Research methods next steps Flashcards
GRAVE
Generalisability
Reliability
Applicability
Validity
Ethics
SCOUT
Supporting evidence
Conflicting evidence
Opposing theories
Usefulness
Testability
What is GRAVE used for?
Research Experiments
What is SCOUT used for?
Theories.
Ecological validity
can it be applied to everyday life?
Population validity
can it be applied to everyone?
Temporal validity
is it outdated? can it still be applied today?
Demand Characteristics
when the participants figure out the aim of the study and change their behaviour according to it.
BPS
British Psychological Society
APA
American Psychological Association
What is the phrase used to remember Ethical Guidelines?
Can
Do
Can’t
Do
With
Participants
Can
Do
Can’t
Do
With
Participants
consent
debrief
confidentiality
deception
withdraw (right to)
protection from harm (physical and psychological)
What happens if a participant withdraws, but they were paid?
Normal procedures must still occur, so if they were supposed to be paid, you still have to pay them, even if the withdraw.
What must be done before an experiment can be carried out?
The board of ethics at your university etc. must approve the experiment before it can be carried out.
Generalisability
whether we can apply the sample to other conditions/populations
Reliability
the consistency of a research study or measuring test
Applicability
the extent to which we can apply the findings to actual use
Validity
the study measuring what it’s intended to measure
Ethics
following the framework provided by the BPS
Supporting evidence
evidence that implies the theory is correct
Conflicting evidence
evidence that implies the theory is wrong
Opposing theories
other theories that suggest contrasting things
Usefulness
whether or not the theory can be applied/is useful
Testability
whether or not the theory can be proven by research
Aim
general statements that describe the purpose of an investigation, developed from theories
Hypotheses
testable statements which predicts the outcome at the start of a study
Directional hypothesis (one tailed)
the research makes clear anticipation of experimental outcome, used when there is sufficient evidence displaying a particular outcome
Non-directional hypothesis (two tailed)
states that there is a difference between conditions but the nature of the difference is not specified, used when there are limited or mixed findings
Null hypothesis
predicts no significant difference between conditions
Alternative hypothesis
predicts a significant difference between conditions
Operationalisation
ensuring variables can be well measured/tested, to establish cause and effect
Four types of variable
Independent variable: variable that is changed
Dependent variable: variable that is measured
Extraneous variable: any variables that might potentially interfere with the IV
Confounding variables: A type of EV, related to the IV and can have an impact on the DV.
What are the four types of experiment?
Laboratory, Field, Natural, Quasi
Lab experiment: meaning and eval
conducted in a room/lab setting where variables are well controlled by the researcher
high control over variables, so clear cause and effect, so high internal validity
standardised procedures, so easy to replicate
lack ecological validity due to high control, artificial environment
demand characteristics - awareness of experiment’s aim
Field Experiment: meaning and eval
IV is manipulated in a natural, everyday environment
high mundane realism and ecological validity due to natural environment.
less likely to have demand characteristics as participants unaware they’re in an experiment
loss of control over CVs and EVs.
ethical issues when participants are unaware of the experiment
Natural Experiment: meaning and eval
researcher does not have control over the IV, but the effect is still measured
necessary cannot manipulate IV due to practical/ethical reasons
high external validity: involving real-world issues and problems.
naturally occurring events might be rare, making it difficult to generalise
difficult to identify causation when IV cannot be manipulated.
Quasi Experiment
based on an existing difference between people, not manipulated by anyone
carried out under controlled conditions - shares some strengths of a lab experiment
difficult to infer causation
What are the three experimental designs?
Independent groups, repeated measures, matched pairs
Independent groups: meaning and eval
two+ separate groups of participants experience different conditions.
two levels of IV: only experimental and control condition.
order effects prevented when participants only participate in one condition.
demand characteristics also prevented because participants only experience one condition
CVs and EVs affecting the results, due to individual differences
time and money-consuming.
Repeated measures: meaning and eval
all participants experience both conditions of the experiment.
individual differences are controlled.
fewer participants needed
order effects: same group of participants participate in two conditions (can be prevented by counterbalancing)
demand characteristics
Matched pairs: meaning and eval
participants put into pairs based on specific similarities (e.g IQ), and the pairs are split into the two different conditions to make the groups as similar as possible
better than independent groups at controlling participant variables and minimising individual differences.
time-consuming and expensive.
extremely difficult to fully eliminate individual differences.
Social desirability
a tendency for participants to present themselves in a generally favourable light.
Investigator effects
undesired influence of the investigator on the research outcome
What are the three investigator effects?
Expectancy effect
Unconscious bias
Leading questions
Expectancy effect
one’s behaviour is affected by how they expected the outcome will be
Unconscious bias
subjective decisions and judgements made by the investigators
Leading questions
asking questions that guide participants to answer in specific ways
Randomisation
the use of chance methods to reduce unconscious bias and therefore investigator effects
Standardisation
all participants are subjected to the same environment, information, and experience, minimising evs
What are the five sampling methods?
Stratified, systematic, opportunity, volunteer, random
Stratified sampling meaning and eval
Reflects the sub-groups of the target population, researchers identify the different strata that make up the population, and work out the proportions needed for the sample to be representative.
Representative as they accurately reflect the population
There could still be individual differences within each sub-group, this means that it is extremely unlikely for a sample to be fully representative
Random sampling meaning and eval
All members of the target population have an equal chance of being selected, sample is selected using chance methods
Potentially unbiased by minimising the decisions made by researchers.
Minimise the effects of extraneous and confounding variables on results, increases internal validity
Can be time-consuming and difficult to create a list of the entire target population.
Might not be representative
Some might refuse to take part, which again makes a fully random approach difficult.
Systematic sampling meaning and eval
Every nth member of the target population is selected
A sampling frame will be created, (a list of people in the target population)
Objective and minimising researcher bias
Similar to random sampling, can be time-consuming and difficult
Volunteer sampling meaning and eval
Participants selecting themselves to be part of the sample
Convenient, saves time and money
Participants tend to be more engaged, providing richer results for the research.
Demand characteristics
Might also be less representative as it can attract people of certain social groups
Opportunity sampling meaning and eval
Participants are selected by whoever happens to be willing and available.
Convenient, saves time and money
Sample can be unrepresentative of the target population as it is normally drawn from a specific area
It can also be unrepresentative due to researcher bias
