UC Irvine
UC Irvine Previously Published Works
Title
Bringing Some Clarity to Role Ambiguity Research
Permalink
https://escholarship.org/uc/item/7k9934ht
Journal
Academy of Management Review, 6(4)
ISSN
0363-7425
Author
Pearce, Jone L
Publication Date
1981-10-01
DOI
10.5465/amr.1981.4285727
Copyright Information
This work is made available under the terms of a Creative Commons Attribution License,
availalbe at https://creativecommons.org/licenses/by/4.0/
Peer reviewed
eScholarship.org Powered by the California Digital Library
University of California
Bringing Sotne Clarity
to
Role Ambiguity esearch
1
JONE
L.
PEARCE
University of California - Irvine
Theorefical
de11e/opmenf
of
tlie
concept
of
role
ambiguity
and
empiricnl
research
on
this
concept
have
proceeded
fairly
independently
of
one
rwofher.
Empiricnl
work
/ms
confounded
role
ambiguity
wif/1
both
job
dissatisfaclion
arid
formnlizafiot1.
A
new
model
of
a,mbiguify
is
offered,
emphasizing
unpredicfnbility,
rnf
her
th911
information
deficiency,
This
model
accords
wifh
expectancy
theory,
Role
ambiguity
has
been
receiving
increased
attention
from
organizational
behavior
researchers.
· There
is
·increased use
of
role
ambiguity
and
its
· frequent
companion,
role conflict, as
intervening
variables
between
the
structural
characteristics
of
organizat1ons
and
a
variety
of
individual
behavioral
and
affective
outcomes.
Unfortunately,
there
is
insufficient
coherence
to
this
growing
body
of
research.
As
role
ambiguity
was
introduced
to
organizational
researchers
by
Kahn,
Wolfe,
Quinn,
Snoek,
and
Rosenthal
f1964l,
it
was
conceptually
well
developed
and
offered
a
multitude
of
testable
hypotheses;
yet
only
a
small
number
of
these
hypotheses
have
been
directly
tested.
Instead,
empirical
research
on
roie
ambi-
guity
has developed
in
several
directions,
focusing
·
on
relationships
between
a
multitude
of
variables.
Throughout
its
use
in
organizational
research,
role
ambiguity
has
usually
been examined
along-
side
other
role
concepts-most
notably
role
con-
flict,
However,
the
very
success
of
role conflict
has
1
tended
to
over.shadow
role
ambiguity.
In
order
to
focus
attention
more
fully
on
role
ambiguity,
I
will
analyze
it in isolation
from
other
role
concepts.
My
discussion will also
be
shaped
by
the fact
that
theoretical
development
of
the
concept
of
role
ambiguity
and
empirical
research
related
to
this
concept
have
proceeded
fairly
independently
of
one
1
Special
thanks
to
Jimmy
Dean for his critlques of earlier drafts
of
this
article,
· r J
98
/
1,y
1/11•
l\crrrlrmy
of
Mnirngfmrul 0,16.3- 7425
----------·
-
665
another,
and
by
the
fact
that
researchers
have
rarely
used
their
findings
to
elaborate
and
test
the
original
conceptualization
of
Kahn
et
al. Yet
an
understanding
of
both
the
theory
of
role
ambiguity
and its empirical
support
is
necessary
to
a
synthesis
of
our
current
knowledge
of
the
concept.
There-
fore,
this
review
begins
with
a
summary
of
the
theory
of
role
ambiguity
as
set
forth by
Kahn
et
al.,
and
concludes
with
a
discussion
of
the
empirical
research
on
role
ambiguity.
These
two
topics are
brought
together
in
the
subsequent
analysis
and
syn
thesis,
in
the
hope
of
providing
dearer
direction
for
future
research.
Role Ambiguity:
Theoretical Development
Kahn
et
al.
visualized
role
ambiguity
and
rol~
conflict as
intervening
variables
between
the
struc-
tural
characteristics
of
an
individual's
organi-
zational position and
personal,
behavioral,
and
affective
consequences,
Their discussion of role
ambiguity
states
that
"the
person
must
be
able
to
anticipate
with
fair accuracy
the
consequences of
his
own
actions.,..
