3. What you need to know:
• Quantitative and qualitative research methods
their strengths and weaknesses
• Sources of data
• Evaluation of research – validity, reliability,
representativeness, generalisability
• Primary and secondary data
• Sampling
• Positivism, interpretivism and their effect on
choice of method
• Ethical considerations
4. Research not common sense
• Research is based on evidence
• Research moves beyond common sense and
experience to the uncommon and factual
account
• Research can be objective
• It can be tested by other people
• Research can compare results to test accuracy
• Quality research enables valid generalisations to
be made about sociological phenomena
5. Durkheim’s study of suicide
• Probably the first sociology research study
• He was interested in the effects of anomie
• His hypothesis was that in anomic situations e.g.
times of war or political crisis, people were more
likely to commit suicide
• He collected lots of statistics from across several
countries and compared the suicide levels
• Many faults in the study – but was the first
systematic attempt to do sociology research
• As sociologists we have a duty to be critical of
any research method and the “facts” that it
generates
6. Correlational studies
• Unlike natural sciences we cannot put people in
a lab and experiment with them
• Sociologists work on associations or correlations
• Association cannot be cause and effect but a
link
• E.g. poverty and early death
• But also increased sales of ice-cream and
drowning (the link is that they are both more
likely to happen on hot days)
7. Theory
• A theory is a collection of linked ideas
which explain some observable facts –
think of Darwin’s theory of evolution
• Sometimes the theory comes first and
sociologist attempts to test it (e.g.
Durkheim and suicide)
• Sometimes the theory comes as a result
of exploratory research (grounded theory)
9. Types of Data
• Primary – those collected by the researcher
themselves – experiments, surveys,
observations, interviews. All of these methods
together are known as Empirical research.
• Secondary – those which have been collected
by other people or organsiations e.g. using
health statistics in the Black Report or even TV
and historical documents
10. Types of data
• Quantitative – That type of data which produces
numeric information in the form of charts,
statistics, tables. And answers questions like :
“how many times”. It is generally “objective”
• Qualitative – That type of data which produces
descriptive information conveying words,
feelings. It is rich, complex and difficult to
analyse e.g. Becker’s work on The Outsiders. It
is generally “subjective”
11. Evaluating Data
• If a research method or approach is weak it will
produce innacurate results
• Remember : GIGO (garbage in garbage out)
• There are five measures of quality:
– Reliability -can it be repeated and get the same result?
– Validity -Does it reflect a true picture of events?
– Representativeness -Does the sample of pps reflect the
target group?
– Generalisability – can the results be generalised from this
research to the whole of the population ? (needs large pps)
– Objectivity – Has the researcher tried to keep their own
values and beliefs out of the research process ?
12. Choice of subject
• Are sociologists more interested in some
people more than others
• Are funds limited to certain areas
(perceived problems)
• Personal bias – to prove a point
• The values beliefs of sociologists
• Access to some groups difficult because
of power – so why should we focus on the
powerless ?
13. Ethical Issues
• Choice of topic – we may draw negative attention to a
disadvantaged group “victim blaming” e.g. Afro
Caribbean youth
• Choice of group – some groups don’t have the power to
say no
• Effects on people to be studied – may find out
negative things about themsleves (the doctors in the
Rosenhan study) or in the case of Humphrey’s Tearoom
Trade, have negative consequences
• Effects on wider society – public policy, victims of
crime etc..
• Legality and immorality – committing illegal acts during
a study e.g. Venkatesh was involved in a beating
• Consent - asking people’s permission to study them and
their lives
• Anonymity – The right to remain anonymus and not to
be able to be identified from the report or study
15. Theory influences choice of
research topic
• Feminists interested in uncovering relationships
between patriarchy and women’s lives
• Marxists interested in how a relatively small
group of people oppress the masses
• Functionalist explore values consensus and
are interested in how society maintains
agreement
• Interactionists are interested in social action
and context
16. Relationship between theory and
method
• Positivism – is the approach based on the
natural sciences. Central is the idea of objectivity
and science. This approach was favoured by
earlier sociologists and tended to be “top down”
• Interpretivism – is the approach which tries to
see the world from the subjects perspective.
Tends to be bottom up and qualitative. This
research approach came originally from
anthropology.
17. Triangulation
• There is no perfect research method
• Qualitative and quantitative methods both have
weaknesses which tend to be the inverse of
each other.
• The use of mixed methods which is known as
triangulation can overcome some of the
weaknesses of a method
• E.g. A quantitative study on health an d lifestyle
could be supplemented by the use of in-depth
interviews or focus groups –thus increasing
validity of the research
18. Quantitative Methods
• Surveys- standardised questionnaire to find out reliable
facts, or test an hypothesis (cross sectional or
longitudinal e.g. The BCS)
• Experiments- to artificially create a situation which tests
a theory or hypothesis e.g. Rosenhan’s study into mental
hospitals
• Comparative research- collecting and anyalysing
statistics e.g. The Black Report or Durkheim’s study on
suicide
• These methods all tend to be high reliability but low
validity
19. Sampling
• Representative sampling – where
attempts are made to accurately sample
the target population to ensure
representative and generalizable results
• Non-representative sampling – where a
valid study of the experiences of a group
is more important than accuracy e.g.
studying prostitution
20. Random sampling
• Where every possible participant has an equal
chance of being chosen
• Advantages: least biased, most accurate
• Disadvantages: need to know the extent of the
target population and be able to get hold of them
e.g. via postcode list or electoral register
• Three main types:
– Systematic sampling every nth on list
– Stratified – divided on known criteria but you have to
know the % of characteristic in population
– Cluster e.g. geographic
21. Quota sampling
• Often used by market researchers
• Sampling based on proportions in UK population
– e.g. age, sex, income, ethnicity
• Each pp would be chosen on the basis that they
fulfil these e.g. if 1% of population is Asian and
female 1% of those sampled will be like that.
