Identification and Meta-Analytic Assessment of Psychological Constructs Measured in Employment Interviews

Abstract

There has been a growing interest in understanding what constructs are assessed in the employment interview and the properties of those assessments. To address these issues, the authors developed a comprehensive taxonomy of 7 types of constructs that the interview could assess. Analysis of 338 ratings from 47 actual interview studies indicated that basic personality and applied social skills were the most frequently rated constructs in this taxonomy, followed by mental capability and job knowledge and skills. Further analysis suggested that high-and low-structure interviews tend to focus on different constructs. Taking both frequency and validity results into consideration, the findings suggest that at least part of the reason why structured interviews tend to have higher validity is because they focus more on constructs that have a stronger relationship with job performance. Limitations and directions for future research are discussed.

Full text

Journal
of
Applied Psychology
2001, Vol.
86, No. 5,
897-913
Copyright
2001
by the
American Psychological Association, Inc.
0021-9010/01/S5.00
DOI:
10.1037//0021-9010.86.5.897
Identification
and
Meta-Analytic
Assessment
of
Psychological Constructs
Measured
in
Employment Interviews
Allen
I.
Huffcutt
Bradley
University
James
M.
Conway
Central
Connecticut
State
University
Philip
L.
Roth
Clemson University
Nancy
J.
Stone
Creighton
University
There
has
been
a
growing interest
in
understanding what constructs
are
assessed
in the
employment
interview
and the
properties
of
those
assessments.
To
address these issues,
the
authors developed
a
comprehensive taxonomy
of 7
types
of
constructs that
the
interview could
assess.
Analysis
of 338
ratings
from
47
actual interview studies indicated that basic personality
and
applied social skills were
the
most
frequently
rated constructs
in
this taxonomy, followed
by
mental capability
and job
knowledge
and
skills.
Further analysis suggested that high-
and
low-structure interviews tend
to
focus
on
different
constructs.
Taking both frequency
and
validity results into
consideration,
the
findings suggest that
at
least part
of the
reason
why
structured interviews tend
to
have higher validity
is
because they focus more
on
constructs
that
have
a
stronger relationship
with
job
performance. Limitations
and
directions
for
future
research
are
discussed.
Much
of the
employment
interview
research
published
in the
past 10-15 years
has
focused
on
interview validity. There have
been
a
number
of
primary studies
(e.g.,
Campion, Campion,
&
Hudson, 1994; Johnson, 1991; Pulakos
&
Schmitt, 1995; Walters,
Miller,
&
Ree,
1993),
several
meta-analyses
of
interview validity
(e.g.,
Huffcutt
&
Arthur, 1994; McDaniel, Whetzel, Schmidt,
&
Maurer, 1994; Schmidt
&
Rader, 1999; Wiesner
&
Cronshaw,
1988; Wright, Lichtenfels,
&
Pursell, 1989),
and a
meta-analysis
of
interview reliability (Conway,
Jako,
&
Goodman,
1995).
These
works collectively suggest that interviews
can
predict performance
on
the
job.
Although
a
great deal
is now
known about interview reliability
and
validity, much less
is
known about
the
constructs captured
by
employment interviews (Schmidt
&
Rader, 1999).
A
number
of
possible constructs have been suggested
in the
interview literature,
including cognitive ability (Campion, Pursell,
&
Brown, 1988),
motivation
(Ulrich
&
Trumbo,
1965), social skills
(Arvey
&
Cam-
pion, 1982),
and
person-organization
fit
(Harris,
1999).
However,
the
extent
to
which most
of
these constructs
are
actually assessed
in
interviews
is
unclear.
The
only
construct that
has
been investi-
gated empirically
to any
real extent
is
cognitive ability,
and
meta-
analytic
research suggests that,
on
average,
it
represents less than
20% of the
variance
in
interview ratings
(Huffcutt,
Roth,
& Mc-
Daniel,
1996).
Allen
I.
Huffcutt,
Department
of
Psychology,
Bradley
University;
James
M.
Conway, Department
of
Psychology, Central Connecticut State Uni-
versity;
Philip
L.
Roth, Department
of
Management, Clemson University;
Nancy
J.
Stone, Department
of
Psychology, Creighton
University.
Correspondence concerning this article should
be
addressed
to
Allen
I.
Huffcutt,
Department
of
Psychology, Bradley University,
Peoria,
Illinois
61625.
Electronic mail
may be
sent
to
huffcutt@bradley.edu.
The
lack
of
research
on the
constructs
assessed
in
employment
interviews
is not
surprising. Unlike many other psychological
tests, such
as
those
for
personality
and
mental ability, interviews
are not
designed
to
assess
specific constructs (Bobko, Roth,
&
Potosky, 1999; Campion, Palmer,
&
Campion,
1997).
Rather, they
are
individually designed
to
assess
work-related characteristics
for
a
given position,
and
various constructs
may be
embedded
in or
associated with those work-related characteristics. Given
the di-
versity
of
jobs
for
which interviews
are
developed,
it
would
not be
surprising
to see
considerable variation
across
interviews
in
both
the
number
and the
type
of
constructs
assessed
(Klimoski,
1993).
Understanding
the
constructs involved
in
employment inter-
views
is
potentially important.
Significant
advances have been
made
in
interviewing methodology, including
the
situational inter-
view
(Latham,
Saari,
Pursell,
&
Campion, 1980)
and
behavior
description interview (Janz, 1982) formats,
but
many continued
advances
are
likely
to
come
from
analysis
of
what constructs
interviews
measure,
not
from
methodological variants.
In
particu-
lar,
analysis
of
constructs
may
provide
greater
insight into
why
formats
such
as the
situational interview predict performance
and
may
allow interviews
to be
optimally designed
to
achieve specific
outcomes such
as
high incremental validity
and
minimal impact
on
protected groups.
It is our
belief that
four
fundamental issues
(or
steps)
are
involved
in the
important process
of
understanding
the
constructs
captured
by
employment interviews. First,
a
taxonomy
of
possible
constructs that could
be
assessed
in
employment interviews should
be
constructed. Such
a
taxonomy would provide
a
common
and
systematic framework
for
identifying
and
classifying interview
constructs.
