research methods Flashcards
hypothesis
a clear, precise, testable statement - shows relationship between variables- stated at the outset of any study
aim
a general statement of the research- the purpose of the study
directional hypothesis
shows the direction of the difference between 2 conditions
based on previous research findings
non-directional hypothesis
states that there is a difference between 2 conditions, but direction not specified
null hypothesis
what is assumed is true during the study
data collected either supports or rejects the null hypothesis
it is often a prediction of no correlation between variables
alternative hypothesis
used when the null hypothesis is rejected
can be (non) directional
levels of the IV
control: without IV
experimental: with IV
operationalisation
clearly defining variables in terms of how they can be measured
to make the hypothesis clear and testable
extraneous variable
any variable other than the IV that may have an effect on the DV if not controlled
do not vary systematically with IV
confounding variable
essentially an uncontrolled EV
any variable other than the IV that may have affected the DV so that we cannot be sure of the true source of change to the DV
vary systematically with the IV
demand characteristics
any cue from the research situation that may be interpreted by participants as revealing the purpose of the investigation
this may lead to participants changing behaviour (please/screw-U)
investigator effects
the effect of the investigator’s (un)conscious behaviour on the DV
may include design, selection, interaction
randomisaiton
the use of chance to control for the effect of bias and investigator effects when designing materials and when deciding the order of conditions
also can avoid order effects
standardisation
using exactly the same formalised procedures and instructions for all participants in a study, including the environment
writing a hypothesis
- state whether there will be a significant difference/ correlation
- include the IV + DV
+ LEVELS - operationalise
- state direction if necessary
experimental design
the different ways in which the testing of participants can be organised in relation to the experimental conditions
types of experimental designs + definitions
independent measures: participants allocated to different groups where each group represents 1 experimental condition
repeated measures: all participants in all conditions
matched pairs design: pairs of participants are matched on variables that may affect the DV, one of each pair in each condition
- to control for the confounding variable of participant variables
- may necessitate a pre-test
independent measures- evaluation
strengths
- no order effects (learning/fatigue)
- only aware of 1 condition- less likely demand characteristics
weaknesses
- participant variables- may cause confounding variables
- no. of participants- 2x as many are needed compared to repeated
random allocation
to control for participant variables in an independent group design- ensures each participant has the same chance of being in 1 condition as another
repeated measures- evaluation
strengths
- no participant variables- same people in each condition
- no. of participants is fewer for same amount of data
weaknesses
- order effects- repeating tasks may lead to learning/fatigue effects + demand characteristics
counterbalancing
control for the effects of order in repeated measures
- half experience in 1 order, half in the other
matched pairs- evaluation
strengths
- no order effects
- participant variables- important differences are minimised by matching
weaknesses
- no, participants- 2x as many needed compared to repeated
- practicalities- time consuming, expensive, difficult to pre-test and identify pairs
- participant confounding variables are still likely
in experimental design, consider
order effects, participant variables, number of participants, demand characteristics. practicalities
types of experiment + definitions
lab: in a controlled environment, researcher manipulates IV and records effect on DV, whilst maintaining strict control of EV
field: takes place in a natural setting- researcher manipulates IV and records effect on DV
natural: the change in IV is not brought about by the researcher, but would have happened had they not been present
researcher records effect on DV.
IV = natural and not necessarily the setting so participants can be tested in a lab
quasi: IV is based on an existing difference between people- it is naturally occurirng, such as gender
lab experiment- evaluation
strengths
- high control over EV- can ensure that any effect on DV is a result of manipulating the IV
- can therefore establish cause and effect (causal relationships)- high internal validity
- replication = possible due to high control
ensures new EV are not introduced when repeating an experiment
replication = vital to check whether results are valid
weaknesses
- may lack generalisability- the lab environment is artificial and therefore does not measure real life behaviour
- so lacks ecological validity
participants may act in unusual ways due to the context, so their behaviour cannot be generalised outside the research setting- low external validity.
