Study sets, textbooks, questions
Upgrade to remove ads
sociology - research methods
Terms in this set (83)
information collected by sociologists themselves for their own purpose
information collected by someone else for their own purpose, but sociologist can then use.
refers to information in a non-numerical form
refers to information in numerical form
primary data evaluation
strengths - can control all aspects of experiment. representative as there is control over who collects data so can be applied to population
weaknesses - time consuming, needs to design and construct info. expensive
secondary data evaluation
strengths - less time consuming, data already been collected. less expensive, not collecting data themselves.
weaknesses - lack of control as data already exists so may have something that sociologist doesn't need so invalid. unrepresentative as info only comes from certain people so is biased.
qualitative data evaluation
strengths - valid data, in-depth, fully explained. less desirable answers given as researcher isn't imposing their meaning.
weaknesses - hard to analyse as isn't numerical, could be subjective. unreliable as can never be repeated (conversations, interviews)
quantitative data evaluation
strengths - easy to analyse, less detail so patterns / comparisons can be easily shown. reliable as consistency in results.
weaknesses - invalid, no detail or depth. meaningless as no understanding
what factors influence choice of methods?
practical issues: ease of access
how easy it is to get to ppts
practical issues: time and funding bodies
length and magnitude of study. money needed to research from funding bodies.
practical issues: personal skills and characteristics
individual characteristics of sociologist, ppts may make it harder or easier to research
practical issues: subject matter
more likely to research useful subjects or lead to promotional/financial gain
practical issues: research opportunities arising
some areas may not be popular as there could be risks
ethical issues: vulnerable groups
children, people with disabilities etc. need extra care in research. researcher needs to ensure consent is given, need to understand issues around group's needs
ethical issues: confidentiality and privacy
researcher cannot use personal info to prevent any negative effects from research
ethical issues: informed consent
ppts have to agree to take part. researcher needs to describe full nature of study before ppt agrees
ethical issues: effects on research participants
researchers should know possible consequences of their research on ppt. e.g. social exclusion,damage to employment prospects.
theoretical issues: reliability
replication of study. more factual
theoretical issues: validity
authenticity and truthfulness of study. more non-factual.
methodical perspective: positivism
belief that society can be studied scientifically.
focus on hard data (quantitative)
positivism: social facts
external forces which shape people's behaviour
positivism: macro approach
study large number of people
methodological perspective: interpretivism
belief that behaviour is influenced by individual's beliefs.
focus on soft data (qualitative)
understanding people's experiences
interpretivism: micro approach
focus on small number of people
names of all people in study population
decide how you are going to sample your ppts
representative group of people to study om
every member of population has equal chance of being selected
random sampling evaluation
+: truly random sample has most chance of being representative
-: time-consuming, being completely random is extremely difficult
ppts selected from sampling frame, regular intervals (nth term)
systematic sampling evaluation
+: generally representative
-: not truly random, every nth person
researcher breaks ppt into strata then randomly selected
stratified sampling evaluation
+: improves on random sampling, ensures it is representative
-: samples influenced by researcher (bias)
researchers look for and select right number of each sort of person required
quota sampling evaluation
+: ppts are closely linked with researcher's intention so is more valid
-: biased, not representative
snowball effect sampling
ask volunteers to choose other volunteers
snowball effect evaluation
+: self-selecting isn't time-consuming
-: lacks representivity as only ppl who want to take part join
types of document
personal: private to person
public: purpose is for public knowledge
historical: created in past. could be public or personal
photographs, letters, diaries, parish records, company accounts
acronym used to evaluate methods
acronym evaluate methods meaning
Practical - ease of access, funding bodies
Ethical - consent, harm etc
Reliability - repeated
Validity - truthfulness/authenticity
Examples - e.g. documents in education
Representative - reflection of whole population
Theory - positivism, interpretivism
Types of data - primary, secondary, qualititative, quantitative
documents evaluation: disadvantages
practical: some are expensive
ethical: can't get consent if historical
reliability: some can't be repeated
validity: problems with authenticity of historical
representative: doesn't apply to population if personal
theory: interpretation of meaning is different
type of data: secondary (loss of control), qualitative (hard to analyse)
documents evaluation: advantages
practical: easy to access
ethical: vulnerable groups
reliability: company accounts can be repeated
validity: fully understood/explained
representative: can be applied to some groups
theory: interpretivists favour documents
type of data: secondary (cheap), qualitative (valid)
content analysis process (plus example)
turning secondary, qualitative data (like documents) into primary, quantitative data. divides data into categories and count number of times categories appear.
