Chapter 9: Multivariate Correlational Research

Question 1

Multivariate Designs

Answer 1

Involve more than two measured variables

Question 2

Multiple Regression Analysis

Answer 2

Can control for some internal validity/third variable problems
Statistical technique
Use BETA to test for third variables
Criterion variable (dependent variable) typically specifies in either top row or title of regression table
Can evaluate whether a relationship between two key variables still holds when they control for another variable (identify subgroups)

Question 3

Longitudinal Designs

Answer 3

Measuring the same variables in the same people at several points in time
Can satisfy temporal precedence
E.g. used in developmental psychology to study the changes in a trait or ability over time

Question 4

Pattern & Parsimony Approach

Answer 4

Pattern of results that are best explained by a single, parsimonious causal theory
Specify a mechanisms for the causal path
All diverse predictions about one phenomenon are tied back to one central principle
E.g. example with cigarettes causing cancer

Question 5

Ways To Interpret Results From Longitudinal Designs

Answer 5

Cross Sectional Correlations
Autocorrelations
Cross Lag Correlations

Question 6

Cross Sectional Correlations

Answer 6

Test to see whether two variables, measured at the same point in time, are correlated
Does not establish temporal precedence

Question 7

Autocorrelations

Answer 7

Evaluate correlation of each variable with itself over time

Question 8

Cross Lag Correlations

Answer 8

Show if the earlier measure of one variable is associated with the later measure of the other variable
Addresses directionality problem and can help establish temporal precedence

Question 9

BETA Basics

Answer 9

Beta value for each predictor variable (similar to correlation coefficient)
When p value is less than 0.05 is not statistically significant
Measure the correlation between different variables while controlling for others
Language: controlled for, considering, adjusting for, etc
Drawback: can’t control for variables that aren’t thought of/measured

Question 10

Mediator/Mediating Variable

Answer 10

Is variable of direct interest to researcher
Linking variable between two others
E.g. following doctors orders links good conscience and health
Can be tested with multiple regressions

Question 11

Third Variable

Answer 11

Proposed third variable is external to the two variables in the original correlation
Often seen as accident or nuisance
E.g. each variable is separately associated with a third variable

Question 12

Moderator

Answer 12

Association between two variables differs depending on level of third moderator variable
Makes association stronger/weaker, or positive/negative
E.g. social support is moderator between stress and depression