0% found this document useful (0 votes)
40 views16 pages

Mediation Analysis in SPSS Explained

Uploaded by

vvvsn8pbmh
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPSX, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
40 views16 pages

Mediation Analysis in SPSS Explained

Uploaded by

vvvsn8pbmh
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPSX, PDF, TXT or read online on Scribd

• Exercises

• How to report
the results
Exercise 2
• Take away
messages
The indirect effect:
Simple Mediation

TWQDA Seminar 3 PROCESS.


sav
Mediating effect
HOW/ WHY are IV (X) and DV (Y) related?
To understand the mechanisms by which
an IV causally influences a DV.
• Does putative causal X influences outcome Y through
one or more mediator variables (M)

X Y
Research
Questions
1. Do perceived broken promises lead to employees’ intention to quit?
Direct effect = c1’
2. Can this effect be explained by organizational commitment?
Indirect effect = a1b1
3. What is the effect of the perceived broken promises on employees’
intention to quit, both directly and indirectly through perceptions of the
organizational commitment?
Total effect (c1) = c1’ + a1b1
Organizational
Commitment

Psychological Exit
Contract (Turnover
Breach Intentions)
• When to use it?
• How to do it in
SPSS?
• How to read the
output?
Control
variables
might be
inserted
here
• When to use it?
• How to do it in
SPSS?
• How to read the
output?
Always check
this!
a1

a1  Two employees that differ by one unit on PsyCB are


estimated to differ by a= - 0.17 units on OrgCom. The sign of
a is negative, meaning that those relatively higher in PsyCB
are estimated to be lower in their organizational commitment.
b1  Two employees who experience the same level of PsyCB (i.e., who are
equal on PsyCB) but that differ by one unit in their level of OrgCom are
estimated to differ by b= - 0.96 units in intention to quit. The sign of b is
negative, meaning that those relatively higher in OrgCom are estimated to be
lower in their intention to quit.

b1
c1’ direct effect = Two employees that differ by one unit on PsyCB and
who experience the same level of OrgCom (i.e., who are equally committed
to the organization), are estimated to differ by 0.47 units in their turnover
intentions (to be 0.47 units more prone to quit) .

R² = quantifies the proportion of the total variance of Exit (Y)


explained by the overall model. This solution explains the 36% of
the variance of Exit, which is statistically significant (p < 0.01).

c 1'
c1

c1 total effect = Two employees who differ by one unit in PsyCB are
estimated to differ by 0.63 units in their reported turnover intentions. The
positive sign means the person perceiving greater PsyCB reports higher
intentions to quit.
*Note: This total effect can also be estimated by regressing Y on X alone.
a1 b1

a1b1 indirect effect = Two employees who differ by one unit in their
PsyCB are estimated to differ by 0.16 units in their reported intentions to
quit as a result of the tendency for those whose promises were broken to
feel less committed to the organization, which in turn translates into greater
turnover intentions.
Consequent

OrgCom (M) Exit (Y)

Antecedent Coeff. SE p Coeff. SE p

PsyCB (X) a1 -.168 .031 <.001 c1’ .471 .073 <.001

OrgCom (M) --- --- --- b1 -.965 .147 <.001


• When to use it?
constant i1 3.443 .082 <.001 i2 4.120 .539 <.001
• How to do it in
SPSS? R2 = .114 R2 = .363
• How to read the F(1,224) = 28.960, p<.001 F(2,223) = 63.563, p<.001
output?
• How to report
the results?
Effect SE p LLCI ULCI

Direct effect c1’ .471 .073 <.001 .326 .615

Total effect c1 .633 .075 <.001 .485 .781

Boot SE Boot LLCI Boot ULCI

Indirect effect a1b1 .162 .040 .098 .265


The effect of PsyCB on OrgCom a1 = - 0.17 means that two employees that differ by one
unit on PsyCB are estimated to differ by 0.17 units Consequent on OrgCom. The sign of a 1 is negative,
meaning that those relatively higher in PsyCB are estimated to be lower in their
organizational commitment. This effect is statistically different from zero, t=Exit
OrgCom (M)
-5.381,
(Y)
p
= .000, with a 95% confidence interval from -0.229 to -0.106.
The effect b1 = -0.96 indicates
Antecedent Coeff. that
SE two employees
p who experience Coeff.the same
SE level of p
PsyCB but that differ by one unit in their level of OrgCom are estimated to differ by b 1 = -
PsyCB (X) a1 -.168 .031 <.001 c1’ .471 .073 <.001
0.96 units in intention to quit. The sign of b 1 is negative, meaning that those relatively
higher
OrgCom in (M)
OrgCom are--- estimated---to be lower --- in their intention
b1 to-.965
quit. This.147
effect is <.001
• When to use it? statistically different from zero, t= -6.542, p = .000, with a 95% confidence interval
constant
from .326 i1 3.443
to .615. .082 <.001 i2 4.120 .539 <.001
• How to do it in The indirect effect of 0.162 means that two workers who differ by one unit in their reported
SPSS? PsyCB are estimated to differ byR0.162 2 = .114
units in their reported turnover intentionsR2 = .363
as a result
• How to read the of the tendency for those who=perceive
F(1,224) that their promises have
28.960, p<.001 been broken
F(2,223) = 63.563,top<.001
feel less
output? committed to the organization, which in turn translates into greater turnover. This indirect
• How to report effect is statistically different from zero, as revealed by a 95% BC bootstrap confidence
interval that is entirely above zero (0.098 to 0.265).
the results?
The direct effect of PsyCB, c′ = .471, is theEffect estimatedSEdifference p in turnoverLLCI ULCI
intentions
between two workers experiencing the same level of organizational commitment but who
differ by one unit inDirect
theireffect c1’
reported PsyCB, .471
meaning .073
that <.001
the person .326
feeling more .615
PsyCB but
who is equally committed is estimated
Total effect c1 to be.6330.471 units
.075higher <.001
in his/her reported
.485 turnover
.781
intention. This direct effect is statistically different from zero, t= 6.427, p = .000, with a 95%
confidence interval from 0.226 to 0.614. Boot SE Boot LLCI Boot ULCI
The total effect of PsyCB on turnover is c = .633, meaning two workers who differ by one
unit in PsyCB are Indirect
estimatedeffect a1bby
to differ 1 .162 units in.040
0.633 their reported turnover.098 .265
intentions.
The positive sign means the person perceiving greater PsyCB reports higher intentions to
quit. This effect is statistically different from zero, t = 8.429, p = .000, or between 0.485 and
0.781 with 95% confidence
RECAP: Steps to run Mediation
Analysis
1. Run mediation analysis in Process
2. Check the p-value of the indirect effect: If it is not
significant: the mediation effect does not take place
3. Check the p-value of the direct effect: partial vs full
mediation
4. Comment on all the relevant statistics, including
constant, main effects, and interaction term

You might also like