Paper 2 Research Methods Flashcards
Co-variables?
comparisons of two variables to see for correlations
Types of hypothesis?
Experimental
Directional
Non-directional
Null
Null hypothesis?
A statement of no difference or relationship
Non directional hypothesis?
A statement where a difference/relationship is present however the direction is not stated
Directional hypothesis?
A statement that states the direction of the difference/relationship
When is a directional hypothesis used?
When previous research suggests a direction in the difference/relationship
Experimental hypothesis?
A statement where the relationship of the IV and DV is stated and the direction of the relationship
Independent groups?
One group of participants do condition A and the second group do condition B
Repeated measures?
All participants take part in all conditions of the experiment
Matched pairs?
Two groups of participants are matched based on characteristics relevant to the study
Types of experimental designs?
Matched pairs
Repeated measures
Independent groups
What are the strengths of independent groups?
Order effects do not interrupt performance
Demand characteristics are reduced
What are limitations of independent groups?
More Ps are needed
Confounding variables may be present
individual differences
What are strengths of repeated measures?
Few Ps are needed
All Ps variables are controlled
What are limitations of repeated measures?
Order effects disrupt performance
Demand characteristics
What are strengths of matched pairs?
Ps variable can be controlled better
What are limitations of matched pairs?
Time consuming
Difficult to match accurately
Only known variables are matched
Counterbalancing?
When half of Ps do condition A then B whilst the other half does visa versa balancing out effect of the order
Order effects?
The change in performance due to the order of conditions
Extraneous variables?
Any variable other than the IV that can potentially affect the DV. They dont vary systematically with the IV.
How can we control demand characteristics?
Single blind design where Ps are unaware of condition they are in
Deception by lying or distracting questions
Investigator effects?
Cues from the investigator that encourages specific behaviours
Situational variables?
Features that influence Ps within situation/environment
Participant variables?
Individual differences between Ps that affect the DV
What happens if you fail to control variables?
They can become confounding variables
Preventions of investigator effects?
Double blind design - both investigator or Ps dont know aims/hypothesis
Inter rater reliability
Preventions of situational variables?
Standardisation so keep everything the same
Procedure, instructions, counterbalancing
Preventions of participants variables?
Consider your experimental design and use matched pairs or repeated measures
If using independent groups do random allocation
Types of experiments?
Field
Quasi
Natural
Laboratory
Labouratory experiments?
Variables are controlled
Special environment
Ps are aware of experiment
Field experiments?
Researchers go to Ps usual environment so they r not aware
IV is manipulated in a natural setting
Natural experiments?
Researcher has no control over the IV as it vaires due to something/someone (natural)
Measured in field or lab
Quasi experiments?
IV can’t be manipulated as based on existing differences (age)
DV can be naturally occurring
What are the strengths of laboratory experiments?
Control over variables
Establish cause and effect relationship
Ethics
Internal validity
What are the strengths of field experiments?
High ecological validity
Reduced demand characteristics
Higher mundane realism
What are the strengths of natural experiments?
High external validity
Research opportunities
What are the strengths of quasi experiments?
Research opportunity
High external validity
Ethics
What is the limitation of quasi and natural experiments?
Confounding variables due to lack of random allocation
What are the limitations of field experiments?
Lack of control over variables
Low internal validity
Ethics
What are the limitations of laboratory experiments?
Low ecological validity
Increased demand characteristics
Low mundane realism
What is a hypothesis
Cleat and concise prediction about the possibly outcomes of the result
3 types types of measurements of the DV
Nominal
Ordinal
Interval
Nominal level?
Categorical data E.g participants choosing the responses yes or no
Ordinal variable?
Used to rank data or put the data in some sort of numerical order E.g. 0-10 scale
Interval variable?
Where data consists of equal intervals or equal measurements. This is much more objective and scientific E.g. swimmers in a race can be measured in mins and secs
Operationalising variable
Where the experiment needs to define the variables and state exactly how accurately and precisely they intend to measure or manipulate it
What are the issues with experiments
Demand characteristics
Researcher bias
Order effects/practice effects/fatigue effects
How do you overcome issues with experiments
-single blind test
-double blind test
- standardised procedures
- counterbalancing
- randomisation
What are participant variables
Variables the participants bring to the study that may affect the measurement of the DV
Situational variables?
Variables that should be controlled in the set up of the experiment E.g. temperature, lighting, time of day
Extraneous variables
Variables that may affect the measurement of the DV and therefore should be controlled by the experimenter
What are demand characteristics
Participants are unsure how to behave and they try working out what is required of them
What is researcher bias
When either consciously or subconsciously researchers may impose a bias on the collection of data to be favourable towards there hypothesis
Randomisation?
Random allocation of participants to a group to reduce the chances of individual differences
What are pilot studies
Small scale prototype of a study
They use smaller scale sample of the target population
What is Reliability
A measure of consistency
What is the test retest method
Where researchers ask the same participants to complete the same test at different times and expect to see the same results
How can you improve the reliability of experiments
Pilot studies, standardised procedures and controlling extraneous variables
internal validity?
the extent to what is intened to be measured, is measured IN DIRECT CONSEQUENCE to the IV
Ways to increase internal validity?
control over cofounding/extraveous variables, bias from researchers/Ps
standardisation
operationalisation
External validity?
how outside factors are relevant to the experimen
e.g. generalisation
types of external validity?
ecological - to what extent can this be generalised to other settings
population - to what extent can this be representative to populations
temporal - to what extent can this be applied over a range of time
ways to assess validity?
face validity - self report using intutive
concurrent validity - comparision with well established test
what is a high concurrent validity?