He needs to
have
useable
knowledge
about
means-ends
connections
in
situa-
tions
where
he
can
produce
or
withhold
the
means"
[p,
72]. They
use
the
term
ambiguity
to
refer
to
the
relative
unpredictability
of
the
outcomes
of
an
individual's
behavior,
a
usage
similar
to
Goffman's
[1963}
unanchored
infemction
and
Seligman's
l19751
1111predicf11hility.
Kahn
et
al.
link role ambiguity with
other
con-
cepts. Briefly, they hypothesize
that
certain organi-
zational positions
or
jobs
will
be
characterized by
greater
role
ambiguUy
and
conflict-those in which
t.he
incumbents
must
(1) cross boundaries, (2) pro-
duce innovative solutions to nonroutine problems,
and (3) be responsible for the work of
others.
Kahn
et
al. expect the consequences of expe-
rienced role ambiguity to be
greater
tension, job
dissatisfaction, a sense
of
futility, and lower self-
confidence.
The
relationship between experienced
role ambiguity and affective outcomes
is
expected
to
be influenced by
an
individual's "need for clar-
ity."
That
is, individuals· experiencing role ambi-
guity
who
have a low need for clarity will
not
feel
the
aversive outcomes as powerfully as will those
who
have a
greater
need for clarity,
The
original
work
by Kahn
et
al.
has been
expanded by Rizzo, House, and Lirtzman 119701,
who
further
developed
the
definition of role ambi-
guity. In addition to
the
"unpredictability"
of
be-
havioral outcomes, Rizzo
et
al.
added a second com-
ponent
to their definition: "[a lack of)
the
existence
or
clarity
of
behavioral requirements,
often
in
terms
of
inputs
from
the
environment
which
would serve
to
guide behavior
1
and provide knowl-
edge
that
the behavior
is
appropria
te
1
'
[pp.
155-156}.
This
alteration
is
especially important because the
scale they developed is
the
role ambiguity opera-
tionalization in most empirical research. Because
these
two
components
of
the concept are central to
the
subsequent
analysis,
the
former
will be called
"unpredictability" and
the
latter
"information
deficiency." ·
Role Ambiguity: Empirical Research
Empirical research
on
role ambiguity has ad-
vanced in
many
directions;
the
complexity of
the
results
is
compounded by the fact
that
different
studies examining a particular link
between
·role
ambiguity and
another
variable frequently produce
contradictory results. Because a detailed examina-
tion of
the
many nuances of this body
of
empirical
work
is
beyond
the
scope of my discussion, mixed
results will merely
be
identified. (See Van Sell,
Brief, and Schuler
[in
press] for a more comprehen-
sive review
of
the
empirical studies.) Role ambi-
guity has been linked
with
a
great
number
of
differ-
666
ent
variables; this review will
therefore
be
organ.
ized
by
studies
of
antecedents, consequence::,,
and
mediator analyses.
Structural Antecedents
of
Role Ambiguity
Although the original hypotheses of Kahn
et.al.
concerning the
structural
antecedents
of
role
ambiguity and role conflict have
not
been tested,
several researchers have measured
the
associations
between
other
structural
antecedents and
role
ambiguity, allowing
the
data,
rather
than
previous
theory development, to indicate
the
antecedents,
Rizzo
et
al. found moderate associations between
such structural variables as "adequacy of communi-
cation" and "organizational emphasis on personal
development,"
no
significant relationships
with
several
other
structural
variables,
but
a strong
association between role ambiguity and "formaliza-
tion" and "goal consensus and clarity." House
and
Rizzo [1972] found a negative relationship between
"task-oriented leadership" and "formal practices"
and role ambiguity. Rogers and Molnar
[1976]
found a moderate correlation between role
ambi-
guity and
an
index
of
"formalization,"
mixed
results for "intra-organizational contacts"
and
"administrative perceptions," and little
or
no
asso-
ciations
with
other
variables, Morris, Steers,
and
Koch [1978] found moderate correlations between
role ambig~ity and "formalization" and
"partici-
pation"
but
none for "supervisory span," '
1
work
group
size," and "functional independence."
