Studied by 2 people
positivism
focuses on producing quantitative data
based on the belief that sociology should use the same research methods as the natural sciences
sociologists: auguste comte, emile durkheim
objectivity and lack of bias leads to accuracy
favour experiments, social surveys and questionnaires
quantitative data
interpretivism
argue the difference between sociology and natural sciences
focus on understanding how humans see reality rather than ability
dismiss positivist methods
use unstructured interviews and participant observation
the main steps in devising and implementing a research strategy
research aims and topic selection
hypothesis setting and revision
pilot studies
sampling
the main steps in devising and implementing a research strategy - research aims and topic selection
sociologists identify a sociological problem
aim: understand them - not providing answers
best kinds of problems are ones without understanding
research projects are always related to or arise from a different research
factors influencing research topic:
personal interests, experiences and observation
what is already known about the topuc
social changes and developments
availability of funding
practicality
ethical issues
cost and time on research is important - research is carried out by sociologists employed by educational and research bodies
sociologists need to convince funding bodies that their research is worthwhile
the main steps in devising and implementing a research strategy - hypothesis setting and revision
review available evidence
find previous information
literature review
draw on other ideas to clarify issues
have a clear hypothesis
hypothesis: statement of research in attempt to find evidence to support or disprove
find how 2 or more variables connect
cause and effect relationship
establish correlation
positivist approach: has hypothesis
interpretivist approach: looser broader aim
decide research method and plan implementation
choose from different methods based on practical, ethical and theoretical issues
should provide material and evidence supporting or disproving hypothesis
the main steps in devising and implementing a research strategy - pilot studies
test with a small number of respondents to see problems in design or improve plan
putting problems right early can save money, time and effort
the main steps in devising and implementing a research strategy - sampling
choose appropriate sample by selecting possible respondents
expensive and impractical to include all respondents
samples make research manageable
people that research is about - survey population
list of everyone in the population - sampling frame
good sampling frames are hard to find
samples aren’t necessary sometimes
many countries have a census which collects information from the whole population
sample has to be a cross-section of the population
commonly used sampling frames
electoral roll - list of everyone registered to vote in elections
telephone directories
school registers
types of samples
random samples
stratified samples
systematic samples
cluster samples
opportunity samples
quota samples
snowball samples
random samples
everyone in the sampling frame has an equal chance of being chosen
not always representative
stratified samples
overcomes problem of random samples not always being representative
sampling frame is divided and then a random sample is taken from each division
divided by different characteristics
systematic samples
regular to the choice e.g. every tenth person
cluster samples
used when population is spread out over a large area
certain areas are chosen for sampling frame from which random samples are taken
reduces expense and time of travelling around whole country
opportunity samples
people who are available at the time to take part in the research
researcher makes decisions
often used when short on time
quota sampling
researcher has instructions to find people with certain characteristics
used in market research
snowball sample
finding one respondent and getting them to put you in touch with more or others
no sampling frame
what happens after a pilot study
research is carried out
data and information is collected
sociologist has to analyse and work out problem from data
data raises further questions
research findings are reported, read and used
findings published in books
end of individual research
research stages of difficulties in implementing a research strategy
identifying a topic
reviewing existing evidence
developing an aim
choosing a method
implementing research method
identifying a topic - potential problem
practical problems - finding respondents, funding
reviewing existing evidence - potential problem
difficult to find - many sources, outdated
developing an aim - potential problem
difficult choices about overall approach
choosing a method - potential problem
method chosen may not produce data, may not confirm or disprove hypothesis or doesn’t fulfill aim
implementing a research method - potential problem
practical, ethical, theoretical
ethical issues affecting the choice and implementation of a research strategy
harm
informed consent
privacy
deception
anonymity
confidentiality
harm
participants musn’t be harmed in a study - includes but isn’t limited to physical, emotional and mental harm
informed consent
partipants’ informed consent must be obtained - respondents must agree to take part, understood what’s involved, purpose of research, findings and can refuse to take part in answering particular questions
researcher shouldn’t try to persuade participant
not necessary to get informed consent from everyone involved
rule of consent rules out covert observation
amount of information which needs to be given out is also unclear
researchers break this guideline quite frequently in minor ways
e.g. underestimate how long a research will take
privacy
researcher shouldn’t invade participant privacy - researcher needs to respect privacy after consent and researcher cannot force participant to give information
deception
partipants shouldn’t be decieved - researchers shouldn’t present research different from what it is
lying about research topic
not giving full information in order to get the partipant to respond naturally
anonymity
partipant’s name or defining traits mustn’t appear on survey form unless for further contact
confidentiality