+.80
ways to improve validity?
experiments - standardise, control group
questionaires - lie scale
observations - covert
qualitative research - triangulation
ways to assess reliability?
test retest
measuring reliability (0.80)
inter-observer reliability
test retest?
measuring tools such as IQ tests/questionaires used for a retest after a sufficient amount of time
correlation coefficient?
0.80 (correlation analysis)
inter-observer reliability?
subjective data obtained by 2+ observers from observation through pilot studies
ways to improvement reliability?
Questionaires - no complex/subjective questions
interviews - structural
observations - behavioural catergories
experiments - precise replication
samples are used for?
representation
less time consuming and costly
sampling techniques?
volunteer
systematic
random
stratified
opportunity
so what? volunteer sampling?
less time-consuming
more enaged sample
volunteer bias limits generalisation
so what? random sampling?
increases internal validity
unbias
time-consuming
potentially unrepresentative
so what? systematic sampling ?
objective method
time-consuming
need sample framework
sample units can refuse
so what? stratified sampling?
time-consuming
representative so generalistion
need sample framework
so what? opportunity sampling?
convient
less costly
reseacher bias
bias so no generalisation
ethical issues?
present when there is conflict between Ps rights and the goals of the research conducted
BPS code of ethics?
quasi legal document that instructs the behaviours acceptable when dealing with Ps
BPS 4 principles?
responsiblity
respect
intergrity
competence
informed consent?
what? Ps are aware of aims of the study + what dat will be used for
how? consent form/letter or parental consent
so what? informed consent?
could result in less natural behaviour
types of consent?
presumptive - simple group of ppl are asked if study is acceptable ‘presumed’
prior general - Ps give permission to be in multiple studies which could involve decepton
retrospective - Ps are asked at debriefing so can be subjected to deception
Deception?
what? Ps are deliberately lied to or have info withheld
how? no informed consent provided
so what? deception?
Ps need a full debreif entailing true aims and what data will be used for IF given the right to use
protection from harm?
what? Ps shouldn’t be put in risk of harm psychological/physical
how? embarassed, under preesure, inadequate
so what? protection from harm?
BPS code of conduct to see if ethically acceptable by ethical comittees
privacy and confidentiality?
what? Ps have to right to privacy so ares/ppl shouldn’t be named
how? confidentiality should be put in place
so what? privacy and confidentiality?
keep anonymity by using no record of personal details
in de/breifing make sure data is protected
event sampling?
an event is recorded every time it occurs
established by target behaviour/event
why would event sampling be used?
when event is not frequent
time sampling?
recording behaviour within a fixed time frame
a tagert individual/group needs to be established 1st
why would time sampling be used?
to reduce the number of observations though this could be unrepresentative
so what? qualitative data?
richness of detail so increase external validity
difficult to analyse so subjective interpretation due to no idetifcation of patterns/comparsions
so what? quantitative data?
easier to analyse so comparisions can be made
narrower in meaning as less detail so decreases external validity
so what? primary data?
information is directly relevant to research aims as designed to be
time consuming/costly
so what? secondary data?
less costly
less quality as may be outdated or incomplete decreasing validity
so what? meta analysis?
increase validity of conclusions so generalisation can be made
publication bias so may lack validity
so what? questionaires?
straight forward to analyse so can be converted into grapghs/charts for comparision
can be distributed quickly so reduces effort
response bias
social desirability
structured interview?
list of questions issued in a pre-fixed order
unstructured interview?
no set of questions instead a general topic is discussed so free-flowing
so what? structured interview?
easy to replicate so standardised format
interviewers cannot give subjective observation
so what? unstructured interview?
greater flexibility so gain insight on unexpected info
risk of interviewer bias so opportunity for unconscious cues
semi-structured interview?
same as structured THOUGH can ask further questions BASED OFF previous questions
correlation?
the strength and direction of an association between 2 variables
so what? correlation?
no cause & effect due to no manipulation so could lead to false conclusions
intervening variables could lead to false conclusions
less time-consuming
could start for a hypothosis if there is a strong correlation
behavioural catergories?
set of observable catergories used in observing target behaviour
so what? behavioural catergories?
difficult to make clear and unambiguous
so what? observations?
insight onto sponteneous behaviours
risk of observer bias
types of observations?
naturalistic
covert
overt
controlled
non/participant
naturalistic observation?
takes place where the target behaviour would normally occur
controlled observation?
some control/manipulation of variables including EV/CVs
covert observation?
Ps are unaware they are being observed
overt observation?
Ps are aware the are being observed
P observation?
researcher becomes apart of the group being observed
non-P observation?
researcher remains separate from the group
so what? non-P observation?
more objective so increases internal validity
loss of insight reducing external validity
so what? P observation?
greater insight so increases external validity
lose of objectivity so threatens internal validity
so what? overt observation?
more ethically acceptble as given consent so right to withdraw
demand characteristics reduces internal validity
so what? covert observation?
reduced demand characteristics so increases internal validity
not ethically right
so what? controlled observation?
can be replicated due to standardised format
decrease in external validity
so what? naturalistic observation?
high external validity so generalisation
low control as may uncontrolled CV/EVs so difficult to detect patterns