- demand characteristics may bias the result as participants may respond to what they thing is being investigated
- low mundane realism- tasks are not representative of real life
- deception = often used, so informed consent is difficult
field experiment- evaluation
strengths
- higher mundane realism than lab- natural setting increases ecological validity
- demand characteristics can be avoided if participants are unaware they are in a study, increases internal validity
- can still establish causal relationships by manipulating IV and measuring DV
weaknesses
- less control over EV- confounding variables = more likely
so causal relationships are more difficult to establish
- if deception is used, participants cannot give informed consent
natural experiment- evaluation
strengths
- provide opportunities for research that may not be undertaken for practical or ethical reasons
- high external validity- involve the study of real life issues
- high ecological validity- less artificial than lab
weaknesses
- naturally occurring experiments may only happen very rarely- this decreases opportunities for research- also hard to generalise
- difficult to establish causal relationships- IV is not directly manipulated and confounding variables are likely
- deception is often used, making informed consent difficult
confidentiality may be compromised if community = identifiable
quasi experiment- evaluation
strengths
- often conducted under controlled conditions
- higher ecological validity than lab- research less artificial
weaknesses
- no control over participant allocation to each condition so confounding variables e.g. where they live may effect results
- hard to establish causal relationships because IV is not directly manipulated
population
a group of people who are the focus of the researcher’s interest, from which a smaller sample is drawn
in types of experiment, consider
control of variables/ conditions
participant variables
validity; ecological, internal, external
generalisability
demand characteristics
deception - consent
causal relationships
replication
sample
a group of people who take part in a research investigation
the sample is drawn from a target population and are representative
types of sampling + definitions
random sampling: every member of the target group has an equal chance of being selected e.g. random number generator
opportunity sampling: samples with whoever is available and willing
volunteer sampling: people actively volunteer by responding to a request
systematic sampling: every nth name from a sampling frame
stratified sampling: all the important subgroups are identified and a proportionate number of each are randomly obtained from each strata
random sampling- evaluation
strengths
- free from researcher and volunteer bias so more likely to be representative
weaknesses
- does not guarantee a representative sample- some majority subgroups may not be selected
- if the group is particularly large, assigning people numbers may be especially time consuming
- selected participants may refuse to take part- leads to opportunity sampling
opportunity sampling- evaluation
strengths
- quick and practical
weaknesses
- unlikely to be representative of target population due to researcher and volunteer bias
- also difficult to generalise the findings- the sample is taken from a specific area
volunteer sampling- evaluation
strengths
- quick, easy, cheap
- a lot of people may respond to the advert, increasing sample size = more accurate
weaknesses
- volunteer bias so sample is unlikely to be representative of target population
a volunteer is a certain type of person- so hard to generalise
systematic sampling- evaluation
strengths
- simple and effective
- no researcher/volunteer bias
- more likely to be representative than e.g. volunteer or random
weaknesses
- subgroups may be missed
- not representative if pattern used for samples coincides with pattern in population
stratified sampling- evaluation
strengths
- produces a representative sample with no bias- so possible to generalise
weaknesses
- expensive and time consuming
- important subgroups may still be missed
- can be difficult to identify traits and characteristics to properly stratify
in sampling, consider:
representative
bias (researcher/ volunteer)
generalisable
cheap/ practical
equal chance of being chosen?
why are there ethical issues in psychology
ethical issues arise when a conflict exists between the rights of participants and the goal of research- to produce authentic, valid, worthwhile data- so the correct rules of conduct are necessary
who publishes ethical guidelines for research
BPS
what are 6 ethical issues in psychology
informed consent
deception
protection from harm
debriefing
confidentiality
right to withdraw
what should be in a consent form
purposes of study, aims
the conditions, what the study involves
the duration of study
the right to withdraw
protection from harm
difficulties with obtaining informed consent
children under 16 need parental consent
mentally unwell people cannot give informed consent, but doctors and family members can on their behalf.