it shouldn’t be possible to trace an individual’s answers from published findings
main methods used in sociological investigation
questionnaires/social surveys
interviews
experiments
case studies
longitudinal studies
participant observation
non-participant observation
content analysis
triangulation
questionnaires/social surveys
highly reliable
not always valid
the UK census survey
questionnaires/social surveys - types of questions
closed or pre-coded
scaled
open
questionnaires/social surveys - closed or pre-coded questions
researcher provides set of answers
respondents can choose from answers
researcher limits responses
each answer is coded by a value used for analysis
questionnaires/social surveys - closed or pre-coded questions (advantage)
easy to analyse results and produce statistical tables
questionnaires/social surveys - closed or pre-coded questions (disadvantage)
some respondents may want to give answers not in the options
questionnaires/social surveys - scaled questions
form of close-ended question
debate: odd/even number of respondents
odd: respondents choose middle one to avoid making a choice
even: respondents are pushed to make decisions
this data proves things
risks forcing people
questionnaires/social surveys - open questions
provides quantitative data and limited qualitative data
uses for deep data so respondents can provide a response
difficult to analyse data
questionnaires/social surveys - ways of administering surveys
self completion questionnaires
structured interviews
ways of administering surveys - self completion questionnaires
respondents answer questions without additional guidance from researcher
researcher is absent: questionaires are returned with some questions empty, inappropriate answers, misunderstood
postal questionnaires - large sample, more representative, generalised, low response rate
argued that those who return questionnaires are different from those who don’t
self completion questionnaires - ways of improving response rate
sending a letter explaining research in detail and obtaining informed consent
sending a stamped, addressed envelope
following up those who don’t respond
making the questionnaire short and easy
giving relevant and clear, short questions
using people’s names for personal involvement
sign each letter
offering an incentive
ways of administering surveys - structured interviews
researcher reads out questions and answers allowed and records them
telephone, face-to-face questionnaires
face-to-face questionnaires: administered location, interviewee should be at east
questions should be standardised
higher response rate
researcher can explain purpose and reassure partipants
take up more time and are more expensive
questionnaires/social surveys - guidelines for good questionnaire design
short
clear layout
easy instructions
start with short questions
minimum needed questions
enough alternative answers
questions shouldn’t be leading
questions should avoid terms that aren’t understood by everyone
only ask things partipants know
best to leave personal information to the end
questionnaires/social surveys - strengths of self-completion questionnaires
cheap
can be sent to those who are geographically distant
researcher cannot influence answers given
convenient as participants can choose when to answer
questionnaires/social surveys - limitations of self-completion questionnaires
low response rate
questions may be understood
researcher cannot be sure who answered the questions
participants leave questions answered
questionnaires/social surveys - strengths of structured interviews
interviewer can explain questions
interviewer can ask additional questions or avoid irrelevant ones
higher response rate
good relationship with participants = trust and valid answers
questionnaires/social surveys - limitations of structured interviews
interviewers may influence answers through participant interaction
time-consuming
expensive
participants may give socially desirable answers
several interviewers may approach work in different ways
interviews
unstructured
semi-structured
focus group
group
interviews - features
qualitative
no schedule of questions
flexible and conversational
interviewer can follow up
hard labour
interviews - unstructured
brief set of prompts
aim: get interviewee to talk freely
interviews - semi-structured
interview guide - list of questions, topics
order of questions varies
all questions may not be asked
similar wording for questions
interviews - guidelines for good interviews
interviewers should be flexible and good listeners
make interviewee comfortable
create an order so questions flow well
understandable and relevant language
avoid leading questions
keep a fact sheet record
make sure interview location is quiet and private
use a good quality recording machine to record and transcribe interviews
interviews - types of good questions
introductory
follow-up
probing questions
specifying questions
indirecrt questions
silence
interpreting questions
focus group interviews
group interviews focusing on a particular topic
researcher has interview guide with different types of questions
researchers find out how groups respond to each other
closer to real social life
researcher can decide how much they want to be involved
risk: irrelevant discussion
researchers should know how to deal with silence, no speakers and people who speak too much
difficult to record and transcribe
interviews - strengths
interviews provide detailed and valid data about respondents
flexibility allows interviewer to probe deeply
interviewers can assess the honesty and validity of answers
information for further investigation can be brought out
interviews - limitations
time consuming to carry out, transcribe and analyse
difficult to make generalisations without standardised questions
less reliable - difficult to replicate
interviewer should be highly skilled
interviewer bias
interviewer effect - affected by CAGE
experiments
positivist method
produce quantitative data
used to find cause and effect/correlations
lab experiments
controlled variables, minimised subjectivity, quantified data
viewed as flawed
aren’t helpful as people live in society, not in labs
hawthorne effect
bandura and the bobo dolls
sociologists prefer field experiments
pygmalion in the classroom
experiments - field experiments
ethical problems
risk
researcher may lose control