example: Glasgow Media Group (media favoured managers portrayed as peaceful, and prejudice against workers, portrayed as noisy)
content analysis evaluation: advantages
practical: cheap, easy to access
ethical: don't need consent
validity: primary data is
representative: large documents are
types of data: primary (more valid) quantitative (easy to analyse)
content analysis evaluation: disadvantages
practical: bias in personal skills
ethical: psychological harm, job prospects
reliability: categories chosen cannot be repeated
validity: quantitative isn't
representative: small documents are not
theory: no soft data
types of data: primary (time-consuming) quantitative (meaningless)
define official statistics
quantitative secondary form of data. collected by government.
types of official statistics
Census - every 10 years, completed by law, housing composition, occupation, family, transport, leisure
suicide statistics - incidence, coroners, doctors
health statistics - doctors, medical institutions. inaccurate (not all sick people go to doctors, inaccurate diagnosis, children go to doctors more often)
crime statistics - reported and recorded by police. inaccurate (dark figure, bias from the state)
official statistics evaluation: advantages
practical: easy to access, cheap
ethical: confidential, don't need consent, little harm
reliability: can be repeated many times
validity: allows comparisons
examples: health statistics
representative: large scale can be generalisable
theory: positivists, show social facts
types of data: quantitative (easy to analyse), secondary (not time-consuming)
official statistics evaluation: disadvantages
practical: personal skills of sociologist
ethical: bad for vulnerable groups
reliability: different researchers may interpret info differently.
validity: not true picture, dark figure, bias
examples: crime statistics
representative: small scale are ungeneralisable
theory: positivists - meaningless
types of data: quantitative (invalid) secondary (bias)
used to test hypotheses to establish cause and effect. has control and experimental group. if change occurs during task, it must be due to experimental group as two groups were identical before experiment.
example of experiment
Harvey and Slatin - showed pictures of children of different classes, asked teachers to rate child's performance. Lower class - less favourable ratings
A change in a subject's behavior caused simply by the awareness of being studied
Hawthorne Effect example
Mayo 1972- told workers they were being studied. Made environment unpleasant, found they increased productivity no matter what. Responded to simply being watched.
P - hard to isolate a single cause (not practical), but might be only way to test subject matter e.g. crime
E- control group is treated differently, but ensures confidentiality
R- has internal reliability but not external due to real-world specificity
V- Hawthorne effect is reduced so better than lab, but is still there and only removed due to deception which is unethical
E- Mayo 1972, Harvey and Slatin
R- small scale cannot be generalised, large scale are too time-consuming and expensive
T- positivists prefer to establish social facts, lacks validity so interpretivists dont prefer
field experiments definition
similar to labs (have 2 groups), but in real-world situations, more concerned with meaning
field experiments examples
Rosenthal - sane in insane places. said they were hearing voices (indicate schizophrenia) and all were admitted to hospital. showed that doctors misdiagnosed patients.
Rosenthal and Jacobsen - paradigms in classrooms. told teachers 20% of pupils were going to spurt ahead. test a year later to find they achieved more. unethical to control group.
field experiments evaluation
P - real world settings lack control so harder to do, minimise research opportunities as researcher doesn't get involved
E - no informed consent (vulnerable groups), but ensures confidentiality as researcher has control
R - internal reliability, but no external reliability
V - Hawthorne effect still occurs unless deceived, but better than lab experiments
E - Rosenthal, Rosenthal and Jacobsen
R - small sample not generalisable but too expensive to do large sample
T - interpretivists prefer as focus on meaning and experiences, but some argue they cant establish social facts due to Hawthorne effects
comparative method of experiments
collects secondary data
compares differences in social groups
can be done historically to compare past and present data
establishes causes of social behaviour
comparative method example
Durkheim study into suicide
compared suicide rates in Catholics vs Protestants using Official statistics.
found Catholics committed suicide less as they were more integrated in society.