Second,
the
constructs that interviewers attempt
to
assess
should
be
identified,
and
information
on the
relative fre-
quency
of
those attempts should
be
compiled. Such
an
analysis
would indicate which constructs
in the
taxonomy
are
actually rated
897
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

898
HUFFCUTT,
CONWAY,
ROTH,
AND
STONE
in
employment interviews and, perhaps more important,
which
constructs
are the
most commonly rated. Third,
the
degree
to
which
the
ratings
for
these constructs reflect
the
intended charac-
teristic
should
be
evaluated. Some ratings
may
represent
a
more
accurate
measurement
of the
intended construct than other ratings,
possibly
due to the
nature
of the
construct itself
(e.g.,
extroversion
vs.
creativity)
or to
differential
influence
from
general factors such
as
mental ability, personality,
job
experience,
or the
type
of
ques-
tions
(see
Huffcutt,
Weekley, Wiesner, DeGroot,
&
Jones,
in
press). Fourth,
the
general properties
of
these ratings should
be
considered, including validity, incremental validity,
and
impact
on
protected groups. Some construct ratings
may be a
stronger pre-
dictor
of job
performance than other ratings
and may
differ
in
regard
to
incremental
validity
as
well
as the
level
of
group
differences.
The
purpose
of
this investigation
was to
begin
the
process
of
identifying
and
evaluating
the
constructs assessed
in
employment
interviews.
The
task
of
completing
all
four
of the
steps outlined
above
is
substantial
and
cannot
be
accomplished
in a
single study;
rather,
it
probably
will
take decades
to
complete these analyses.
Our
intent
was to
thoroughly address
the
first
two
steps described
above
and
then
to
initiate
work
on the
third
and
fourth
steps.
To
address
the first two
steps,
we
constructed
a
comprehensive tax-
onomy
of
possible interview constructs based
on
current psycho-
logical
literature
and
then used this taxonomy
to
identify
con-
structs
commonly rated
in
interview studies.
To
initiate work
on
the
third
and
fourth
steps,
we
accumulated
and
meta-analytically
summarized
the
empirical data that
are
currently available
for
ratings
of the
various constructs.
As
evident later
in
this article,
data
on the
degree
to
which construct ratings reflect
the
intended
characteristics tend
to be
very sparse.
In
fact,
there
was
only
one
construct
(general mental ability)
for
which
we
found
enough data
to
analyze. Fortunately, there
was
a'much
more reasonable amount
of
data
on the
properties
of the
construct ratings, specifically
for
validity
and
group
differences,
although
the
number
of
data points
was
small
for
some
of the
constructs.
Development
of a
Taxonomy
of
Possible
Interview
Constructs
As
a
first
step
in
understanding employment interview con-
structs,
we
created
a
taxonomy
of
possible constructs.
To
help
develop this taxonomy,
we
examined
the
psychological literature
in
areas such
as
cognitive, social, personality,
and
industrial-
organizational psychology
and
identified constructs relevant
to the
employment
interview.
In
particular,
we
looked
for
established
constructs
that
have
a
long history
in
applied psychology, includ-
ing
direct applications
in
personnel selection
and
placement
(e.g.,
mental
ability,
personality, interests
and
preferences);
for
other
constructs that could
be
measured
in an
interview (e.g., organiza-
tional
fit,
social skills);
and for
characteristics
in
which many
employers
are
routinely interested
(e.g.,
job
knowledge).
This
search required judgment
on our
part
and was
guided
by our
background
and
experience
with
interview studies
and the
inter-
view
literature.
Our
collection
of
relevant constructs fell into seven
major
categories.
The
categories,
the
constructs within each cate-
gory,
and
even subfactors
for
some
of the
constructs
are
presented
next.
The first
category,
and
probably
the
best place
to
start
a
discus-
sion
of
established psychological constructs,
is
mental capability.
Performing
mental operations
is an
important part
of
most jobs
(Hunter
&
Hunter, 1984),
and
many employers
are no
doubt
interested
in how
well applicants
can do
these operations.
The first
individual construct that
we
found
in
this category
was
"general
intelligence," also called
"general
mental ability"
or
"gen-
eral cognitive
ability"
(see
Herrnstein
&
Murray, 1994; Schmidt
&
Hunter,
1998),
and it
reflects
the
overall ability
to
learn
and
process information. Earlier
in the
20th century, Spearman (1927)
proposed
the
idea that
a
central processing capability, which
he
called
g,
underlies much
of our
common mental
functioning
(see
also Thurstone, 1938).
The
existence
of g is
empirically supported
by
the
finding
that primary
areas
of
mental
functioning
such
as
math, verbal, spatial, perceptual,
and
mechanical skills correlate
moderately
with
one
another
and
form
a
single,
psychometrically
meaningful
factor.
Not
surprisingly, measures
of
general intelli-
gence have been found
to be
related
to
performance across
a
wide
range
of
jobs (Hunter
&
Hunter, 1984; Schmidt
&
Hunter, 1998).
Although
a
variety
of
psychological measures
are
available
to
assess general intelligence
(or one of its
primary areas),
a
number
of
employers still appear
to use the
employment interview
for its
assessment.
Spychalski
(1994),
for
example, developed
an
inter-
view
for
corrections
officers
that included assessment
of the
ability
to
learn procedures.
In a
similar manner, Huse (1962) evaluated
managerial applicants
on
intellectual capacity
in his
interview.
Although
ability tests
may be
superior psychometrically, many
employers continue
to
assess
mental traits
in
interviews
for a
variety
of
reasons, including logistical considerations, habit, legal
concerns,
and
even
a
basic belief
in the
accuracy
of
human
judgment
(see Dipboye, 1992).
The
second construct that
we
found
under mental capability
can
be
called
"applied
mental
skills,"
and it
reflects
the
application
of
mental
ability
to
solve organizational problems
and
address vari-
ous
organizational issues. Specific areas
of
this construct include
judgment,
decision making, problem solving,
and
planning.
