Conducting experiments Flashcards
Diode population into subgroups (strata) and randomly sample from each strata
- ensures small groups are included
Stratified sampling
Every member of a population has equal chance of being sampled
Simple random sampling
First, randomly determine number of provinces, then health boards, then sample
Cluster sampling
Get a sample that meets some pre-defined purpose
Purposive sampling
Measure manipulated by experimenter
Independent variable
Like stratified sampling, but using convenience procedures. (Strata + convenience)
Quota sampling
Follows instruction of the experimenter, not a real subject - create a particular state
Confederate
The extent to which the two groups differ (in a standard two group design)
Manipulation strength
What the experimenter is measuring
Dependent variable
Percent correct (type of measure)
Accuracy measure
Response time.
Response duration.
(Type of measure)
Latency measures
When performance on the DV is so low that is cannot go lower
Floor effect
When the performance on the DV is so high that it cannot go higher
Ceiling effect
Any characteristic that might inform the subject about the purpose of the study
Subjects tend to (consciously and unconsciously) change behavior to confirm hypothesis
Demand characteristics
Experimenter can unintentionally bias the subject to behave in certain ways
Experimenter bias