Two
conclusions can be
drawn
about
the
antece-
dents
of role ambiguity. First, the'se findings
reveal
a consistent result: in each study, formalization
is
negatively associated with role ambiguity.
Second,
most
correlation coefficients for
structural
vari•
ables
and
role ambiguity are not significant.
Since
all of these researchers
measured
many antece•
dents
(from six
for
Morris
et
al.
to
dozens
for
Rizzo
I
et
al. and
for
Rogers and Molnar), it could be argued
1
that
the
few significant coefficients could
be
expected
by
c~ance alone.
We
can conclude
that
there
is only weak evidence
that
certain
organiza-1
tional
structural
characteristics lead individuals
to
experience role ambiguity. ·
Although we can have little confidence in
the
scattering
of
other
significant associations,
the
consistent negative association between role
ambi·
guit-y
and
formalization
cannot
be
ignored.
Yet
this
association
should
not
be surprising. In
each
of
the
above
studies,
the role ambiguity scale developed
by
Rizzo
et
aL
was
used,
and
with
scale
items
such
as
"I
know
what
my
responsibilities
are"
and
"clear
planned goals
and
objectives
for
my
job,"
its
~trong
correlation
with
"the
extent
to
which
procedures
and
rules
are
written"
-formalization-might
be
assumed,
This
redundancy
will
be
fully
explored
in
the
analysis.
Individual
Consequences
of
Role
Ambiguity
The
hypotheses
offered
by
Kahn
et
al.
concern-
ing
the
affective
consequences
of
experienced
role
ambiguity
have
been
extensively:
tested.
There
is
a
wealth
of
support
for
the
association
between
role
ambiguity
and
stress
{Brief & Aldag,
1976;
Caplan
& Jones,
1975;
Hamner
&
Tosi,
1974;
Kahn
et
al.,
1964; Lyons, 1971; Miles,
1975,
1976;
Rizzo
et
al.,
1970];
only
Tosi
(1976]
found
a
nonsignificant
rela-
tionship.
That
experienced
role ambiguity leads an
individual
to
feel
stress,
anxiety,
or:
tension
receives
additional
support
from
research
on
the
physiologi-
cal
effects
of
unpredictability
[Seligman, 1975]
and
anecdotal
support
from
Goffman [1963].
The
researchers
who
have
examined
the
rela-
tlonship
between
role
ambiguity
and
self-confi-
dence
have
found
evidence
of a
negative
association
[Beehr,
1976;
Kahn
et
al., 1964],
Only
Kahn
et
al.
provide
data
supporting
the
hypothesis
that
expe-
rienced role
ambiguity
leads
to
lowered
self-
confidence
and
a
sense
of
futility.
Therefore,
we
have
little
substantial
empirical evidence
of
an·
association
between
experienced
role
ambiguity
and
lowered
self-confidence
and
a
sense
of
futility.
In
addition,
several
researchers
have tested
the
relationship
between
experienced
role
ambiguity
and
lower
job
performance-?
behavioral
outcome.
Unfortunately,
the
support
for
this
hypothesis
is
mixed,
Schuler
[1975] and Szilagyi, Sims,
and
Keller [1976)
found
a
significant
negative
associa-
tion be~ween role
ambiguity
and
performance,
but
Schriesheim
and
Murphy
[1976],
Schuler,
Aldag,
and
Brief
I 1977],
and
Schuler
[1977]
found
no
rela-
tionship.
Because
these
tests
were
not
based
on
the
original
conceptualization
of Kahn
et
al.,
and
because
these
authors
do
not
provide
extended
dis-
cussion
of
the
reasons
why
we
should
expect
an
667
association,
these
mixed
results
are
difficult
to
interpret
theoretically.