The
roots
of
this construct
go
back
to the first
half
of the
20th century,
including
the use of a
practical judgment subtest
in the
U.S. Army
Alpha
Examination during World
War I
(Taylor, 1949;
Terman,
1918)
and
Cardall's
(1942) development
of an
industrial screening
test
to
measure practical
judgment.1
Prominent ability tests such
as
the
Wechsler
and the
Stanford-Binet continue
to
include items
in
which
a
specific situation
is
described (e.g.,
fire in a
movie theater)
and
test takers must indicate what they would
do in
that situation
(see Murphy
&
Davidshofer, 1998),
and the
concept
of
applying
mental capability
to
real-world contexts continues
to be
discussed
in
the
literature
(Sternberg
&
Kaufman,
1998).
Given
that most
jobs
have
at
least
some problems
to
solve
and
some decisions
to
make,
it is not
surprising that
a
number
of
employers appear
to use the
interview
for
evaluation
of
applied
mental skills. Examples
of
such applications include
Pulakos
and
Schmitt's
(1995) assessment
of the
ability
to
evaluate information
and to
make
decisions
from
that information
in
their interview
for
investigative agents
and
Berkley's
(1984)
assessment
of the
ability
to
solve problems
in his
interview
for a
corrections
position.
1
Cardall
(1942) defined practical judgment
as
"recognition
of
possible
alternatives
of
action
and the
ability
to
select
the
best"
(p. 1).
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

INTERVIEW CONSTRUCTS
899
The
third construct that
we
found
under mental capability
was
"creativity,"
and it
reflects
the
capability
to
generate innovative
ideas
and
solutions. Creativity
is
different
from
the
traditional
conceptualization
of
mental capability
in
applied psychology
(i.e.,
learning,
retaining,
and
processing information) because
it
requires
flexibility
of
thought, originality,
and the
ability
to see
beyond
current
structures
and
operations (Cohen
&
Swerdlik, 1999).
Nonetheless, many believe
it
still constitutes
a
mental operation
and
thus
is
best placed
in the
mental capability
category.2
Given
the
sparsity
of
paper-and-pencil
tests
of
creativity,
it is not
sur-
prising
that
at
least some employers
use the
interview
for its
assessment. Examples
of
such assessment include
Hoffman
and
Holden's
(1993)
evaluation
of
innovation
in
their study
of a
management position
in a gas
company
and
Chapman
and
Rowe's
(1998)
evaluation
of
creativity
in
their study
of
cooperative edu-
cation workers.
The
second category
of
psychological constructs relevant
to the
interview
is
knowledge
and
skills. Rather than focusing
on the
capability
to
process information, these constructs revolve around
information
already stored
in
long-term memory
and
include both
declarative (i.e., terms, values, names,
and
dates)
and
procedural
(i.e.,
actions, skills,
and
operations) components (see Winograd,
1975).
Not
surprisingly, knowledge
and
skills appear
to
have
at
least
some
relationship
with
general
intelligence
in
that
people
with
higher intelligence
often
can
retain more information (Bor-
man,
White,
Pulakos,
&
Oppler,
1991),
although other factors such
as
exposure, experience,
and
interest
can
influence
knowledge
retention
as
well.
The
usefulness
of
knowledge constructs
is
supported
by
evi-
dence that direct measures
of job
knowledge
and
skills
are
predic-
tive
of job
performance (Hunter
&
Hunter, 1984; Schmidt
&
Hunter,
1992). Some employers
use
background credentials
(e.g.,
education, training,
or
experience)
as
indirect
measures
of
knowl-
edge
and
skills, although such credentials
do not
have
as
strong
of
a
record
for
predicting performance (Hunter
&
Hunter, 1984)
and
are
often
used more
in the
initial screening process (Riggio, 1996).
Assessment
of
knowledge
and
skills
has
been included
in at
least
some interviews,
including
Adorno,
Binning,
Srinivasagam,
and
Williams's
(1997)
evaluation
of
"technical
knowledge"
in
their
interview
study
of
assembly workers
and
Landy's
(1976)
evalua-
tion
of
"experience"
in his
interview study
of
police
officers.
The
third category
of
selected psychological constructs
is
basic
personality
tendencies. Commonly called
"traits,"
these tendencies
reflect
long-term predispositions
to
act in
certain ways. Although
research
on
personality
has
been ongoing
for
most
of the
past
century
(e.g.,
Allport,
1937), there appears
to be a
growing con-
sensus
that
there
are five
main personality dimensions. Collec-
tively
known
as the
"Big
Five,"
these dimensions include
Extroversion, Conscientiousness, Agreeableness, Openness
to
Experience,
and
Emotional Stability. Descriptions
of
each
of
these
traits
is as
follows
(see Costa
&
McCrae, 1992; Digman, 1990;
Oliver, 1989). Extroversion
is the
basic tendency
to be
socially
active
and
includes elements
of
both basic sociability (high need
for
and
enjoyment
of
social activity)
and
power-related tendencies
(assertiveness
and
dominance). People high
on
this trait
are
typi-
cally
described
as
warm, gregarious, energetic, assertive, domi-
nant,
driven,
and
competitive. Conscientiousness reflects
the
drive
to
accomplish assigned tasks
and
duties
to the
best
of
one's
ability
and
to do so
within
the
confines
of
established procedures
and
protocols.
People
high
on
this trait
are
often
described
as
respon-
sible, dependable, competent, punctual,
deliberate,
and
respectful
of
authority.
Agreeableness
is the
basic
desire
to be
liked
by and
to
fit
in
with other people.
People
high
on
this trait
are
commonly
described
as
likable,
friendly,
warm, caring, polite,
tactful,
and
helpful.
Openness
to
Experience reflects
the
tendency
to be
open
to new
ideas
and
flexible
in
one's
thinking.
People
high
on
this
trait
are
usually
described
as
open, curious,
flexible, and
imagina-
tive. Finally, Emotional Stability reflects
the
regulation
and
man-
agement
of
one's
emotions, including doing
so in
stressful condi-
tions.
People
high
on
this trait
are
typically described
as
calm,
poised,
composed,
confident,
and
stable.
There
is a
growing body
of
literature that suggests that personality tendencies
can
predict
job
performance
(Barrick
&
Mount, 1991; Mount
&
Barrick,
1995;
Tett,
Jackson,
&
Mitchell, 1991).
