Research Methods Flashcards
Types of experiments?
Lab, field, natural
Evaluate a lab experiment ?
S-control of extraneous variables
W-can lack validity because people are overly aware they are being observed
S-more reliable as conditions are kept similar
Evaluate a field experiment?
S-natural setting provokes natural behaviour
W-may lose control of evs so cause and effect is less shown
Evaluate natural experiment?
S-high validity due to real life variables
W-more likely to have evs
W-hard to find opportunities for natural experiments
Types of experimental design?
Independent groups, repeated measures, matched pairs
Evaluate independent groups design?
S-order effects are not a problem
W-participant variables can act as evs
Evaluate repeated measures?
S-no participant variables
S- less participants=less expensive
W-order effects can act as evs
Evaluate matched pairs?
S-no order effects, fewer participant variables
W-lengthy process to match participants
Evaluate a questionnaire?
S-thousands can be sent out and done at the same time
W-can’t use follow up questions to get into further detail
Evaluate an interview?
S-you can use follow up questions to get more information when needed
W-information is more valid as it’s face to face
Independent variable?
Manipulated by the experimenter
Dependent variable?
Depends on the IV, measured
Aim?
A general statement that explains the purpose of the study
Operationalisation?
Making variable measurable
Alternative hypothesis?
Statement of relationship between variables or affect on eachother
Null hypothesis?
A statement of no relationship between the variables
Extraneous variables ?
Things that have an affect on the DV other than the IV
Instructions to participants?
Should be standardised, the same instructions should be told to every participant. A researcher should write down exactly what they will say.
Standardised procedures?
The only thing that should vary in the studies are the IVs, easiest in lab setting, same environment, same instructions
Randomisation?
Randomisation includes using chance in the experiment. Aspects of the experiment should be allocated randomly to avoid bias
Sample?
Has to be representative of the target population
Generalisation?
Your sample has to be able to be generalised to the target population. This means it’s applicable.
Random sampling?
Names in a hat, completely chance
Random sampling evaluation?
S-includes chance, no bias
W-takes more time and effort
Opportunity sampling?
Researcher selects people who they happen to come across
Opportunity sampling evaluations?
S-easy to do
W-unrepresentative of sample
Stratified sampling?
Participants are selected from different subgroups of a target population
Stratified sampling evaluations?
S-most representative
W-lengthy process
Systematic sampling?
Selecting every nth person from the target population
Systematic sampling evaluations?
S-avoids researcher bias
W-representativeness not guaranteed with chance
Ethical issues examples?
Deception, privacy, confidentiality, harm, informed consent
Deception?
Too much lying to the participant(difference between lying and withholding information) can decept if benefits outweigh. Can debrief at end.
Protection from harm?
Participants should not be placed at risk physically or psychologically
Privacy?
People should be able to control information about themselves
Confidentiality?
If someone’s privacy had been invaded their identity should be protected
Informed consent?
The participants should know enough about the study to be able to give informed consent. If not they will have to be debriefed at the end of the study and have a right to withdraw their data.
Observation types?
Naturalistic vs controlled, covert vs overt, participant vs non participant
Naturalistic observation?
Observing a behaviour where it would naturally occur, taking place in it’s natural setting and not changing anything about the environment.
Controlled observation?
Observation with aspects manipulated by the observer or in a ‘fake’ setting
Covert observation?
Participants are not aware they’re being observed(must be public place)
Overt observation?
They’re aware they’re being observed
Participant observation?
If the observers are taking part with the participants in order to observe
Non-participant observation?
Observers aren’t taking part with participants, simply observing and recording
Inter observer reliability?
Must be more than one observer, compare observations they got and should have similar records written to be reliable.(to avoid investigator bias)
How to check inter-observer reliability?
Make categories of behaviour, compare records, alter behavioural categories if they aren’t getting similar findings
Observation strengths?
High validity- in questionnaires or interviews they can lie, we see what they would actually do in action
Especially high in naturalistic settings or covert
Observation weaknesses?
Ethical issues- invading privacy if covert(but observation is supposed to be natural setting)
Observer bias
What do correlations show?
Show a relationship or link between two co-variables, tell us strength and direction of link
