Research Methods Flashcards
Aims definition
A statement of what the researcher intends to find out
Define independent variable
Variable that is manipulated to cause effect on the DV
Define dependent variable
Variable that is measured and changes as the IV is manipulated
Define hypothesis. What are the three types?
testable statement of expected findings. Directional, Non- directional, null
What is a directional hypothesis?
States expected direction eg better, higher.
Used when there is PREVIOUS RESEARCH to indicate direction
What is a non- directional hypothesis?
States difference but not direction eg difference, affects.
Used when NO PREVIOUS RESEARCH
What is a null hypothesis? What has to be added on the end?
a statement of no relationship between population and respect to variables.
… but if there is change it is due to chance
How would you start a written hypothesis?
There will be a significant…. (increase/ decrease/ effect/ positive correlation/ difference)
What are standardised procedures?
each pp does EXACTLY THE SAME THING IN EACH CONDITION. Consider: ethical issues, informed consent, debriefing
Define extraneous variable
A variable that is not the IV or DV but is controlled so it doesnt effect the DV
Define confounding variable
Variables that are missed and so not controlled so effect the DV and lower the validity
What is meant by balancing control and realism?
Control variables yet be be real enough to generalise to real life
Define validity
ACCURACY of results
Define internal validity and give examples
degree to which the research tested what it was intended eg demand characteristics, socail desireablility bias
Define external validity and give the three types
extent to which results can be generalised to the real world.
1) Population- generalise to population
2) ecological- Generalise to other situations
3) historical- genralise old studies to modern day
Define the scientific method
observation, identification, description, experimental investigation and theoretical explanation of phenomena
Define empirical methods
evidence gained through direct observation rather than thoughts and beliefs
Why does psychology opt for science?
no bias/ expectation, controlled, validated or falsified
Define objectivity
not effected by personal expectations
Define replicability
repeating a study to verify findings- usually different pp and different task
Define theory construction and state the difference between the two models
Set of principles to explain facts and enable predictions to be made an tested.
Inductive= theory at end
Deductive= theory after observations
What is hypothesis testing?
Conducting research studies to assess the validity of a hypothesis
Define falsifiability
Ability to prove something is wrong via tests- that an alternative may be correct
Define paradigm
Shared set of assumptions about a scientific matter
Define paradigm shift
progress in science causes questioning of accepted paradigm and when enough contradictory evidence paradigm shift occurs
What is a confederate?
person who ‘plays a role’- is part of the research set up but not a participant
What is a pilot study?
small scale trial run of a research design to make appropriate changes eg timings, variables
What is the repeated measures design? what are the limitations and how are they dealt with?
every pp carries out each condition.
L= order effects, guess aim
D= counterbalancing (each condition tested 1st and 2nd in equal amounts
What is the independent groups design? what are the limitations and how are they dealt with?
each group does one condition.
L= individual differences. more pp needed
D= random allocation
What is the matched pairs design? what are the limitations and how are they dealt with?
Two groups, pairs matched of key characteristic and put in different groups.
L= time consuming, difficult
D= restrict no of variables, pilot study
What is a lab experiment? what are some limitations?
controlled environment, pp aware of study, may alter behavior, unrealistic
What is a field experiment?
natural environment with controlled IV to measure DV
One strength of lab experiment
high internal validity, highly controlled, BUT low ecological validity
One strength of field experiment
more natural, BUT less control of variables- lower internal validity
What is a natural experiment?
natural environment but IV is naturally occurring and not manipulated
What is a quasi experiment?
IV naturally occurring as naturally existing difference eg gender
Two limitations of natural/ quasi experiments
1) IV not directly manipulated so results may be confounding variable.
2) pp not randomly allocated- may be other variables, bias
What are demand characteristics?
CONFOUNDING VARIABLE. cues that make pp unconsciously aware of aims of study so change behaviour
What are investigator effects?
CONFOUNDING VARIABLE. cues from investigator that cue behaviours that fulfill their expectations- direct or indirect
Two ways to deal with demand characteristics
1) single blind
2) double blind
What is opportunity sample? one + and one -
convenience or availability
+ easiest- less time
- bias
What is random sample? one + and one -
equal chance of selection eg computer program
+ unbiased
- takes time
What is volunteer sample? one + and one -
advertise for pp
+ variety of pp
- VOLUNTEER BIAS
What is statified sample? one + and one -
collecting pp in proportion to group they’re representing
+ representative
- time consuming
What is systematic sample? one + and one -
predetermined system eg every nth person
+ unbiased
- not completely bias
What are ethical issues?
Conflict between what the researcher wants to do an the pp rights
How many ethical guidelines are there? What are they?
6
- informed consent
- right to withdraw
- deception
- protection from harm
- confidentiality
- privacy
What is informed consent
pp know true aim so can fully consent
What is deception?
necessary to avoid demand characteristics, difference between withholding info and giving false info
What is the right to withdraw?
important id pp distressed
What is protection from harm?
leave study in same state as arrived
What is confidentiality?
personal info only recorded is pp not identifed
What is privacy?
not observed in certain situations eg own homes/ without awareness- invasion of privacy
How do you deal with ethical issues in general (4)?