The
weak
connection
b~tween
the
theoretical
conceptualization
of
role
ambiguity
and
empirical
research
using
a role
ambiguity
measure
leads
to
several
problems
and
merits
discussion.
This
lit
era-
.
ture
aptly
fits
Torgerson's
characterization
of
the
social
and
behavioral
sciences:
"We
have
a
wealth
of
observables
and
certainly
no
lack
of
constructs.
There
is,
however,
a
rather
serious
shortage
of
important
connections"
[1967, p. 5).
The
Kahn
et
al.
model
of
role
ambiguity
has
not
been
tested
in its
entirety,
yet
we
have
many
studies
reporting
the
assoc;iation
between
one
particular
operational
definition
of
ambiguity
and
a
multitude
of
vari-
..
ables.
Thus,
in
Torgerson's
terms,
the
Kahn
et
al
conceptualization
is
"immune
from
rejection"
on
the
basis
of
empirical evidence,
while
atheoretical
studies
accumulate.
The
relationship
between
role
ambiguity
and
job
satisfaction
has
generated
the
most
research.
The
research
evidence
tends·
to
indicate
a
moderate
negative
association
between
role
ambiguity
and
job
satisfaction
[Beehr,
1976;
Beehr,
Walsh, &
Taber,
1976;
Greene
&
Organ,
1973;
Hamner
&
Tosi,
1974; Ivancevich_&
Donnelly,
19.74;
Kahn
et
al.,
1964;
Keller,
1975;
Lyons,
1971;
Miles, 1976;
Rizzo
et
al:
1
1970;
Schriesheim
&
Murphy,
1976;
Schuler,
1975;
Schuler
et
al., 1977; Szilagyi
et
al.,
1976;
Valenzi
&
Dessler,
1978].
Contradictory
evi-
dence
comes
from
Tosi
[1976]
and
Brie.f
and
Alda-g
[1976).
However,
as
with
formalization,
there
is a
redundancy
in
the
measurement
of
the
two
vari-
ables
of
role
ambiguity
and
job dissatisfaction,
The
argument
to
be
developed
in
the
following analysis
is
that,
as
they
have
been
measured,
role
ambiguity
and
job
dissatisfaction
are
indistinguishable.
Several
conclusions
about
the
consequences
of
experiencing
role
ambiguity
are
·warranted.
There
is
strong
support
for
the
hypothesis
that
role
cJmbi-
guity
causes
an
individual
to
experience
stress,
and
weaker
support
for
the
hypothesis
that
those
who
experience
role
ambiguity
will
have
less
confidence
in
their
ability
to
influence
their
environments
(a
sense
of
futility).
No
definiHve
conclusions
can be
drawn
about
the
effect
of
role
ambiguity
on
job
performance,
since
the
results
are
mixed, Finally,
there
is
substantial
evidence
that
rote
ambiguity
is
associated
with
job
dissatisfaction,
a
relationship
that
will be
analyzed
more
fully
with
respect
to
how
these
variables
have
been
measured.
Role Ambiguity in Mediator Analyses
Turning
to
the
various
studies
in
which
mediator
variables
have
been
examined,
we
begin
with
those
studies
in
which
role
ambiguity
has
been
examined
as a
mediator
of
other
relationships,
followed by
the
studies
concerned
with
mediators
of
the
rela-
tionships
between
role
ambiguity
and affective
consequences.
In his pa th/goal
theory,
House
[1971}
hypothe-
sized
that
the
responses
of
subordinates
to
leader
"initiating
structure"
behaviors
varied
because
role
ambiguity
acted
as
a
mediator:
for
subordinates
with
unambiguous
tasks,
leader
structuring
be-
havior
would
be
redundant
and
!~ad
to
less job
satisfaction,
whereas
for
subordinates
experienc-
ing
role
ambiguity,
the
structuring
behavior
would
be
valued
and
lead
to
greater
satisfaction.
House
provides
data
supporting
this
hypothesis,
but
Schri~sheim
and
Murphy
[1976]
found
no
signffi-
can
t media ting.
effects.