This category
is
important
to
study
in
relation
to the
interview
because,
in
addition
to
mental capability
and
accumulated knowl-
edge, many employers also seem interested
in how
potential
em-
ployees would typically
act on the
job. Employers appear
to use
interviews, similar
to use of
ability tests,
for
assessment
of
per-
sonality
traits
for
reasons such
as
logistical considerations
and
habit
even though established
and
psychometrically
superior mea-
sures
are
available. Personality
has
been evaluated
in a
number
of
interview
studies,
including
Zedeck,
Tziner,
and
Middlestadt's
(1983) assessment
of
"self-confidence"
in
their interview
for fe-
male military officers; Dipboye, Gaugler, Hayes,
and
Parker's
(1992) assessment
of
"responsibility"
in
their interview
for an
entry-level corrections position;
and
Chapman
and
Rowe's
(1998)
evaluation
of
"friendliness"
in
their interview
for
cooperative
education workers.
The
fourth
category
of
relevant psychological constructs
is
applied
social
skills,
a
category related
to
basic personality ten-
dencies.
This
category reflects
the
ability
to
function
effectively
in
social situations,
the
skills
for
which
may be
influenced both
by
underlying
personality structure
and by
acquired competencies.
Historically,
the
roots
of
this category
go
back
to the first
half
of
the
20th century.
In
1920,
Thorndike
proposed
the
concept
of
"social
intelligence,"
a
concept that
he
defined
as the
ability
to
understand
others
and to act
wisely
in
human relations (see also
Thorndike, 1921). More recent derivations
of his
concept include
Gardner's (1983) concept
of
"personal intelligences" (i.e., inter-
personal
and
intrapersonal skills)
and
Sternberg,
Conway, Ketron,
and
Bernstein's (1981) concept
of
"social competence." Compared
with
basic personality, research
and
literature relating
to
applied
social skills
are
much more sparse
and
underdeveloped.
In
fact,
we
were able
to find
only
one
established measure relating
to
these
constructs—Stevens
and
Campion's (1999) test
for
teamwork
skills—and
that measure
was
designed specifically
for
selection
in
autonomous
and
semiautonomous work teams.
It is
possible that
their
test
may be
predictive
of
performance
in
more
individual-
based jobs
as
well, although
that
would have
to be
established
in
future
research.
Given
the
importance
of
social components
to
many jobs
and
the
lack
of
established measures,
it is not
surprising that many
employers
use the
interview
for
assessment
of
skills
in
social
situations.
We
were able
to
identify
four
specific social skills that
2
We
thank
one of the
reviewers
for
making this
suggestion.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

900
HUFFCUTT,
CONWAY, ROTH,
AND
STONE
employers have evaluated
in an
interview: oral communication
skills,
interpersonal skills, leadership,
and
persuasiveness. Oral
communication
reflects
the
ability
to
express (and receive) ideas
and
information
clearly, accurately,
and
convincingly,
and it has
been assessed
in
studies such
as
Robertson, Gratton,
and
Rout's
(1990)
interview
for
financial services representatives
in
England.
Interpersonal
skills refer
to the
ability
to
relate
to,
understand,
work
with,
and
develop rapport with others,
and
they have been
assessed
in
studies such
as
Dougherty,
Ebert,
and
Callender's
(1986)
interview
for
entry-level positions
at a
large energy corpo-
ration.
Leadership
is the
ability
to
direct
and
motivate others,
and
it
has
been assessed
in
studies such
as
Wiesner, Latham, Bradley,
and
Okros's (1992) interview
for
Canadian Naval officers. Finally,
persuasiveness
is the
ability
to
change other people's opinion
in
important
matters,
and it has
been assessed
in
studies such
as
Hoffman
and
Holden's
(1993)
interview
for a
management posi-
tion
in a gas
company.
The
fifth
category
of
psychological constructs
is
interests
and
preferences,
and it
represents
an
inclination toward certain areas
or
activities.
Items
in
this category would include
a
preference
for a
particular
type
of
work
or
profession,
a
preference
for a
specific
company
or
geographical area, involvement
in
related hobbies,
and
an
interest
in
certain topics
or
subjects. Research
on
interests
and
preferences
goes back
to the
turn
of the
last century, including
G.
Stanley Hall's development
of a
questionnaire
to
assess
recre-
ational
interests
in
1907 (see Cohen
&
Swerdlik, 1999).
In a
similar
manner,
the
Strong-Campbell
Interest Inventory,
a
popular
instrument
today,
was
first
published
in
1928 (see Cohen
&
Swerdlik, 1999).
Although
standard measures
of
interests
and
preferences
do not
have
a
strong record
for
predicting
job
performance (Hunter
&
Hunter,
1984), some employers still appear
to
include character-
istics
related
to
interests
and
preferences
in
their interviews. John-
son
(1991),
for
example, evaluated "interest
in
medicine"
in his
interview
for
medical residents,
and
Roth
and
Campion (1992)
assessed
"interest"
in the
position
in
their interview
for
refinery
technicians.
More recently, "organizational fit"
has
emerged
as a
unique
and
potentially
important concept related
to
organizations,
and it is our
sixth
category.
The
idea behind organizational
fit is
that each
organization
takes
on its own
unique culture
or
climate, defined
by
characteristics such
as
values, goals, norms,
and
attitudes (Cable
&
Judge,
1997;
Kristof,
1996).
The
closer
the
values
and
attitudes
of
an
individual
correspond
to
those
of the
organization,
the
better
the
fit
is
between them. Measures
of
organizational
fit
have been
shown
to
correlate
with
a
number
of
criteria, including work
attitude,
organizational tenure, prosocial behaviors,
and
work per-
formance
(Kristof, 1996;
see
also Rynes
&
Gerhart,
1990).
At
least some employers appear
to use the
interview
for
evalu-
ation
of
organizational
fit.
Such instances include
DeGroot's
(1997) assessment
of
"appreciation
of
diversity"
in his
interview
study
of
first-level managers
and
Bradley,
Bernthal,
and
Thomas's
(1998)
assessment
of
"quality orientation"
in
their interview study
of
process operators. Some have even argued that
the
ability
to
assess organizational
fit is a key
reason
for the
continued popu-
larity
of the
interview
(Karren
&
Graves,
1994).
