Research Methods- y12 Flashcards
experimental method- key concepts
operationalised hypothesis (statement of what researcher believes to be true)
1.directional- states whether changes are greater/lesser pos/neg
1.non directional-doesn’t say direction, just says there is a correlation
-researcher causes IV to vary and measures effect on DV
experimental method- research issues
-EV’s= nuisance variables that make it hard to detect effect
-CV’s=change with IV so we cannot be sure if change in DV is due to CV or IV
-demand characteristics= any cue from researcher that may reveal aim and change behaviour
-investigator effects= any effect they have on DV
-randomisation= use of chance when designing investigation to control bias
-standardisation= same procedure for all pps
experimental method- pilot studies
-small scale trial run to road test procedures
-control groups/conditions set to make comparisons
-single blind means pps dont know aim double blind means investigator also doesn’t
-these things reduce demand characteristics
experimental method- independent groups
-one group does A and another does B
-randomly allocated groups
pros=no order effects, will not guess aim
cons=participant variables, less economical
experimental method- repeated measures
-same pps do all conditions
-order should be counterbalanced to reduce order effects
pros= participant variables, fewer participants
cons= order effects, pps may guess aims
experimental method- matched pairs
-2 groups of pps used who are paired on pps variables related to experiment
pros=participant variables, no order effects
cons= matching is not perfect, more pps
types of experiment- lab
-controlled environment where EV’s and CV’s can be regulated
-Iv manipulated, effect on DV recorded
pros= EV’s and DV’s controlled, easily replicated
cons=may lack generalisability, demand characteristics
types of experiment- field
-natural setting
-Iv manipulated, effect on DV recorded
pros=natural environment, pps unaware of study
cons=difficult to control CV’s and EV’s, ethical issues
types of experiment- natural
-Iv changes but not because of experimenter
-DV may be naturally occuring or devised, measured in field or lab
pros= practical and ethical, external validity
cons=event may only occur rarely, pps not randomly allocated
types of experiment- Quasi
-IV based on pre existing difference such as age or gender
-DV may be naturally occurring or devised
pros=high control, comparisons can be made between people
cons= pps not randomly allocated, casual relationships not demonstrated(cannot say change in DV was due to IV)
Sampling
-population= large group we are interested in
-sample= smaller part of pop
-generalisation= sample should be representative so we can generalise
-bias= certain groups are over or under represented
random sample
-everyone has equal chance of being selected
-lottery method/ randomiser
pros=potentially unbiased
cons= rime consuming and may not work if pps refuse
systematic sample
-pps selected using set pattern e.g. alphabetical order
-every nth person selected
pros=unbiased
cons=time and effort, might as well use random
stratified sample
-sample reflects proportions of people in certain sub groups
-sub groups identified and relative percentages chosen which represent population
pros= representative
cons=stratification not perfect as not all types of people can be represented
opportunity sample
-people who are easiest to obtain
-ask people nearby
pros=quick method
cons=biased and cannot be generalised
volunteer sample
-self selecting
-advertisements
pros= participants are willing and likely to engage
cons= volunteer bias, likely to share certain traits
ethical issues- conflict
-conflict between rights of pps and aims of research
-bps code of conduct, document to protect pps
-ethics committee weigh up costs and benefits before allowing a study
ethical issues- informed consent
-get permission
-signed consent for or;
1.presumtive-ask similar group
2.prior general-agree to be decieved
3.retrospective-get consent after
ethical issues- deception
-misleading
-cannot deliberately mislead or withold info, mild deception is ok
-debrief at end=
1.true aims
2.details not given
3.what data will be used for
4.right to withold date
ethical issues-protection from harm
-risk
-pps should be at no more risk then everyday life
-right to withdraw at every stage
-reassured behaviour is normal
-counselling provided if needed
ethical issues-privacy and confidentiality
right to control -we have right to withold data, confidentiality should be respected
-personal details must be protected or not recorded
-refer to pps as numbers
-personal details not to be shared with other researchers
correlations
-No manipulation of variables, so cause and effect cannot be demonstrated
-illustrates association between two co-variables
pros=useful starting point, relatively economical as second hand data can be used
cons=no cause and effect which can lead to false conclusions, intervening variables