alternatives of getting consent
presumptive consent rather than obtaining from participants, a similar group of ppl = asked if a study is acceptable- if this group agrees, consent = ‘presumed’
prior general consent: participants give permission to take part in a number of different studies, including ones with deception
retrospective consent: asking for consent during debriefing
deception - why use it, why is it good, what is the problem with it
BPS: if participants have been deceived, they have not given informed consent
deception usually avoids demand characteristics- increases internal validity
BPS: deception is only acceptable if there is a strong scientific justification- and there is no alternative procedure
nb- the severity of deception can differ between studies
protection from harm
BPS: the risk of harm should be no greater than they would face in normal life
minimise distress inc. feelings of inadequacy, stress
minimise long term effecrs i.e. no frightening, endangering, offending
no permanent -ve impact
can receive counselling
right to withdraw @ any point
debriefing
return the participants to the state they were prior to the research
this is especially important if deception is used
the researcher must fully explain the research and results. and answer any questions
participant has right to withdraw
confidentiality
no participants should be identifiable from reports
data collected must be confidential
participants must be informed if data is not completely anonymous
includes the right to privacy of where the experiment took place
right to withdraw
participants can leave at any time, can withdraw their data
problems with ethics in psychology
even when guidelines are followed, it is difficult to asses effects like psychological harm, or justify the use of deception
deciding whether the ends justify the means is not easy
animal rights
- animal research has provided valuable information for psych and medicine
- some believe it is ethically wrong to inflict harm and suffering on animals- they cannot give consent
- some believe it is cruel to experiment on animals with similar intelligence to humans
- animal research could be reserved for less developed animals, but they are then so different from humans that this becomes too difficult to generalise
self report
any type of data that involves the participants reporting their own perspectives
interview
researcher directly asks the participant questions
structured interview + evaluation
fixed set of questions that are the same for all participants
strength
- like questionnaires, straightforward to replicate due to standardised format, also minimises differences between interviewers
weakness
- not possible to elaborate on points so limited information is collected
unstructured interview + evaluation
there may be discussion topics but it is less constrained- can elaborate
strength
- much more flexibility than structured- can gain a more well rounded view
weakness
- analysis of data is more complex, drawing conclusions is more difficult
semi-structured interview + evaluation
set questions but follow up questions can be asked
strengths
- like questionnaires, straightforward to replicate due to standardised format, also minimises differences between interviewers
- much more flexibility than structured- can gain a more well rounded view
weaknesses
- analysis of data is more complex, drawing conclusions is more difficult
interview- evaluation
strengths
- rich data: detailed information, fewer constraints than questionnaires
- useful in pilot studies- easy way to obtain data
- unlike observations, can access people’s thoughts and feelings that cannot be seen but can be asked about
- unlike questionnaires, can clarify what qs mean, increasing validity
weaknesses
- self report: unreliable, effected by social desirability bias
- impractical: time consuming, requires skilled researchers
- analysing lots of qualitative data is hard
- relies on people being able to explain their thoughts and feelings
thematic analysis
analysing qualitative data by identifying patterns in material (not pre-determined categories)
content analysis
analysing qualitative data using coding units e.g. themes
create checklist of relevant categories- tally behaviours
questionnaires + what to consider
participant given a set of pre-set questions they must respond to
to consider
1. type of data: qualitative/ quantitative
- open qs: reply in any way and in detail- hard to analyse
- closed qs: limit the answers that can be given- less detail but easier to analyse
- ambiguity: avoid question and answer options that are not clearly defined
- double barreled questions: participants may wish to answer differently for each part
- leading questions: may lead ppt to particular answer
- complexity: use clear English, avoid jargon
questionnaires- evaluation
strengths
- practical: can collect lots of data quickly and cheaply
- can be completed without researcher present
- can make comparisons easily
- participants more willing to answer honestly bc. anonymous
- decreases investigator effects- reactions not visible
weaknesses
- leading questions
- biased samples- some people are more likely to respond so unrepresentative sample
- self-report- social desirability bias