Finally,
we
felt
it
necessary
to
include
a
"physical attributes"
category
of
constructs because some employers appear
to use the
interview
to
assess physical characteristics. Some
of the
physical
characteristics that employers
assess
are
general
in
nature, such
as
Raza
and
Carpenter's (1987) evaluation
of
attractiveness. Other
physical
characteristics that employers assess
are
more
job-related,
such
as
Grove's (1981) evaluation
of
stamina
and
agility. Although
physical attributes
are not
really psychological constructs
in a
true
sense,
we
included them
in our
construct framework
to be
thor-
ough
because they
are
rated
in at
least some interviews.
Collectively,
the
aforementioned
categories
and
constructs pro-
vided
a
workable framework
from
which
to
begin
the
process
of
identifying
the
constructs rated
in
employment interviews.
As we
noted
earlier,
these constructs were selectively chosen
and
repre-
sent
only
a
fraction
of the
constructs available
in the
psychological
literature.
There
are
other popular constructs that
we did not
use,
such
as
Festinger's
(1957)
concept
of
cognitive dissonance
and
Rotter's
(1966)
concept
of
locus
of
control.
Moreover,
there
are
alternative
ways
to
look
at
some
of
these constructs, such
as
Gardner's
(1983) theory that there
are
seven largely independent
types
of
intelligence. Regardless,
it is our
opinion that
the
con-
structs
presented above
are the
ones most relevant
for the
task
at
hand, namely, identifying
the
constructs captured
in
employment
interview
evaluations.
Degree
of
Structure
The
interview
process
can be
influenced
by a
number
of
factors,
including
use of an
interview panel, availability
of
background
information, type
of
questions asked,
and
type
of job
analysis (see
Campion
et
al.,
1997; Dipboye,
1992).
Among
these
factors,
the
degree
of
structure
is
generally considered
to be the
most impor-
tant,
not
only because
of its
effects
on the
interview
process
itself
but
also
because
of its
impact
on
reliability
and
validity (Campion
et
al.,
1997;
Huffcutt
&
Arthur, 1994; Wiesner
&
Cronshaw,
1988).
There
is a
real possibility that structure could
also
influence
the
constructs that
are
captured
in
interviews. High-structure inter-
views
differ
from
low-structure interviews
in a
number
of
ways,
many
of
which could influence construct measurement.
For in-
stance,
the
dimensions
(i.e.,
constructs) rated
in
structured inter-
views
are
more likely
to be
based
on a job
analysis than
are the
dimensions rated
in
unstructured interviews. Consequently,
it
would
not be
surprising
to see
measurement
of
more general,
impression-based constructs
in
low-structure interviews
(e.g.,
mo-
tivation,
ability
to
think)
and
measurement
of
more specific, job-
related constructs
in
high-structure interviews
(e.g.,
job
knowl-
edge, ability
to
solve problems
or
come
up
with creative solutions).
In
addition
to
tapping different constructs, unstructured interviews
tend
to
have considerably lower reliability (Conway
et
al.,
1995),
so
their ratings
may
represent less accurate measurement
of the
intended constructs.
Accordingly, when analyzing
the
frequency
and
properties
(e.g.,
validity)
of
interview constructs,
we
broke
as
many
of our
analyses
down
by
structure
as
possible.
Analyzing structural differences
was
not
always easy given limitations
in
sample size
for
some
of
the
constructs. Nonetheless,
we
felt
it was
important
to
include
analyses
of
structure given
its
centrality
to the
interview
process
and
the
outcomes
of
this
process.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

INTERVIEW CONSTRUCTS
901
Method
Search
for
Primary Data
We
conducted
an
extensive search
to
locate
usable employment inter-
view
studies
for our
investigation. Interview studies included
in
previous
interview
meta-analyses
were reviewed (Conway
et
al.,
1995;
Huffcutt
&
Arthur, 1994; McDaniel
et
al.,
1994; Wiesner
&
Cronshaw,
1988),
more
recent issues (since 1994)
of the
Journal
of
Applied Psychology
and
Personnel Psychology were examined,
the
databases PsycLIT
and
ABI-
INFORM
were searched,
and
recent conference programs were
checked.
Supplemental inquiries were
also
made
to
prominent researchers
in the
interview
area
to
obtain
any
additional studies
not
included
in the
afore-
mentioned sources. Finally, authors
of
several recent studies
in
which
the
desired information
was not
reported were contacted
to see if
that infor-
mation
was
available.
Four main
criteria
guided
our
search. First,
a
study
had to
list
the
specific
dimensions (i.e., characteristics) assessed
in
that interview. Studies
nor
listing
the
specific dimensions were excluded, such
as
that
of
Open
(1985),
who
developed
a
structured interview around
six
behaviorally
defined
dimensions
but did not
indicate what those dimensions were.
Second,
a
study
had to
involve
a
real position
in
business
and
industry
(either applicants
or
incumbents)
or a
position
in a
professional training
program that included duties
in a
real setting
(e.g.,
medical school
resi-
dent). Here,
we
excluded Walsh's (1975) study because
the
subjects were
college football players
and
also
several other studies involving high
school students being interviewed
as
part
of the
general admissions process
to
college.
Third,
a
study
had to
represent
a
typical interview.
We
dropped Hilton,
Bolin,
Parker, Taylor,
and
Walker's (1955) study because
the
interview
ratings were actually made
by
psychologists
who
read notes made
by the
original interviewers. (The original interviewers
did not
make
any
ratings.)
We
also dropped
the two
older
studies
from
Martin
(1972)
because
the
main
purpose
of
those interviews
was to
interpret
and
augment psycho-
logical test results.
Fourth,
a
study
had to
provide
at
least some supplemental information
for
the
individual dimensions
in the
interview. Consistent with
our
second
objective
to
begin exploring
and
establishing relationships
for the
con-
structs,
we
looked
for any
possible data relating
to
their properties. This
exploratory information included correlations with
job
performance (i.e.,
validity
coefficients), correlations with psychological
tests,
group
differ-
ences (race
or
sex),
and
interrater
reliability.
In
total,
we
were able
to
locate
47
employment interview studies that
met
the
four
aforementioned search
criteria.3
Citations
for
these studies
are
included
in the
general
list
of
references, identified
by an
asterisk. They
included
a
wide range
of job
types, interview designs, companies,
and
products.
