research methods Flashcards
questionnaires
good:
- quick
- no ethical issues
- high reliability can be repeated as questions are standardised
- high representativeness if a large sample is used (can generalise findings)
- high validity as they can be done anonymously (more honest answers)
bad:
- low response rate
- low validity (imposition problem, if closed questions are used people cannot give detailed answers)
- low validity (people can lie)
EXAMPLE: 2001 UK Census 390,000 people declared their religion as Jedi
questionnaires methods in context
good:
- higher response rates as headteachers can ensure teachers and students cooperate
- higher validity (anonymous so topics like bullying can be studied
bad:
- trouble operationalising concepts (putting questions into language students understand)
- choice of topic (studying anti school subcultures would be hard as students may not want to cooperate)
- low response rate (students can lose sheets in bag)
structured interviews
good:
- high representativeness (the illiterate can be studied)
- high reliability as standardised questions are used so its easy to repeat
- objective (standardised questions mean less interviewer bias)
bad:
- time consuming as they have to be done face to face
- low validity if people lie due to social desirability
structured interviews methods in context
good
- more representative as students with poor literacy can be studied
bad:
- low validity as students may lie
group interviews
good:
- less time consuming as several people are interviewed at once
- high validity as people may feel more relaxed around friends
- more representative if more diverse people are used
bad:
- low reliability as difficult to repeat the same
- low validity as people may lie to fit in with the group (conformity)
- low representativeness if a few people dominate the conversation
group interviews methods in context
good
- high validity as students may feel more comfortable surrounded by friends
bad
- low validity as students may copy eachothers answers (peer pressure)
- difficult as finding a time/space to interview a group of teachers may be difficult
unstructured interviews
good:
- flexible
- high validity as people can give detailed responses
bad:
- transcribing interviews is costly and time consuming
- low reliability as not standardised
- low validity as people can lie
unstructured interviews methods in context
good:
- can build a rapport so can talk about sensitive subjects (bullying)
bad:
- interviews with children cant be too long due to a short attention span
- low representativeness as small sample sizes will be hard to generalise the results across the education system
lab experiments
good:
- high reliability (easy to repeat)
- controlled environment as high level of control over the variables
bad:
- low validity as it isn’t a real social setting
- low validity due to Hawthorne effect
- unethical as harm may come to participants
lab experiments methods in context
good:
- high validity as researches can isolate variables (what causes labelling)
bad:
- ethical issues if students are used
- may be difficult to get informed consent from parents
field experiments
good:
- high validity as no hawthorne effect, real life setting
- high in reliability as can be easily replicated
bad:
- unethical as no informed consent
- low validity as cant control all variables so harder to establish cause and effect
field experiments methods in context
good:
- high validity as can be carried out in school (natural setting)
bad:
- ethical issues as no informed consent
- low reliability as differences between schools make it difficult replicate
structured observations
good:
- fixed behaviour categories mean high reliability and that its quick to analyse
bad:
- low validity as counting behaviours doesnt give the meanings behind them
structured observations methods in context
good:
- high reliability as can be repeated in different schools
- high representativeness as results can be generalised across schools
bad:
- low validity if overt as Hawthorne effect
participant observation
good:
- high validity as no Hawthorne effect
bad:
- time consuming
- unethical if covert (deception)#
- guilty knowledge, could get participants in trouble
participant observation methods in context
good:
- teachers are used to being observed (Ofsted) and so may be more willing
bad:
- difficult to do covert in a school, safeguarding, would have to have a job at school
- low representativeness as researcher only able to study one/two schools (time consuming)
non participant observation
good:
- high validity as objective, just watching what happens
- wont have to take part in illegal action
bad:
- low validity if overt as Hawthorne effect
- covert raises ethical issues of deception and no informed consent
non participant observation
methods in context
good:
- easy to do overt as schools will allow it (its like Ofsted)
bad:
- impossible to do covert in a school due to safeguarding issues
- students might not want someone following them around
official statistics
good:
- free source of huge amounts of data
- high representativeness as large samples used
bad:
- low validity for soft statistics (crime statistics may be wrong as not all crime is reported)
- subjective as the government defines concepts in certain ways to make themselves look better
official statistics methods in context
good:
- high reliability for statistics related to education
- high representativeness for education statistics (league table results)
bad:
- interpretivists say official statistics lack meanings (they don’t explain results)
