Ethics
C in Cardud?
Consent
lab/true?
Very controlled conditions for accurate measurements, standardized procedure, random allocation, manipulation of IV, can control environment- extraneous variables
volunteer, highly motivated, unrepresentative of population
Repeated Measure?
one group does both conditions of the expirement
D in carduD?
Deception
U in cardUd?
Undue stress and harm
Field?
Conducted in natural setting, manipulation of IV, less control of enviroment
Opportunity?
convenience, easy to find because group exists, homogenous, unrepresentative most of time
Independent?
members are randomly allocated to one of the 2 conditions
Hap Hazard sample?
assumes that the population units are alike, can occur over a period of time, includes anyone
A in cArdud?
Anonimity
Quasi?
occurs in natural setting, no IV manipulated, based on personal traits that set them apart,
Random?
All members of population have chance to be picked, "ideal sampling technique", not for large groups
Demand Characteristics?
when people assume the purpose and change their attitude
expectancy effect, screw you effect, cofounding variable, social desirability effect, placebo effect
Construct validity?
investigating if the measure is really measuring what it says it is measuring
D in carDud?
Debriefing
Natural?
no IV manipulated, occurs in natural setting, IV changed by natural occurance in enviroment
Purposive?
looking for people with specific traits, self-selected/volunteer sample
Snowball is a type of purposive, people get more people
Extraneous variables?
variables that influence the result of the study
in/external validity?
internal- relationship that you are testing are not being influenced by anything else
External- extent that the results can be generalized to other people, ties into ecological validity can the results be generalized beyond this situation, artificial or not
R in caRdud?
Right to withdraw
Evaluate?
Testable
Evidence
Application
Constructive variables
Unbiased
Predictable
Stratified?
tries to make a sample of sub-groups of population, random select from start
Matched Pairs design?
sample design where people are not randomly allocated, pretested with regard to the variable, guarantees that each group has a range of abilities
population validity?
sample is not representative of the population