A
total
of 338
assessment characteristics
(i.e.,
dimensions) were
identified
from
these studies.
The
mean number
of
dimensions
per
study
was
7.2, with
an
actual range
from
3 to 18.
Mapping
of
Interview Constructs
Using
the
framework
of
psychological constructs presented
earlier,
Allen
I.
Huffcutt,
James
M.
Conway,
and
Philip
L.
Roth attempted
to
code
each
of the 338
interview dimensions
as to the
construct that
it
reflected
(first
the
category
and
then
the
individual construct).
We
based
our
coding
on
both
the
label
of the
dimensions
and any
definitions provided
in the
studies.
Our
general strategy
was to
independently code
the
dimensions
and
then
to
discuss
and
resolve
any
differences.
To
assess
the
consistency
of the
initial coding
process,
we
compiled data
on
the
level
of
agreement among
the
three
of us.
Specifically,
we
recorded
each agreement
as a
"+"
and
each disagreement
as a
"—."
In
total, Allen
I.
Huffcutt
had an 86%
agreement rate with James
M.
Conway
and an 85%
agreement rate
with
Philip
L.
Roth. James
M.
Conway,
in
turn,
had an 82%
agreement rate with
Philip
L.
Roth.
These
results
suggest
that
the
initial
process
of
identifying
constructs
was
sufficiently
consistent.
To
further ensure that
all
coding
was as
accurate
and
free
from
bias
as
possible,
we had a
fourth
individual independently code
all
interview
dimensions.
This
person
was an
experienced applied psychologist
who
kept
current with
the
general
selection
literature
but was not
familiar with
our
present study prior
to
being contacted.
We
gave
her a
copy
of our
construct
framework
and a
list
of the
dimensions
from
the
interview studies
and
then
asked
her to
code
the
dimensions
as to the
construct they reflected.
The
subsequent agreement between
the
consensus codings
by the
first
three
authors
and
those
of the
independent rater
was
88%, again suggesting that
the
construct codings were accurate
and
unlikely
to be
biased.
Recording
of
Exploratory
Information
We
recorded whatever supplemental information
was
available
for
each
of
the 338
interview dimensions
in our
data set.
It is
important
to
note that
the
level
of
analysis with this information
was for
each individual rating
characteristic
and not for the
overall interview
scores.
As we
expected,
not
every study reported
all of the
supplemental information
for the
individual
dimensions being rated, which reduced
the
overall amount
of
information
available.
For the
validity coefficients,
we
recorded
the
unconnected
correlation
between interviewer ratings
on a
given characteristic
and
evaluation
of
overall
job
performance
by a
supervisor
or
manager.
For the
ability
correlations,
we
recorded
the
unconnected
correlation between interviewer
ratings
on a
characteristic
and
scores
on
some type
of
mental ability test.
For
racial group differences,
we
recorded
the
mean
and the
standard
deviation
of the
ratings
for a
characteristic
for
White
and
Black subjects,
respectively. From this information
we
computed
a d
score, which reflected
the
number
of
within-group
standard deviations that
the
mean
of the
White
subjects
was
different
from
the
mean
of the
Black subjects (see Hunter
&
Schmidt,
1990).
In a
similar manner,
for sex
group differences
we
recorded
the
mean
and the
standard deviation
of the
ratings
for
male
and
female
subjects,
respectively,
and
computed
a d
score
from
this information. Here,
the d
score indicated
the
number
of
standard deviations that
the
mean
of the
male subjects
was
different
from
the
mean
of the
female subjects.
The d
values
for
both race
and sex
were computed such that
a
positive value
indicated that
the
unprotected group (Whites
and
male subjects, respec-
tively)
received higher ratings,
on
average, than
the
protected group
(Blacks
and
female subjects, respectively) whereas
a
negative value indi-
cated that
the
protected group received higher ratings.
For
analysis
of
construct validity,
we
recorded
the
unconnected
correla-
tion
between interviewer ratings
and
scores
on an
established psycholog-
ical measure
of the
same construct.
We
found
a
fair
amount
of
this
information
for
mental ability,
but
corresponding correlations
for
person-
ality, applied social skills,
and the
other constructs
in the
framework were
minimal.
Therefore,
we
were forced
to
limit
this
analysis
to
mental ability
and
leave analysis
of the
remaining constructs
for
future
research.
Data
on
interrater reliability were similarly sparse. There were some data
on
interrater reliability
for
overall interview
scores
but
considerably less
data
for the
individual characteristics assessed. Moreover, what data there
were
at the
characteristic level varied
in
format, with some representing
the
reliability
of
multiple ratings
from
the
same interview
and
others repre-
senting
the
reliability
of
ratings
from
different
interviews. Accordingly,
we
were forced
to
leave this issue
for
future
research
as
well.
3
There
are
actually
only
45
studies identified
in the
reference list.
The
reason
is
that Johnson (1991)
and
U.S.
Office
of
Personnel Management
(1987)
each
had two
usable
studies.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

902
HUFFCUTT, CONWAY, ROTH,
AND
STONE
Analysis
of
Exploratory Information
and
Artifact
Corrections
For
each construct
for
which data were available,
we
computed
the
mean
of
the
exploratory statistics, namely,
the
mean validity,
the
mean cognitive
ability
correlation,
the
mean
d
value
for
race,
and the
mean
d
value
for
sex.
These means were computed
by
giving each study coefficient equal weight.
We
decided
to
compute means
in
this
manner
out of
concern that
a
handful
of
studies
with
a
relatively large sample
size
could dominate
the
results
for
a
number
of
constructs. This
was
particularly important
for a
study such
as
ours
in
which
a
number
of
cells
had
comparatively
few
studies, because
a
single
large sample-size study would
be
extremely
influential.
For
exam-
ple,
if we
used sample weighting,
the 2
largest studies
for
interpersonal
skills
(see
Table
3,
which
is
presented later) would count almost
as
much
as the
other
17
studies combined. Research
in
this area suggests that equal weighting
provides
more stable estimates than sample weighting when
a
large-sample
study
is
present (Fuller
&
Hester,
1999;
Osburn
&
Callender,
1992).