BPS give guidelines of which behaviours are acceptable.
Cost- benefit analysis,
Ethics committees,
Advice not punishment (should not must)
How do you deal with informed consent?
give signed consent, use presumptive consent
How do you deal with deception?
approved by ethics committee, fully debriefed
How do you deal with right to withdraw?
informed at start that they can withdraw
How to deal with protection from harm?
stopped if harm suspected
How to deal with confidentiality?
use false names
How to deal with privacy?
have to have informed consent
What is a naturalistic observation? One strength
natural situation, researcher doesn’t interfere.
S- high ecological validity especially if covert
What is a controlled observation?
some environmental aspects controlled, reduces realism
What is an overt observation?
pp aware of study, effects validity.
What is a covert observation? One problem
pp not aware of study.
L- ethical issues (no informed consent, respect privacy)
What is a non-participant observation?
observer watches from distance
What is participant observation?
observer joins group, INTER-OBSERVER RELIABILITY
What are unstructured observations? Two problems
record all relevant behaviour
1) too much to record
2) most noticeable not most relevant
What are structured observations? One positive
decide relevant behaviour- behavioural categories
1) objective observations
What are behavioural categories?
behaviour operationalised and put into categories
- objective
- comprehensive (cover all possible)
What are sampling procedures? What are the two types?
Ideally record every instance of behaviour. Gain estimate of frequency.
1) Event sampling
2) Time sampling
What is Event sampling?
counting how often a certain behaviour occurs in a period of time
What is time sampling?
Record what target is doing every 30 seconds (eg) by ticking checklist
Self-report techniques: What are questionnaires? one advantage
pre-determined set of written questions that find out what people think and feel- objective and scientific if correct.
A- Easily distributed- large sample, more impersonal so personal info, reduces social desireability bias
Self-report techniques- What are structured interviews? One limitation and one advantage
Pre-determined set of questions asked face to face. No deviation from questions.
L- interviewer bias- might be unconscious so should be trained
A- easier to analyse and form statistics
Self-report techniques- What are unstructured interviews? One limitation
Starting questions but conversation changes depending on answers given
L- interviewer bias- might be unconscious so should be trained
What is reliability?
CONSISTENCY of measurements
Three ways to assess reliability
1) test-retest (same group, same task, time interval)
2) inter-observer (2 or more to agree on observations)
3) split- half (questionnaire- 2 sets of differently worded questions
Three ways to improve reliability
1) reduce ambiguity of questions
2) operationalised behavioral categories
3) use standardisation of measures (same procedure for each pp)
Two ways to assess validity
1) face (surface level- does it look accurate?)
2) concurrent (comparing existing results with the area you are interested in)
Four ways to improve validity
1) (low face) make questions more relevant
2) (low concurrent) remove irrelevant questions
3) improve design eg double blind
4) pilot study (check ethics, methodology, feedback)
Define probability
numerical measure or the likelihood or chance that certain events will occur
Define chance
something without cause
What level of probability do psychologists mainly use and what does it mean?
p= 0.05 (95% probability that the outcome is accurate, 5% due to chance)
What level of measurement is used in more serious research or challenging research?
p= 0.01 ( 1% probability result was due to chance- results more accurate)
What is a TYPE I error?
FALSE POSITIVE
significance level too low, results always significant so null hypothesis rejected falsely
What is a TYPE II error?
FALSE NEGATIVE
significance level too high, results never significant so null hypothesis falsely accepted
What is central tendancy? What are the THREE measures of central tendancy? Evaluate each
AVERAGES
- mean (sensitive to anomolies)
- mode (useful for nominal data)
- median (ignores extremes)
What are the FOUR LEVELS of measuremtent? (NOIR)
- Nominal (categories)
- Ordinal (ordered/ ranked)
- Interval (equal intervals between units and CAN go below 0 eg. temperature)
- Ratio (equal intervals between units but CAN’T go below 0 eg. money)
What is dispersion? What are the TWO measures of dispersion? Evaluate each
SPREAD OF DATA
- standard deviation- difference of each value from mean (effected by extremes)
- range (doesnt account for data distribution)
What are parametic tests?
more powerful. criteria:
- interval or ratio data level
- conform to normal distribution
- no significant variance (spread of data is not 4x different)
LEARN TABLE
LEARN TABLE- PHOTO
THREE things to consider when writing a questionnaire
1) Clarity of questions
2) Bias- no leading questions/ social desireability bias
3) Analysis (open questions better but harder to analyse)
TWO things to consider for an interview
1) recording (notes, audio, video)
2) listening skills- don’t probe questions, allow time to answer
Advantage and disadvantage of open questions
A- qualitative and detailed answers that are insightful
D- harder to analyse
Advantage and disadvantage of closed questions
A- easier to analyse and produce stats
D- forced to choose unrepresentative answers