Quantitative Research Methods Flashcards
Positivism and quantitative methods
- argue that society is an objective social reality that shapes our behaviour
- use research methods that produce quantitative data (information in numerical and statistical form) e.g. experiments, questionnaires, structured interviews, official stats
- structured nature of these methods mean that the data produced is reliable and representative
Experiments
laboratory experiments, field experiments and comparative method or ‘thought experiment’
Laboratory experiments
artificial environment where the researcher controls variables to discover their effect, aim of discovering a casual law
two groups:
experimental group and control group
The experimental group
are exposed to a variable (independent variable) to test its effect
The control group
not exposed to the independent variable
Lab experiments have man PET limitations
Practical
Ethical
Theoretical
Practical issues - lab experiments
closed systems - lab experiments are only suitable for studying closed systems where all relevant variables can be controlled
Individuals are unique - can’t ‘match’ members of the control and experimental groups exactly
Studying the past - can’t control variables that were acting in the past
Small samples - can only use small samples
Hawthorne effect - subjects’ behaviour may change because they know they are being studied
Ethical issues - lab experiments
Informed consent - gaining subjects’ agreement to take part, explained the nature, purpose and risk of experiment to them
Harm to subjects - research shouldn’t harm subjects w/o compelling justification
Treating subjects fairly - experimental group are seen to be benefiting e.g. pupils gaining from a trial teaching method, unfair to the control group
Theoretical issues - lab experiments
positivists favour lab experiments for their reliability
However, even positivists recognise problems. Interpretivists criticise lab experiments as lacking reliability
Reliability and hypnosis testing - lap exp
reliability is imp. because replication enables us to check researcher’s results. Positivists regard the lab experiment as highly reliable cus:
> original researcher can specify steps in the experiment, others can re-run it
> produces quantitative data, results can be compared
> detached and objective - researcher just manipulates variables and records results
Representativeness - lab experiments
imp. to positivists because they aim to make generalisations about behaviour, findings may lack representativeness or external validity
> small samples may mean findings can’t be generalised
> lack of external validity: the more control, the more unlike the real world the experiment becomes
internal validity may be lacking because of the artificiality of the lab environment
Interpretivism and freewill - lab exp
clim humans are diff from natural phenomena, we have free will and choice
> actions are based on meanings, not ‘caused’ by external forces, means they can’t be explained through the cause and effect relationships experiments seek
Field experiments
may be used to overcome the lack of validity of lab experiments
> they differ from lab experiments in two ways, take place in the subjects’ natural surroundings, and the subject don’t know they are in an experiment
> manipulates variables to see what effect they have e.g. Rosenthal and Jacobson manipulated teachers’ expectations by misleading them about pupils’ abilities
Actor tests and correspondence tests
are field experiments, to test for discrimination in employment. Brown & Gay sent a white and black actor for interviews for the same posts. they were both matched for age, qualifications etc, so any diffs in job offers could have been due to discrimination
The comparative method
another alternative to lab experiments
carried out only in the mind of the sociologist - ‘thought experiment’, it identifies two groups that are alike in all major respects except for the one variable we are interested in
It them compares them to see if this one diff has any effect e.g. Durkheim’s study of suicide
Questionnaires
written/self completion questionnaires are the most widely used form of social survey e.g. Census
respondent must provide answers to pre set qs, usually closed ended w/ a limited range of pre set answers, but can be open ended, where respondents are free to answer in their own words
Practical issues - questionnaires
- cheap to gather large amounts of data
- no need to recruit interviewers (do themselves)
- data easily quantified and can be computer processed to reveal relationship between variables
- data is often limited and superficial, need to be fairly brief so ppl take them
- low response rates are a major problem, can be caused due to faulty questionnaire design e.g. complex lang
- inflexible and can’t explore any new areas of interest, only snapshots
Theoretical issues: positivism - questionnaires
positivists prefer questionnaires, seen as representative, reliable, objective and detached method for producing quantitative data, testing hypotheses and developing casual laws.
Hypothesis testing - questionnaires
questionnaires are attractive to positivists because they yield quantitative data about variables, this enables researchers to test hypotheses and to identify correlations and cause and effect relationships between variables
Reliability - questionnaires
(replicability) other researchers can easily replicate the original research by using the same questionnaire
Representativeness - questionnaires
> positivists favour questionnaires because they are large scale, distributed quickly and cheaply by post/email over wide geographical areas and usually use sampling techniques that give a representative sample
representativeness can be undermined by low response rates, especially if those who do return their questionnaires differ in some way from those who don’t
Sampling - questionnaires
sample is drawn from a sampling frame - list of members of the research pop.