- ethics- confidentiality around sensitive topics = problem
acquiescence bias
the tendency to agree with items on a questionnaire, regardless of the contents of the questions
designing questionnaires- scales/ choice options
likert scale: strongly agree –> strongly disagress
rating scales: 1–> 5
fixed choice options: choose applicable from list
in self-report, consider:
amount/ quality of information
bias (social desirability, volunteer, researcher, acquiescence)
validity (clarifying answers)
practical / ease of analysis
validity
how well a test measures what it claims to
reliability
how consistent and dependable a test is
the Hawthorne effect
someone interested in what they’re doing and the attention they are receiving from researchers- they show more positive responses - results may be artificially high
social desirability bias
people try to show themselves in the best possible light to make them appear socially acceptable
researcher bias
researcher’s expectations influence how they design, measure and analyse and the questions they ask
investigator effects
anything the researcher does to affect the behaviour of the participant- could lead to demand characteristics
correlation
the measure of the relationship between two co-variables
correlation co-efficient
number from -1 to 1, shows how closely to variables are linked
3 types of distribution curves
normal: symmetrical, mean=median=mode, width depends on S.D
positive: cluster at lower end, tail on right, mode<median<mean
negative: cluster at higher end, tail on left, mode>median>mean
correlation studies- evaluation
strengths
- variables can be studied that would be unethical to manipulate
- gives ideas for future research
- can be used to test for reliability and validity
weaknesses
- correlation does not mean causation, only link
- 3rd variable may be responsible- controlled conditions are necessary
naturalistic observation + evaluation
observing subjects in their natural environment- no interference with subjects
strengths
- ecological validity- no demand characteristics, natural behaviour increases validity
- theory development- can develop ideas about behaviour that could later be tested in more controlled conditions
weaknesses:
- EV- cannot be controlled, so replication and generalisability decreases
- observer bias-expectations may influence what they focus on and record
- ethics- must respect privacy, only occur where people would expect to be observed
controlled observation + evaluation
carried out in an environment set up by the researcher, managing variables
strengths
- variables are more controlled, EV minimised and cause-and-effect easier to establish
- replication = easier and reliability easier to establish
weaknesses
-lower ecological validity
-demand characteristics- ppts aware they’re observed.
participant observation + evaluation
researcher participates in activity under study
strengths
- develops a relationship with group- greater understanding may increase validity
weaknesses
- may lose objectivity by building relationships- may identify with some more than others
- participants may change behaviour if they know there is a researcher among them
non-participant observation
observes behaviour, but not involved with group
strength
- objective, does not develop bias
weakness
- not part of group dynamic, may decrease understanding/ validity
participant and non participant observation can be
overt/ covert
evaluate both
overt: researcher’s presence = obvious
strength
- much more ethically sound
weakness
- demand characteristics- change behaviour
covert: researcher’s presence = unknown
strength
- demand characteristics minimised so validity increased
weakness
- ethical approval = difficult
recording data:
unstructured observation
structured observation
written notes
video/audio recordings
unstructured observation: writing down everything you see
- rich in detail
- useful in small scale observation w/ few ppts
structured observation: behaviour categories pre-defined
- easy to gather relevant data
- BUT may miss something significant
written notes- useful for qualitative data
video/audio: more accurate, permanent record
categorising behaviour
operationalise what you intend to observe- how it will be measured
behavioural categories as checklist- observable, measurable, self-evident
observations therefore objective, clear focus, increases reliability, easy to analyse
sampling behaviour
unstructured / structured (+types of structured)
unstructured: continuous recording of behaviour
structured: systematic method of sampling observations
- event sampling, time sampling
event sampling + evaluation
only record particular events that you are interested in
strength
- focus exactly on specific behaviours
weakness
- interesting behaviours could be ignores if there are many complex behaviours occurring at one time
time sampling
if behaviours occur over a long period, they may be recorded at set time intervals - chosen randomly
strength
- convenient to carry out
- avoids researcher bias
weaknesses
- interesting behaviours outside time samples are not recorded
so recorded behaviours are unrepresentative of samples as a whole
pilot studies + uses
small scale version of an investigation- takes place before real investigation
