Causal Relationship
One variable directly or indirectly causes changes in another.
Hope to show this between IVs and DVs.
Unidirectional
Type of causal relationship.
Changes in A cause changes in B, but not vice versa.
Bidirectional
Type of causal relationship.
Changes in A cause changes, but changes in B cause changes in A.
Correlational Relationship
Changes in one variable accompany changes in another.
Independent Variable
Active; Manipulated with at least 2 levels.
Value is determined by researcher.
Dependent Variable
Measured.
Variable whose value you observe and measure in an experimental design.
Value is determined by subject's behavior.
Experimental
Group where subjects receive treatment.
Control
Group where subjects do not receive experimental treatment.
Serves as baseline of behavior.
Attribute Variables
Naturally occur.
We have no control over manipulating these in a study.
Ex. Pre-existing characteristics.
Incomplete/Lost Data
Can compromise the effect of random assignment and decrease the power of a study.
Steps to reduce:
-Specific eligibility criteria
-Thorough consent process
-Ongoing support during study
On-Protocol/Completer Analysis
Involves sticking with the participants that actually completed the study/adhered to the protocol.
Intention to Treat Analysis
Data analyzed according to the original random assignment regardless of the treatment subject actually received.
Includes all subjects and reduces bias of dropouts regardless of group.
More reflective of actual practice where we typically see some degree
Blinding
Protection against the influence of participants knowledge of their treatment status or researchers expectations on performance or recording/reporting of results.
Blocking
Systematically manipulate attribute variables by building them into the experimental design as an IV.
Matching
Match subjects on specific characteristics across groups.
Limits generalizability.
ANCOVA
Statistical technique to control for inter-subject differences; helps to eliminate influence of extraneous factors on IV by conceptually removing confounding effects by making it artificially equivalent across groups and basing analysis of differences on