> diff sampling techniques - random sampling, quasi random sampling, stratified random sampling, quota sampling, non representative sampling
Quasi random sampling
selecting every 10th name
Stratified random sampling
subdivides the population into the relevant categories and randomly selects a sample of each
Quota sampling
look for the right number (quota) of people required in each category
Non representative sampling
may be used where there is no sampling frame for the population
Detachment and objectivity
scientific research is objective (unbiased) and detached: values are kept separate from the research and not allowed to ‘contaminate’ findings. Positivists see questionnaires as a detached, scientific approach, since the sociologist’s personal involvement with respondents is detached, scientific approach, since the sociologist’s personal involvement with respondents is kept to a minimum, unlike in an interview, there is no researcher to influence the answers
Ethical issues - questionnaires
> parental consent for questionnaires with children
confidentiality is assured, since most questionnaires are completed anonymously
Interpretivism and questionnaires
seek to discover the meanings underlying our actions and from which we construct social reality, they tend to reject questionnaires because they can’t yield valid data about actor’s meanings
Detachment, Imposing the researcher’s meanings, Lying, forgetting and trying to impress
Detachment
reject the detachment and objectivity of questionnaires because they believe this fails to produce valid data, instead we need a subjective understanding of actor’s meanings
Imposing the researcher’s meanings
aim to reveal actor’s meanings, questionnaires prevent this by imposing the researcher’s framework of ideas on the respondent
Lying, forgetting and trying to impress
create validity problems and the researcher is often unaware how far these issues affect responses
Types of interviews
structured (SIs) - standardised questionnaires, same questions, word, tone
unstructured (UIs) - open ended, free to vary wording and tone
semi structured - same questions but interviewer can probe with additional qs
group - relatively unstructured, researchers asks in a group setting
Practical issues - SIs
> quick and cheap to administer to large numbers of people
suitable for gathering factual info
closed ended qs with pre coded answers are easily quantified
r response are usually higher than for questionnaires
inflexible, no new leads can be followed up
Ethical issues - SIs
> interviewees may feel under pressure, feminists regard them as potentially oppressive to women interviewees
make it clear they have a right not to answer
Theoretical issues: positivism - SIs
Hypothesis testing, reliability, representativeness
Hypothesis testing - SIs
> positivists model their approach on the natural sciences, seeking laws of cause and effect
structured interviews enable them to test hypotheses by revealing correlations and possible cause and effect relationships
Reliability - SIs
> interviewees are asked the same qs, so the answers can easily be compared to identify similarities and diff
Representativeness -SIs
relatively quick and cheap which enables larger numbers to be conducted
Interpretivist criticisms - SIs
criticise SIs for lack of validity
> give interviewers little freedom to explain qs or clarify misunderstandings
> may lie or exaggerate
> researcher has decided in advance what is imp
Feminist criticisms - SIs
relationship between interviewer/ee reflects the exploitative nature of patriarchal gender relationships
Oakley argues that this positivistic ‘masculine’ approach to research values objectivity and regards ‘science’ as more imp than the interests of the people it researches
Graham claims SIs distort women’s experiences because they impose the researcher’s categories on women
Interviews as social interactions
status diffs - between interviewer/ee may affect honesty and willingness to cooperate
cultural diffs - may lead to misunderstandings
social desirability - may give false answers to make themselves seem more interesting
interviewer bias - may ask ‘leading’ qs, influence answers by tone of voice or identify too close w/ interviewee
Official statistics
created by the gov or other official bodies
various non state organisations e.g. churches and charities also produce non official stats
Practical issues - offstats
A-> only the state has the resources to conduct expensive large scale surveys and sociologists can access these for free to use in their research
- only the state has the power to compel individuals to supply certain data e.g. parents are required by law to register births, this reduces the problem of non response
- official statistics are collected at regular intervals showing trends and patterns
D-> there may be no statistics available on the topic the sociologist is interested in
- the definitions the state uses may be different from the sociologist’s.
- the state may change the definitions it uses, making comparisons over time difficult
Theoretical issues: positivism - offstats
Positivists assume official statistics are reliable, objective social facts - a major source of representative, quantitative data that allows them to identify and measure behaviour patter
and trends.
* Statistics can be used to develop and test hypotheses, identify correlations and discold causal laws
Representativeness - offstats
> they are very large-scale, some official statistics are extremely representative
stats from off surveys may be less representative because they are often only based on a sample of the relevant pop. However these surveys are still much bigger than most sociologists could afford to carry out
great care is taken with sampling procedures to ensure representativeness when conducting official surveys