- checks that procedures, materials, measuring scales work- changes can be made
- checks that questionnaires/interviews= worded in unambiguous manner
- in observational studies, good way to check coding systems before irl observation
trials
single/double blind, placebo, control
single blind: only the researcher knows what conditions participants are in - control for demand characteristics
double blind: neither the participant nor researcher know who is in what conditions- control for observer bias + demand characteristics
placebo: can be used to check the condition being tested is actually having an effect
control: used for comparison
if the change in behaviour of the experimental group is significantly greater than the control group, the researcher concludes that the cause of this effect = due to manipulation of IV, assuming confounding variables = constant
peer review + main aims
aspects of the investigation being scrutinised by a small group of experts in that field
main aims
1. allocate research funding to proposed research project
2. validate quality and relevance of research + accuracy of hypotheses, methodology and conclusions
3. suggest amendments/ may suggest minor revisions orb that it is completely inappropriate
peer review evaluation
strengths
- minimises bias therefore:
establishes accuracy and validity
makes research more statistically significant
promotes the scientific process
weaknesses
- anonymity
some reviews may use their anonymity as a way of critiscising rival researchers- many researchers are in competition for limited research funding
some journals favour ‘open reviewing’ where names are public for this reason
- publication bias
editors want to publish significant, headline- grabbing findings to increase credibility and circulation
also prefer to publish positive results- this can lead to the ‘file in drawer’ effect
selective publishing creates a false impression of current state of psychology
- burying ground breaking research
peer review process may want to maintain the status quo within particular scientific fields
because reviews are likely long-established researchers, they are perhaps more favourable to research that matches their own view
so new innovative research is more likely to be discarded
therefore, peer review may slow the rate of change within a dicipline
validity - face internal external temporal ecological concurrent
face validity: the degree to which a procedure appears effective in terms of its stated aims
internal validity: whether results obtained are solely effected by changes in the variable being manipulated in a cause-and-effect relationship
external validity: whether data can be generalised to other situations outside the research environment.
temporal validity: research findings successfully apply across time
ecological validity: data collected is generalisable to the real world based on the conditions the research is conducted under.
concurrent validity: whether the test produces the same results as another established measure.
qualitative data + evaluation
expresses thoughts and feelings in a non-numerical way
strengths
- broader in scope - ppt can develop their opinions
- greater external validity- provides researcher with more meaningful insight
weaknesses
- difficult to analyse- to see patterns + make comparisons
- conclusions therefore rely on the subjective interpretation of the researcher
quantitative data + evaluation
data that can be counted numerically
strengths
- simple to analyse and draw conclusions
- data in numerical form is more objective and less open to bias
weakness
- narrower in scope and meaning than qualitative
fails to represent real life
primary data + examples + evaluation
‘field research’- original data, collected specifically for the purpose of the investigations by the researcher - arrives first hand from participants
can be gathered by experimenter e.g. questionnaire, observation, interview
strength
- carefully controlled procedures and operationalised variables
weakness
- expensive to obtain
secondary data + examples + evaluation
‘desk research’ - data that has been collected by someone other than the researcher- it already exists
- has often already been subjected to statistical testing
can be gathered from journals, books, websites
strength
- easier to conduct- a bulk of research has already been carried out
weakness
- higher chance of researcher bias- can choose what data to use
meta analysis + evaluation
results from many different studies brought together to formulate a conclusion
strength
- decreases the problem of obtaining large sample size
weakness
- difficult to formulate conclusions from many conflicting results
case studies + evaluation
- an in-depth study of one person, group or event
- nearly every aspect of the subject’s life and history is analysed to seek patterns and causes of behaviour
- usually carried out in the real world
- idiographic, individualistic
strength
- lots of qualitative data gathered over time
- depth of analysis increases validity
- may stimulate new paths for research
weaknesses
- low control over EV- difficult to establish cause + effect
- poor reliability and replication
- hard to generalise due to unique situation
- retrospective data may not be as accurate or relevant- difficult to verify