Lecture 4

Question 1

How does elaboration work?

Answer 1

1. First, think about how the variables Y, X1 and control variable X2 are
   theoretically associated with each other (make conceptual model)
2. Next, estimate the uncontrolled association between X1 and Y
   This yield a measure of association, e.g. r 1Y
3. Then, keep the association between X1 and Y constant for the influence
   of X2 (control for X2)
4. Finally, inspect whether the (controlled) association between r 1Y has
   changed
5. Depending on the nature of the change in association and keeping in
   mind your theoretical model (step) we decide that there is a particular
   type of association between the variables

Question 2

‘Keeping constant’, how can you do that?

Question 3

‘Keeping constant’, how can you do that?

Question 4

2 ways to control an association

Question 5

Notation

Question 6

How can partial r be computed?

Question 7

Change of r coefficients and types of association

Answer 7

a) No linear association, before and after
controlling
b) Linear association does not change after
controlling (replication)
c) Linear association disappears completely
after controlling (either ‘full mediation’
or ‘full spuriousness’)
d) Linear association gets weakerafter
controlling, but is still significant (either
‘partial mediation’or ‘partial
spuriousness’)
e) Controlled relationship changes direction
or is larger than uncontrolled
relationship (suppression)

Question 8

Full spuriosity

Question 9

3 different models, same empirical result

Question 10

Important observation

Answer 10

“This empirical outcome by itself does not indicate which one among these
several different models is correct. Theoretical and commonsense
understanding of how variables might be related is required to choose
among these three models.”

Question 11

Full and partial mediation

Question 12

Suppression

Question 13

Interpretation of association