• Advantages: economic, requires smaller
sample and can choose to sample in one
location but make it generalisable
• Disadvantages you need to know the
demographics of the population and
innacuracies are hard to spot
22. Non-representative sampling
• Snowball – where each participant is
asked to refer researcher on to someone
of interest the the researcher. Highly
biased but useful for investigating
marginalised groups
• Theoretical – where a researcher selects
an non typical group so that they can
evaluate whether a charaqcteristic is
biological or socially constructed e.g.
mental illnesses
23. Pilot Study
• Essential
• Tests
– sampling frame
– Response rate
– Research instrument e.g. questionnaire
• Allows amendment for full research study
24. Qualitative Methods
• Some sociologists interested in
understanding the qualities of social life
• The favoured methods are:
• Observation
• Informal interviews
• Focus groups
25. Observation
• Not a single method but 4 choices:
• Non participant or participant
• And if participant then
• Covert or overt
26. Participant observation
+ -
• Experiences • Bias
• Getting to the • Researcher influence
truth • Ethics
• Depth • Too close
• Dynamic • Going native
• Getting to hard to • Studying the
reach groups powerless
• Keeping accurate field
notes
27. Non participant
• Less likely to go native
• Less likely to accidentally influence group
• But:
• Desirability bias – or Hawthorne effect
• Observing without being part may be
sterile
28. The overt / covert dilemma
Overt Covert
• Trusted outsider • Access to forbidden
• Honest and ethical areas
• Can use supplemental • See normal behaviour
research methods such • Dangerous
as interviews • Ethical dilemmas
• Can ask questions • Recording
• Can keep
contemporaneous notes
• But people may lead
you on
29. Ethnographic Research
• Ethnography means writings about a
group or culture.
• A researcher joins a group and uses a
range of methods to record the lives of the
people in the group.
• There are issues about how to gain
access, acceptance, recording, maintining
objectivity and if the group changes as a
result of researcher’s presence
30. Informal interviews
• More like a conversation with prompts
• Open questions which will allow subject to
explore or talk rather than give short
answer
• Researcher must be skilled
• Advantages – high validity, lots of info
• Disadvantages- hard to analyse, time
consuming and dependent upon skill of
researcher
31. Focus Groups
• Small group
• Series of prompts
• Facilitated by researcher
• Used a lot in market research to test ideas
• In sociology useful to get people to speak when
supported.
• Weaknessss- can over represent the views of
one or two dominant individuals
• And lost of rich data hard to analyse
33. Questionnaires
• Surveys
• Standardised
• Large numbers of people
• Handed out, internet or posted less often, face
to face e.g. consumer surveys
• Quantitative categories or Likert scales for
opinions
• Can contain a mixture of closed and open
questions
• A good questionnaire is easy to understand and
complete, gets the right info and preferably
short.
34. Inteviews
• Good for complex or sensitive subjects
• Researcher can clarify
• Higher response rate
• But:
• Bias – people want to present themselves
positively
• Lying
• Recording of information
35. Operationalizing concepts
• Turning an abstract idea into something which can be
seen, counted, measured in some way.
• “Health” can be operationalized as:
– Weight / height ratio (because people who are overweight tend
to have more ill health)
– Blood pressure (vulnerability to heart attacks / strokes)
– Healthy diet (what you eat, smoke and drink affects your health)
– Feelings of happiness (measured on a likert scale)
• These concrete measures such as blood pressure are
known as indicators, but their weakness is they are not
the “thing” itself so car needs to be taken when using
them instead
36. Language
• Leading questions must be avoided e.g.
• “wouldn’t you agree that…”
It is hard not to agree
• Emotive language or labelling language avoided
• E.g. describing someone as disabled rather than
talking about the disability – the former implies
that every aspect of their life is “disabled”
38. Secondary data
• Not collected by the sociologist themselves
• Collected by someone else for their own
purpose e.g. Government statistics such as the
census.
• Or produced for personal use e.g. photos and
diaries
• Useful to sociologists, particularly when
formulating research – reference to previous
research in the field
39. Why use secondary sources:
• Information is already available e.g.
SMR’s (standard mortality ratios)
• Historical events where participants can’t
be interviewed
• Cost to high to visit places
• Issues of safety for the researcher e.g.
researching role of women in Afganistan
• Groups who don’t want to be researched
e.g. the super rich..
40. Types of secondary data
• Previous sociological research
• Official publication
• Diaries, letters and personal documents
• Treaties, company records
• Novels
• Oral histories
• Documentaries, newspapers, internet
• pictures
41. Advantages of using secondary
data
• Statistics can illustrate trends e.g. health
improvement, trends in crime
• Personal documents may give a more
valid picture of someone’s life
• Oral histories – rich picture
• Media and content analysis – turn opinion
into “fact” e.g. yes the Daily Mirror does
seek to create moral panics about black
youth. GUMG analysis of topics covered
42. Issues with secondary data
• Statistics/ previous research: may not be exactly
what you need, also bias in collecting e.g.
suicide statistics
• Previous research studies – you can’t check
validity e.g. Margaret Mead and growing un in
Somoa.
• Personal documents – self serving bias or edited
by someone e.g. Otto Frank
• Oral histories – memories may be distorted
• Media and content analyis – different tools can
give different results, huge amount of data.