Partial Med - KolobKreations

Report
Mediation Step by Step
With Bootstrapping
Here is the model we’ll be testing
Step 1: Create Table
Hypothesis
Partial Med
MU-Burn-Satw
Partial Med PartBurn-Satw
Direct Beta w/o
Med
Direct Beta
w/Med
Indirect Beta
Mediation type
observed
Step 2: Direct Effects Significant?
Regression Weights: (Group number 1 - Default model)
Estimate
S.E.
C.R.
P
satw
<---
part
.130
.050
2.594
.009
satw
<---
mu
-.525
.040
-13.102
***
Step 3: Update Table
Hypothesis
Partial Med
MU-Burn-Satw
Partial Med PartBurn-Satw
Direct Beta w/o
Med
Direct Beta
w/Med
Indirect Beta
Mediation type
observed
-.60***
.12**
Estimate
S.E.
C.R.
P
satw
<---
part
.130
.050
2.594
.009
satw
<---
mu
-.525
.040
-13.102
***
Step 4: Set up Bootstrap
Step 5: Run with Mediator
If it throws you an error when you run it with
bootstrapping, just run it again, and again, and again,
until it runs without an error. If this still doesn’t work,
then close and reopen AMOS and try again.
Step 5a: Direct Effects w/Med
Hypothesis
Partial Med
MU-Burn-Satw
Partial Med PartBurn-Satw
Click here first
Then here
Direct Beta w/o
Med
Direct Beta
w/Med
-.60***
-.346
.12**
.052
Indirect Beta
Mediation type
observed
Step 5b: Direct Effects w/Med
Hypothesis
Partial Med
MU-Burn-Satw
Partial Med PartBurn-Satw
Direct Beta w/o
Med
Direct Beta
w/Med
-.60***
-.346**
.12**
.052 (ns)
Indirect Beta
Mediation type
observed
Click here first
**
Then here
Not
significant
(ns)
Step 6a: Indirect Effects
Hypothesis
Partial Med
MU-Burn-Satw
Partial Med PartBurn-Satw
Click here first
Then here
Direct Beta w/o
Med
Direct Beta
w/Med
Indirect Beta
-.60***
-.346**
-.253
.12**
.052 (ns)
.067
Mediation type
observed
Step 6b: Indirect Effects
Hypothesis
Partial Med
MU-Burn-Satw
Partial Med PartBurn-Satw
Click here first
Then here
Direct Beta w/o
Med
Direct Beta
w/Med
Indirect Beta
-.60***
-.346**
-.253**
.12**
.052 (ns)
.067**
Mediation type
observed
Step 7: Mediation?
Hypothesis
Partial Med
MU-Burn-Satw
Partial Med PartBurn-Satw
Direct Beta w/o
Med
Direct Beta
w/Med
Indirect Beta
Mediation type
observed
-.60***
-.346**
-.253**
Partial
.12**
.052 (ns)
.067**
Full
No Mediation
• If Indirect is not significant
• Also if direct effect of IV on Med is insignificant or if direct effect
from Med to Y is insignificant
Indirect Effects
• Both direct effects are not significant, but Indirect effect is significant
Full Mediation
• Given the direct effects were significant prior to adding the mediator
• If Indirect is significant and Direct (with med) is not significant
Partial Mediation
• If Direct (with med) & Indirect are significant

similar documents