A critical evaluation of alternative methods and paradigms for conducting mediation analysis in operations management research

• DOI: http://doi.org/10.1016/j.jom.2014.01.003
• Published date: 01 May 2014
• Author: Manoj K. Malhotra,Guangzhi Shang,Cherry Singhal,Robert E. Ployhart
• Date: 01 May 2014

Journal

of

Operations

Management

32

(2014)

127–137

Contents

lists

available

at

ScienceDirect

Journal

of

Operations

Management

j

o

ur

na

l

ho

mepage:

www.elsevier.com/locate

/jom

Technical

note

A

critical

evaluation

of

alternative

methods

and

paradigms

for

conducting

mediation

analysis

in

operations

management

research

Manoj

K.

Malhotraa,∗,

Cherry

Singhalb,1,

Guangzhi

Shangb,2,

Robert

E.

Ployhartc,3

aJeff

B.

Bates

Professor

and

Chair,

Department

of

Management

Science,

Moore

School

of

Business,

University

of

South

Carolina,

Columbia,

SC

29208,

United

States

bDepartment

of

Management

Science,

Moore

School

of

Business,

University

of

South

Carolina,

Columbia,

SC

29208,

United

States

cBank

of

America

Professor

of

Business

Administration,

Department

of

Management,

Moore

School

of

Business,

University

of

South

Carolina,

Columbia,

SC

29208,

United

States

a

r

t

i

c

l

e

i

n

f

o

Article

history:

Received

20

May

2013

Accepted

8

January

2014

Available

online

5

February

2014

Keywords:

Mediation

models

Methodological

review

and

recommendation

Empirical

a

b

s

t

r

a

c

t

Mediation

as

a

theory

testing

approach

has

witnessed

considerable

adoption

among

Operations

Man-

agement

(OM)

researchers.

Although

mediation-testing

methods

have

evolved

tremendously

in

the

past

decade,

their

dissemination

in

the

OM

field

has

not

seen

parallel

growth.

These

advanced

techniques

facilitate

the

testing

of

existing

and

complex

hypotheses

in

a

more

precise

manner.

With

the

intent

of

critically

evaluating

existing

and

alternative

methods

for

conducting

mediation

analysis

needed

to

sup-

port

sophisticated

empirical

research,

this

paper

first

reviews

OM

studies

that

tested

for

mediation

in

the

past

eleven

years

(2002–2012)

from

top-tier

OM

journals.

Four

commonly

used

mediation

approaches

were

identified.

Based

on

principles

of

good

theory

building,

type

of

mediation

model,

and

properties

of

empirical

data,

we

evaluate

the

existing

methodologies

and

make

recommendations

on

how

to

improve

the

rigor

of

OM

mediation

testing.

Using

published

OM

studies

in

top

journals

as

examples,

we

then

illustrate

the

relevance

and

advantages

of

these

recommendations,

as

well

as

their

ease

of

use.

Further-

more,

we

empirically

show

that

more

robust

and

insightful

results

can

be

achieved

by

adopting

these

techniques,

which

in

turn

have

the

promise

of

leading

to

better

theory

building

and

testing

in

the

field

of

operations

management.

©

2014

Elsevier

B.V.

All

rights

reserved.

1.

Introduction

As

a

straightforward

and

powerful

way

to

construct

thoughtful

observations

into

theories,

mediation

has

received

significant

pop-

ularity

among

Operations

Management

(OM)

researchers.

A

review

of

OM

journal

publications

in

the

past

eleven

years

(2002–2012)

shows

that

on

average,

eleven

studies

per

year

employed

a

media-

tion

perspective

in

their

research.

Mediation

research

has

covered

a

wide

range

of

topical

areas

and

theoretical

domains

in

OM.

These

include

service

operations

(Ba

and

Johansson,

2008;

Goldstein,

2003;

Venkatesh

et

al.,

2010),

sourcing

(Bardhan

et

al.,

2007;

Jayaraman

et

al.,

2013;

Narayanan

et

al.,

2011),

supply

chain

man-

agement

(Bendoly

et

al.,

2007),

new

product

development

and

process

innovation

(Bendoly

et

al.,

2012;

Lee

et

al.,

2011),

and

∗Corresponding

author.

Tel.:

+1

803

777

2712;

fax:

+1

803

777

6876.

E-mail

addresses:

malhotra@moore.sc.edu

(M.K.

Malhotra),

cherry.singhal@grad.moore.sc.edu

(C.

Singhal),

guangzhi.shang@grad.moore.sc.edu

(G.

Shang),

ployhart@moore.sc.edu

(R.E.

Ployhart).

1Tel.:

+1

803

479

0636;

fax:

+1

803

777

6876.

2Tel.:

+1

803

678

9398;

fax:

+1

803

777

6876.

3Tel.:

+1

803

777

5903;

fax:

+1

803

777

6876.

retail

operations

(Li

et

al.,

2013),

to

mention

just

a

few

exam-

ples.

Venkatraman

(1989)

defined

mediation

as

the

existence

of

a

significant

intervening

mechanism

between

antecedent

and

con-

sequent

variables,

whereas

Baron

and

Kenny

(1986)

put

it

in

a

slightly

different

way.

They

state

that

mediation

represents

the

generative

mechanism

through

which

a

focal

exogenous

indepen-

dent

variable

is

able

to

influence

its

dependent

consequence.

Most

OM

researchers

have

adopted

one

of

these

two

perspectives.

Psy-

chometric

literature

on

mediation

analysis

has

made

significant

progress

in

the

past

decade

toward

more

robust

approaches

to

mediation

testing

(MacKinnon

et

al.,

2002;

Preacher

and

Hayes,

2004;

Shrout

and

Bolger,

2002;

Zhao

et

al.,

2010).

However,

OM

research

is

not

synchronized

with

these

advancements.

This

is

probably

due

to

the

fact

that

although

multiple

streams

of

methodological

literature

on

mediation

are

available,

there

is

lack

of

a

concise

synthesis

catered

specifically

to

the

needs

of

an

OM

audience.

The

main

purpose

of

this

paper

is

to

first

assess

and

critically

evaluate

the

status

quo

of

mediation

research

in

the

field

of

OM,

and

then

provide

a

set

of

alternative

methods

and

paradigms

for

improving

the

current

practices

of

mediation

testing.

0272-6963/$

–

see

front

matter

©

2014

Elsevier

B.V.

All

rights

reserved.

http://dx.doi.org/10.1016/j.jom.2014.01.003