(-) suicide stats by coroners, bias
(-) Catholics view suicide as one of main sins so may have committed suicide but made it look like natural death, invalid
comparative method evaluation
P - easy/cheap as data already collected, but might need permission depending on subject matter
E - ethical issues when comparing groups but confidential
R - can be compared again, but actual stats could be bias from state/government
V - no Hawthorne Effect, but meaningless data and no Verstehen so not authentic
E - Durkheim
R - large scale data is generalisable
T - positivists prefer as establishes social fact which leads to social implications, but no detail or explanation.
social survey of pre set questions.
can be postal, face-to-face, internet and telephone
CSEW - 38,000 issued every year. asks if they have been a victim of crime in the past year.
Connor and Dewson 2001 - 4,000 to students asking about their influences for the W.C to go to uni.
P - might be only way to test sensitive subject matter but might have to offer incentives to boost response rate
E - confidential, postal difficult to complete for vulnerable groups
R - reliable for due to standardised procedure, but results only reflect limited time so have to be repeated
V - allows objectivity and detachment so less bias, but misunderstanding of questions
E - CSEW, Connor and Dewson
R - large scale are, but limited samples due to subject matter
T - positivists as they establish social facts and closed questions, no Verstehen
structured: standardised, pre-coded q's.
semi-structured: mix of two
group: can be any of above
Bicknell (2014) - parents of educationally gifted children
Oakley (1981) - women becoming mothers
P - easy to access ppts, characteristics of researcher limits rapport
E - stop interview at anytime if face to face and no peer pressure if not group, limited confidentiality
R - structured can be repeated, unstructured and group can't be repeated
V - unstructured + group are valid (spark discussion), social desirability (depending on subject matter) and interviewer bias and individuals left out in group discussions
E - Bicknell, Oakley
R - groups generalisable, individuals and structured not
T - positivists prefer structured (social facts, standardised, quantitative but setting becomes artificial). interpretivists prefer unstructured (validity, discuss more important things to them, but unique situations)
interviews problems and improvement
interviewer bias, validity
Mazroo (1997) - matched social characteristics of interviewer and ppts. more authentic answers, develops rapport.
Kinsey (1953) - quick fire q's to limit social desirability and lie.
Becker (1971) - played 'dumb' to give ppts the upperhand to expand on answers
get true/valid picture
covert vs overt
ppt (unstructured) vs non-ppt (structured)
Mirza and Reay (2000) - overt non-ppt. Observed classes ran by afro-Caribbean community.
Lyng (1990) - overt ppt. Observed high risk groups. Didn't hide the fact he was researching but took part in sky-diving etc.
observations factors and issues
getting in - may need gatekeeper in relation to group. Links to practical skills of researcher.
staying in - developing rapport but not going native (loses objectivity)
getting out - can't damage the group
observations overt evaluation
+ consent, no deceit
+ openly take notes (valid)
+ other methods can also be used
- hawthorne effect
- researcher not trusted
- denied access by group
observations covert evaluation
+ no hawthorne effect
+ access to 'forbidden' areas
- if discovered, can be dangerous
-no consent, deceive ppts
- may have to view or take part in illegal activities
observations ppt evaluation
+ only method to study group
+ develop rapport
- time consuming/expensive
- bias, lack of objectivity (going native)
- not reliable, qualitative data
observations non-ppt evaluation
+ can't influence group behaviour
- limited understanding
- Hawthorne effect
positivists don't like them overall (unreliable, unrepresentative) but would prefer non-ppt.
interpretivists prefer ppt due to micro approach and Verstehen.
other methods only show snapshot of behaviour. Longitudinal - same sample used over a period of time. Usually a social survey.
cohort - sample
sweep - data collected at regular intervals
case study - individual
life history - uses documents and interviews
Longitudinal study example
Willis (1997) - case study, 12M students, unstructured interviews and observations. Anti-school subculture. 18 months in school and then again 6 months into their working life.
Longitudinal study evaluation
P - people drop out
E - build rapport and ensures consent, but more pressure if they drop out
R - behaviour over longer time is good but difficult to repeat
V - valid, Hawthorne Effect may still occur
E - Willis (1997)
R - sample size that drops out reduces representativeness, but group is generalisable
T - interpretivists prefer but expensive to repeat and have to use same method in case study, qualitative data.
methodological pluralism and triangulation
methodological pluralism - using range of methods for single piece of research
triangulation - using 3 methods to check accuracy, reduce limitations of data and increase validity. e.g. content analysis, experiments, unstructured interviews.
Sets with similar terms
Research methods Sociology AS level AQA
METHODS: key terms
IB Psychology Paper 3 Notes
Sets found in the same folder
sociological methods key words