To
provide information
on the
practical significance
of our
results,
we
computed
a 90%
confidence interval
for
each
of our
mean
estimates.4
Some concern
has
been noted
in the
literature about confidence interval
formulas
that
are
based
on the
assumption that
the
effect
is
consistent
in the
population
and not
moderated
by
features
of the
situation
or the
study
methodology
(see
Erez, Bloom,
&
Wells,
1996;
Hedges
&
Vevea,
1998).
Accordingly,
we
used
a
formula
offered
by
Osburn
and
Callender (1992),
specifically
their Equation
5, to
form confidence intervals
for the
mean
correlations
between interview ratings
and
performance ratings
and for the
mean
correlations between interview ratings
and
mental ability test scores.
As
Osburn
and
Callender noted, this
formula
is
appropriate
for
both
homogeneous
and
heterogeneous situations.
We
used
a
similar equation
pro-
vided
by
Hunter
and
Schmidt
(1990,
p.
430)
to
form confidence intervals
for
mean
race
and sex
effect
sizes,
and
this equation
is
appropriate
for
both
homogeneous
and
heterogeneous situations
as
well.
In
both cases,
the
confi-
dence intervals were formed
by
taking
the
uncorrected
mean estimate plus
or
minus
1.65
times
the
square root
of the
mean sampling variance.
Finally,
as is
typical
in
selection research,
we
corrected
the
mean
observed validity estimates
for the
influence
of
various statistical artifacts.
Given
the
sparsity
of
artifact
data
in the
studies
in our
database,
we
used
the
artifact distribution approach outlined
by
Hunter
and
Schmidt
(1990).
Specifically,
we
corrected
all
mean validity estimates
(a) for
range restric-
tion
in the
interview
by
using
the
average ratio
of .74
found
by
Huffcutt
and
Arthur
(1994)
and (b) for
measurement error
in job
performance evalua-
tions
by
using
the
average
interrater
reliability value
of .52
found
by
Viswesvaran
et
al.
(1996).
There
appears
to be
some debate over
the
accuracy
of the .52
value (Murphy
&
DeShon,
2000),
but at the
present
time,
we
believe
it to be the
best estimate available.
In
a
similar manner,
we
corrected
the
mean observed correlations
be-
tween
interview construct ratings
and
mental
ability
test
scores
by
using
the
artifact
distribution approach. Here,
we
corrected
all
mean estimates
for
range
restriction
in the
interview
by
using
the .74
value noted above
and for
measurement
error
in the
ability
tests
by
using
a
value
of .90 for
their
reliability
(see
Huffcutt
et
al.,
1996). Because ability testing
is
typically
done before selection decisions
are
made,
the .74
value
for
range restriction
in
the
interview
may
have resulted
in a
slight
overcorrection
to
these
mean
effect
sizes.
No
corrections
for
artifacts were made
to the
group differences
d
values
for
race
and
sex,
although artifacts undoubtedly
had at
least some
influence
on
their magnitude
as
well.
Identification
and
Analysis
of
Interview Structure
Huffcutt
and
Arthur (1994) established
a
framework whereby interviews
can
be
classified along
four
distinct levels
of
structure. Because
the
number
of
coefficients available
for
some
of the
constructs
was
relatively
low,
we
decided
to
simplify
our
structural classification
by
collapsing this frame-
work
into
two
more general levels. Specifically,
we
classified studies
as
high
structure
if a
majority
of the
questions were specified beforehand
and
as low
structure
if the
interviewers
had
fairly
wide discretion
in
terms
of
choosing
what topics
to
cover
or at
least what questions
to
ask.
Then
we
recomputed
and
corrected
the
mean exploratory statistics
as
described
above separately
for
high-
and
low-structure interviews.
Results
Frequency
of
Construct Measurement
Results
of the
mapping
process
between psychological con-
structs
and
interview dimensions
are
shown
in
Table
1.
The
data
on
the
left
show
the
number (and percentage)
of
dimensions
in
each
of
the
major construct categories,
and the
data
in the
middle show
the
number
of
dimensions associated with each construct. Com-
mon
labels
for the
interview dimensions
are
shown
on the right.
As
indicated
in
Table
1,
the
largest
number
of
dimensions
reflected
basic personality tendencies
(35%),
followed
by
applied
social skills (28%). Mental capability (16%)
and
knowledge
and
skills
(10%)
were
the
next most frequently rated constructs. Con-
structs
in the
remaining three
categories
were rated less
frequently,
including
interests
and
preferences (4%), physical attributes (4%),
and
organizational
fit
(3%).
Differences
in the
frequency
of
construct
measurement between
low-
and
high-structure interviews
are
presented
in
Table
2. As
shown,
when expressed
as
percentages, there were noticeable
differences
in the
frequency
of
measurement
for a
number
of the
constructs.
For
mental capability, general intelligence
was as-
sessed almost
five
times more
often
in
low-structure interviews,
whereas
applied mental skills were
assessed
more
than twice
as
often
in
high-structure interviews. Among knowledge constructs,
direct knowledge
and
skills were
assessed
more than twice
as
often
in
high-structure interviews, whereas education
and
training
as
well
as
experience
and
general work history were assessed roughly
three times
as
often
in
low-structure interviews.
For
personality, Agreeableness
was
rated more than three times
more often
in
low-structure interviews than
in
high-structure
in-
terviews,
and
Emotional Stability
was
rated more than twice
as
often.
Openness
to
Experience
was
rated more frequently
in
high-structure
interviews, although
a
ratio could
not be
esti-
mated because
it was not
rated
at all in the
low-structure
interviews.
There
did not
appear
to be
substantial differences
between low-
and
high-structure frequencies
for
Extroversion
and
Conscientiousness.5
For
applied social skills, communication
and
interpersonal skills
were rated approximately twice
as
often
in
high-structure inter-
4
A
variety
of
intervals have been used
in the
literature. Viswesvaran,
Ones,
and
Schmidt
(1996),
for
example, used
80%
intervals
because
those
intervals isolated points
at
which
10% of
values would
be
higher
and 10%
would
be
lower.
Wiesner
and
Cronshaw
(1988)
formed
95%
confidence
intervals, which isolated
the
outer
2.5%
points.
