Term
|
Definition
Longitudinal Designs
Multiple-regression Designs
The Pattern and Parsimony Approach |
|
|
Term
|
Definition
This can provide evidence for temporal precedence by measuring the same variables in the same people at several points in time.
Used in developmental psychology.
Also adapted to causal claims. |
|
|
Term
| Cross-sectional Correlations |
|
Definition
Are the two variables measured at the same point in time correlated?
TV Violence 3rd Grade TV Violence 13 grade
I I
I r=.21 I r=.05
V V
Aggression Aggression |
|
|
Term
|
Definition
The association of each variable with ITSELF over time.
TV Violence 3rd Grade ------> TV Violence 13 Grade
r=.05
Agression 3rd Grade -------> Aggression 13 Grade
r=.38 |
|
|
Term
|
Definition
Show whether the earlier measure of one variable is associated with the later measure of the other variable.
The stronger path can help determine which came first.
TV Violence 3rd Grade -----> Agression 13 Grade
r=.31
Aggression 3rd Grade--------> TV Violence 13 Grade
r=.01 |
|
|
Term
| 2 Reasons for Why Not Just Do an Experiment? |
|
Definition
Ethics- unethical to assign people to certain conditions.
EX- Smoking
Practicality- Can't randomly assign preferences.
EX- Liking Violent TV |
|
|
Term
| Correlational studies and the third variable problem.... |
|
Definition
| Measuring other variables controls for their impact. |
|
|
Term
| Multiple-regression Analysis |
|
Definition
Can help rule out some third variables.
Can help address questions of internal validity. |
|
|
Term
| When studying 3 or more variables....... |
|
Definition
Choose a dependent variable (or Criterion Variable), it is the one we are interested in understanding or predicting.
The rest of the variables are called predictor variables. |
|
|
Term
|
Definition
One beta value for each predictor variable.
Similar to r
*Examining the beta weight helps us get a better understanding of which factors most strongly effect the variable we are trying to explain. |
|
|
Term
| Regression in Popular Press |
|
Definition
"Controlled for"
"Taking into account"
"Correcting for" |
|
|
Term
| True or False: Regression establishes causation. |
|
Definition
|
|
Term
| True or False: Regression cannot establish Temporal Precedence. |
|
Definition
|
|
Term
| True or False: Regression cannot control for unmeasured variables. |
|
Definition
|
|
Term
|
Definition
The best explanation that requires the fewest exceptions or qualification.
Assembling converging evidence.
Harder to raise the third variable question.
EX- Smoking causes Cancer. |
|
|
Term
|
Definition
| Variables that EXPLAIN the RELATIONSHIP between 2 other variables. |
|
|
Term
| Mediators vs Third Variables |
|
Definition
| The third variable is related but does not explain the relationship between the 2 variables. |
|
|
Term
|
Definition
| They CHANGE the relationship between 2 variables. |
|
|