observational techniques
-seeing or listening to people without asking, way of assesing DV
pros=capture what people do spontaneously
cons=risk of observer bias
observational techniques- naturalistic
-takes place where target behaviour would usually occur
pros=high external validity therefore more generalisable
cons=low control of EVs and CVs
observational techniques-controlled
-some control/manipulation of variables
pros=easily replicated due to standardised procedures
cons=low external validity
observational techniques- covert
-pps unaware they are being studied
pros=demand characteristics reduced which increases internal validity
cons=ethically questionnable due to right of privacy
observational techniques- overt
-pps aware of being studied
pros=ethically acceptable
cons=demand characteristics
observational techniques-participant observation
-researcher becomes part of group they are studying
pros=greater insight which increases external validity
cons=possible loss of objectivity
observational techniques-non participant
-researcher remains seperate from group
pros=more objective
cons=loss of insight
observational design-behavioural categories
-target behaviour to be observed should be broken up in to set of observable categories
cons=difficult to make clear and unambiguous categories should be self evident and not overlap, dumped behaviours go unrecognised
observational design- event sampling
-target behaviour recorded every time it occurs
pros=useful for infrequent behaviour
cons=complex behaviour is oversimplified, details go unrecorded affecting validity
observational design-time sampling
-observations made at regular intervals
pros=reduces number of observations
cons=may be unrepresentative and not reflect whole behaviour
self report techniques- questionnaires
-pre set list of written questions
-can be used to assess DV
pros=distributed easily, straightforward to analyse
cons=respondednts may portray themselves in positive light, response bias
self report techniques- structured interview
-pre determined questions in a fixed order
pros=easy to replicate, standardised
cons=interviewers cannot elaborate
self report techniques-unstructured interview
-no set questions, just a general topic to be discussed
pros=greater flexibility and insight
cons=increased risk of interviewer bias
self report techniques-semi structured interview
-list of questions but can elaborate questions based on previous answers
designing questionnaires
avoid; jargon, double barrelled questions and leading questions
-closed questions usually lead to quantitative data
pros=easy to analyse
cons=restricted response
-open questions usually lead to qualitative data
pros=not restricted
cons=hard to analyse
designing interviews
-standardised list of questions
-quiet room encourages them to open up
-rapport- begin with neutral questions
-ethics, remind they will be treated in confidence
qualitative and quantitative data
qualitative- rich detail but difficult to analyse
quantitative-easy to analyse, narrow in meaning
primary and secondary data
primary-fits job but requires time and effort
secondary-inexpensive, but may be poor quality such as outdated
meta analysis
-secondary data, combining many studies and calculation of effect size
pros=increases validity and generalisability
cons=publication bias
measures of central tendency
1.mean
pros=sensitive measure
cons=unrepresentative as small or large number can sway it
2.median
pros=less affected by extreme scores
cons=less sensitive than mean
3.mode- most frequent
pros=relevant for categorical data
cons=overly simplified
measures of dispersion
1.range-diff between high and low
pros=easy to calculate
cons=does not account for distribution of scores
2.standard deviation-measure of spread around the mean
pros=more precise than mean
cons=may be misleading as distorted by extreme values
The sign test
1.score from condition b subtracted from condition a to produse sign difference (plus or minus)
2.add up number of pluses and minuses
3.pps who got same in condition a and b should be disregarded and removed from N value
4.s value is total of less frequent signs
5.if s is equal to or less than significant value, s is significant and null hypothesis rejected
peer review
-before publication, investigation scrutinised by experts in field
-aims are to fund research,validate quality and relevance and improve
pros=protects quality of published research
cons=anonymity may be used to criticise rival research, publication bias and ground breaking research may be buried as it contradicts others views
psychological research and the economy
-research stressed importance of fathers in psychological development
-may promote flexible working arrangement in families
-modern parents better equipped to contribute effectively to economy
-one third of days off work caused by mental disorders
-gives access to drugs such as SSRi’s
-people with mental disorders can attend work and contribute to economy