Both
House
and
Rizzo
[1972)
and
Valenzi
and
Dessler
[1978]
found
evi-
dence
of
a
significant
mediating
effect
only
for
leader
"consideration"
behavior,
not
for
structur-
ing
behavior.
Valenzi and
Dessler
disagree with the
path/goal
hypothesis,
since
the
consideration/satis-
faction
relationship
has
more
empirical
support.
There
is,
then,
at
best
mixed
support
for
the
path/
goal
hypothesis;
although
role
ambiguity
does
con-
sistently
mediate
the
consideration/satisfaction
relationship,
no
comprehensive
theoretical
expla-
nation
comparable
to
House's
path/goal
theory
has
been
offered,
Many
researchers
have
sought
mediators
that
would
increase
the
strength
of
the
demonstrated
association
between
role
ambiguity
and
stress/
dissatisfaction. A variety
of
variab_les
have
been
tried, using
the
inductive
approach, and
the
re~ults
have been mixed, so few firm
conclusions
can be
drawn.
On
the
whole,
organizational
level does
not
consistently mediate relationships
between
role
ambiguity
and
other
variables [Ivancevich &
Don-
nelly, 1974; Miles, 1976;
Schuler,
1975, 1977; Szi-
lagyi
et
aL
1976].
Need
for
achievement
[Johnson
&
Stinson,
1975;
contradictory
results
from
Morris
&
Snyder,
19791, need
for
clarity
[Ivancevich &
668
Donnel1y,
1974;
Kahn
et
al.,
1964;
Lyons, 1971;
Miles &
Petty,
1976],
and
group
cohesiveness
[Beehr,
1976]
do
mediate
most
of
the
role
ambigui-
ty/affective
consequences
relationships,
bot
the
additional
variance
explained
is
not
substantially
greater
than
that
explained
in
other
studies
omit•
ting
mediator
variables.
Because
chance
findings
may
be
interpreted
as
real
when
mixed
results
are
obtained,
the
empirical
support
for
most
hypothesized
relationships
ia
weak;
only
three
relationships
have
strong
sup~
port.
The
first-greater
role
ambiguity
increases
stress-is
the
only
one
with
solid
support.
The
other
two-formalization
decreases
role
ambi~
guity,
and
role
ambiguity
ere
ates
job dissatisfac-
tion-found
correlational
support;
yet
a careful
examination
of
the
conceptual
and
methodological
bases
of
this
research
will reveal a
confounding
of
the
three
variables
of
role
ambiguity,
formaliza-
tion,
and' job
dissatisfaction,
We
now
turn
to
this
·
examination.
Confounding with
Job
Dissatisfaction ·
Role
ambiguity
and
job
dissatisfaction,
as
they.
are
currently
conceptualized
and
measured,
are
confounded.
The
problems
begin
with
the
concep•
tualization
of
role
ambiguity.
Role
ambiguity
has
been
defined
as
a
dual-component
variable. Rizzo
et
al.
include
bqth
the
predictability
and
information-
deficiency
components
explicitly,
but
Kahn
et
aL
do
so
only
implicitly;
they
use
these
components
inter-
changeably,
as
can
be
seen
by
their
use
of
deficiency-of-information
examples
of
role
ambl~
g~ity
[pp.
73-74].
The
information-deficiency
component
leads
to
conceptual
difficulties.
No
doubt
organizational
positions
differ
in
the
degree
to
which
they
are
explicitly detailed;
the
availability
of
information
is
not,
however,
synonymous
with
the
predictabi1ity
of
the
consequences
of
behavior.
Information
is
often
made
availab)e
for
reasons
other
than
clarify•
in'g
the
consequ!,!nces
of
individual
action~e.g.,
job
descriptions
can' be
written
solely
to
present
acer•
tain
image
to
outsiders,
to
ma~<e
a case
for
a
pay
raise,
and
so
forth.
We
can
imagine
situations
irt
which
information
availability is
in
the
service
of
enhanced
predictability,
yet
also
imagine
situations
in
which
the
proliferation
of
documents
is asso·