ended, valid, qualitative data

→ Use data to understand individual worldviews in their own words

→ Value verstehen

Methodological preference -- choice of method

POSITIVSTS

→ quantitative

→ structured methods

INTERPRETIVISTS

→ qualitative

→ unstructured methods

Practical factors -- choice of method

→ Time; some research takes more time

→ Finance; money avail, some methods cheaper e.g. postal questionnaires

→ Source of funding; research sponsored by govt/businesses reflects concerns of the funding body

→ Personal factors; may not be able to do lengthy research due to family commitments

→ Research subjects; some groups less open, so cant use structured methods

→ Research opportunity; one may suddenly appear, so no time to prepare questionnaire/interview schedule

→ Personal danger; direct contact w/ research groups may increase danger to researcher

Ethical factors -- choice of method

→ Informed consent; need to gain this due to potential effects of research; shouldnt mislead people

→ Confidentiality; right to anonymity

→ Effect on subjects; findings may be used to damage subjects by political groups/media

→ Vulnerable groups; must take special care with these

→ Covert research; ethical issues with detection when withholding researcher identity, vs difficulty of accessing groups who dont want to be investigated

What factors are most important? -- choice of method

→ Theoretical factors; pos. influence on what kind of data sociologist prefers to have

→ Practical/ethical; limits choice of method

→ Concerns often interrelated; e.g. qual. data has issues of access/trust, quant. has problems of sampling frames etc.

→ Triangulation; some socs see advantages to both types -- combining methods so strengths of one balance out the other

Choice of topics

→ Practical factors; some topics hard to study

• e.g. high-level political decision-making

→ Funding bodies; only fund topics they consider important

→ Societal values; interest in particular topics/issues e.g. BLM

→ Sociologist’s theoretical perspective; feminists study gender, functionalist benefits of TNC

→ Chance; may find themselves in a particular research situation by chance

Process of research

→ Researcher picks topic

→ Formulates aims + hypotheses

• Statement on what they intend to study, and a possible explanation to test

→ Operationalising concepts; defining sociological ideas in ways they can measure

• hard to compare findings if you operationalise concepts differently

→ Pilot study; if using survey methods, produce a draft + trial it

Sampling frames

→ Impractical to study whole research population; cost, time, etc.

→ Researcher creates a ==sampling frame ==(list of members of research pop) to choose sample from

• e.g. electoral register

Samples and representativeness

→ Can be rep. if have same characteristics in same proportion as wider population

→ If sample is a representative cross-section, then what is true for sample will be true for whole group

→ Important to positivists so they can make generalisations

Are all samples representative?

→ Smaller samples less likely to be representative of large pops

→ Interpretivists more interested in meanings held by actors so less emphasis on rep. samples

Types of sampling

→ Random sampling; everyone in frame has equal chance of selection

• eliminates bias, but not all samples large enough to ensure this represents characteristics of frame

→ Systematic/quasi-random; selecting every nth person in frame

• reduces chance of biased sample being randomly selected

→ Stratified; sample in same proportion as wider population

→ quota sample; stratified as above, then interviewer given a quota of characteristics to fill w/ respondents fitting these