effective for getting inside group behaviour
case studies
involved detailed research on one or more topics
involve any combination of methods
usually not possible to generalise or prove/disprove a hypothesis
affluent worker (case study)
case studies - strengths
allow different aspects to be studied using different methods
provide deep and detailed accounts
draws wider conclusions
produce findings that can be tested by other research
case studies - limitations
findings may apply only to the case
may not be generalised
cannot be replicated
researchers involvement can influence results
longitudinal studies
carried out over a period of time
used by government-funded organisations
employs surveys
panel studies
same sample used repeatedly
panel sample
panel members interviewed regularly
NCDS
7 up
longitudinal studies - strengths
gives a snapshot view of society and shows it over time
shows factors for changes
valid data due to committment
longitudinal studies - limitations
considerable amount of time and research
sample attrition: indivduals die, move away, decide they don’t want to partake, movement of individuals
hawthorne effect
participant observation
helps develop an understanding of the world from the point of view of the subjects
researchers put themselves in the same position as those they’re researching
stages of participant observation
getting in
staying in
getting out
gang leader for a day - case study
stages of participant observation - getting in
researchers adopt an overt rule and declare their true identity to the group or adopt a covert rule and produce a cover story
researcher should share the same CAGE as the group
the researcher should gain relationships with key individuals to get access
stages of participant observation - staying in
observer should develop a trusting role
involves learning and listening
taking notes may disrupt the natural behaviour
objectivity
researcher may have to do things they don’t agree with
stages of participant observation - getting out
researcher should be detached enough to write an impartial and accurate account
members of the group shouldn’t be identified in reports
participant observation - covert participation observation
group doesn’t know research is taking place
decieves the group and conceals truth
no informed consent
public research protects anonymity
most likely used where criminal or deviant activities are involved
avoids the risk of changing the behaviour of the group under study
researcher will have to become a full participant in the group as research may be ruined
may involve participating in illegal activities
difficult to ask questions and take notes without arousing suspicion
moral and ethical concerns about observing and reporting without consent
observer has to work hard at passing as a member of the group
participant observation - overt participation observation
group is aware that research is taking place
informed consent
researcher may be able to ask questions or interview people
researcher can avoid participating in illegal or immoral behaviour
ethically and morally right
validity - group being studied may behave differently from normal
participant observation - strengths
high in validity
deep understanding
participant observation - limitations
presence of observer may affect group behaviour
low reliability
cannot be generalised
problems in gaining access, winning acceptance, etc.
researcher should share similar social characteristics
covert: time and energy needed to maintain cover
researchers may lose objectivity and be influenced by the group
non-participant observation
carried out by observation alone without the researcher participating
eliminates the risk that people will be affected by the presence of a researcher or a new member of their social group
used when groups might be unwilling to cooperate in the research - raises ethical issues
quantitative data
avoids hawthorne effect
normal social situations
visible presence of observer may influence the activity
doesn’t let the researcher investigate the meanings people attach to the behaviour that is being observed - issues over reliability and validity
content analysis
studies content of documents and mass media
researcher defines set of categories and classifies the material being studied by how frequently it appears
viewing the world - case study
content analysis - strengths
provides information in statistical form
reliable
avoids ethical issues as it doesn’t involve respondents
content analysis - limitations
quantitative data
doesn’t answer why
difficult to decide categories and allocate media to it
triangulation
researcher uses a variety of methods
triangulation - strengths
supports quantitative data with qualitative data - reliable, valid
validity can be checked
reliability can be checked
cross-referencing
provides balance
triangulation - limitations
time consuming
expensive
researcher should be skilled
difficult to combine positivist and interpretivist approaches
analysing research - 3 types of issues when evaluating
practical - time, money, response
ethical
theoretical - overall approach, validity, reliability, representativeness, bias
validity
extent to which the research findings accurately reflect reality
do interpretivists favour validity or reliability
validity
reliability
extent to which the findings of the research can be confirmed by repeating the study
representativeness
findings of the research should apply not only to the actual sample but to the whole population being studied
sampling error
samples cannot be the same as the whole population so there will always be a difference between the results for a sample and results for the whole population
how is sampling error reduced
by having a large random or stratified random sample
research bias
positivist view: argue researchers should be neutral and object for reliability
interpretivist view: argue sociology is about people so it’s impossible to be completely unbiased
researchers should be open about their bias to readers and let them make their own decisions about validity and reliability - adopted by feminists
imposition problem: the researcher imposing themselves or their values on the research
happens through the social characteristics of the researcher
or happens through the ways in which researchers word questions or analyse data
