Bivariate + Multivariate Correlational Research

Question 1

Multivariate Design

Answer 1

A study design to test an association involoving more than two measured variables

Question 2

Moderators

Answer 2

A variable that, depending on its level, affects the relationships b/w 2 variables . In correlational research, moderators tell us about external validity
(EX: is compulsive texting associated w/ academic performance - Moderators can tell us: when, for whom, or under what conditions two variables are related)

Question 3

Why does correlation NOT imply causation?

Answer 3

There are 3 criteria for a causal claim: Covariation, temporal precedence, internal validity

Question 4

Temporal Precedence (or directionality problem)

Answer 4

Did: A lead to B? or did B lead to A? (if we cannot tell which came first, we cannot infer causation

Question 5

What do we need to determine cause?

Answer 5

Requires an experiment (manipulated variable while holding all others constant)
While we can determine cause using simple bivariate correlation techniques, we can build a case for cause using the multivariate techniques

Question 6

What are the “fancy” multivariate correlational techniques

Answer 6

Pattern and Parsimony approach AND longitudinal research design

Question 6

Pattern and Parsimony

Answer 6

Research can investiage causality by using a variety of correlational studies that point in the same causal direction
Scientists will settle on the explanation that best fits, and most parsimoniously, explains the ENTIRE pattern of data across several studies

Question 7

Pattern

Answer 7

Look at the pattern of results across a variety of correlational studies that all point in a single, causal direction

Question 7

Parsimony

Answer 7

We aim for the simplest explanation of a pattern of data (we want the theory that requires making the fewest exceptions)