We
chose
to use 90%
intervals
because
that
seemed
reasonable
given
the
other intervals
just
described
and
because
they
isolated
the
outer
5% of
values
on
either
side.
5
As
shown
in
Table
2,
there were
36
conscientiousness-type constructs
rated
in the
high-structure studies even though there were only
28
high-
structure studies
in the
data
set.
The
reason
why
there were more consci-
entiousness coefficients than high-structure studies
is
that most
of
these
studies
included
several
conscientiousness-type
dimensions.
For
example,
the
interviewers
in
Roth
and
Campion's
(1992)
study evaluated candidates
on
both reliability
and
dependability.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

INTERVIEW CONSTRUCTS
903
views,
and
leadership
was
rated more than three times
as
often.
Persuading
and
negotiating appeared
to be
rated with about
the
same frequency
in
both low-
and
high-structure interviews. Among
the
remaining
categories,
organizational
fit was
assessed about
one-and-one-half
times more
often
in
high-structure interviews,
physical attributes (both general
and
job-related) were assessed
more than
four
times
as
often
in
low-structure interviews,
and the
frequency
of
ratings
was
fairly
close
for
interests
and
preferences.
Overall, these data suggest
a
potentially important
effect,
namely,
a
tendency
for
high-
and
low-structure interviews
to
emphasize somewhat different constructs. High-structure inter-
views appear
to be
focused more
on
applied mental
skills,
direct
job
knowledge, applied social skills (communication, interpersonal
skills,
and
leadership),
and
organizational
fit,
whereas low-
structure interviews appear
to be
focused more
on
general mental
ability,
background credentials (education, training,
and
experi-
ence), some
aspects
of
personality
(agreeableness
and
emotional sta-
bility),
and
physical attributes.
In
short, these results suggest that
structure
affects
not
only
the
conduct
of the
interview
but
also what
constructs
are
rated.
We
next examine relevant validity evidence.
Validity
of
Construct
Ratings
Mean
correlations
between
ratings
of the
various
constructs
and
overall
performance evaluations
are
presented
in
Table
3. To
enhance
the
stability
and
generalizability
of the
results,
we did not
attempt
to
interpret
the
mean value
for any
individual construct
with
fewer than
four
studies (see
Viswesvaran
et
al.,
1996).
For our
overall analysis
of all
studies,
the
mean corrected
validity
coefficients ranged
from
.24 to .58
(with
a
mean
of
.36).
We
noted
in the
introduction that some construct ratings
may be a
stronger predictor
of job
performance than other ratings,
and
these
results provide support
for
that idea.
The
highest
mean
corrected
validities
were observed
for
ratings
of
creativity (.58), agreeable-
ness (.51), organizational
fit
(.49), leadership (.47), emotional
stability (.47),
job
knowledge (.42),
and
interpersonal skills (.39).
The
lowest mean validities were observed
for
ratings
of
interests
and
preferences
(.24),
general intelligence (.24), communication
skills (.26),
and
applied mental skills
(.28).
Four things should
be
noted
in
regard
to the
aforementioned
findings.
First, some
of
these mean correlations were based
on a
fairly
small number
of
studies
(e.g.,
there were
four
studies
for
creativity).
As a
result, these mean estimates should
be
interpreted
with
some caution because they could include sampling error.
Moreover,
use of
mean
artifact
distribution values
for
interview
range restriction
and
performance unreliability could have resulted
in
over-
or
undercorrection
of
these estimates.
Second, these mean validity estimates included both low-
structure
and
high-structure studies. Structured interviews gener-
ally
have higher validity than unstructured interviews (Wiesner
&
Cronshaw, 1988),
so
mean estimates based
on a
higher proportion
of
structured studies would
be
expected
to
have higher mean
validity.
For
example,
five of the six
studies
in
which
job
knowl-
edge
was
rated were
of
high structure.
Third, these values reflect
the
mean validity across
different
types
of
jobs.
It is
possible that some constructs
may be
predictive
of
performance
in
some jobs
but not in
other jobs, whereas other
constructs
may be
more universal predictors
of
performance. That
constructs such
as
organizational
fit,
emotional
stability,
interper-
sonal skills,
and
creativity
had
high validity across various types
of
positions
may
suggest that they
are
somewhat universal predictors
of
performance. Alternatively,
it may be
that some
of
these high-
validity
constructs
are
easier
to
observe
in
interview situations,
the
"signs"
versus
"samples"
distinction brought
up
earlier
by
Werni-
mont
and
Campbell (1968). Interpersonal skills
and
some elements
of
emotional stability (e.g., maturity, stress tolerance),
for
exam-
ple,
may be
fairly
salient during
an
interview.
Finally, these results suggest that ratings
of
creativity have
higher validity than ratings
of
general intelligence, something that
appears
to be
inconsistent with existing selection literature. There
are at
least
two
possible explanations.
One is
that
the
mean validity
estimate
for
creativity
was
based
on
only
four
total studies
and
thus
may
have contained sampling error.
The
other
is
that
the
nature
of
the
questions asked
to
evaluate creativity
may
have made
it
easier
to
evaluate than general intelligence.
For
example, some
of the
studies
in
which creativity
was
assessed used behavior description
questions (e.g.,
Hoffman
&
Holden, 1993;
Huffcutt
et
al.,
in
press).
Having
candidates describe
a
past situation
in
which they came
up
with
a
creative solution
to an
organizational problem might
be
fairly
easy
to
evaluate.
In
contrast, general intelligence
is
more
abstract
and
general, which could make
it
more
difficult
to
assess
in
an
interview (unless
the
questions posed mathematical
or
verbal
problems
like
those
found
on
typical
ability
tests).
Results
of the
structure analyses
are
also presented
in
Table
3.
As we
expected,
the
mean validities
for
constructs rated
in
more
structured
interviews were higher overall than
the
mean validities
for
constructs rated
in
less structured interviews.
In
particular,
the
mean
corrected validity across
all
constructs rated
in
low-structure
interviews
was
.24, whereas
the
mean corrected validity across
all
constructs
rated
in
high-structure interviews
was
.39. There were
three constructs that
had
four
or
more studies
for
both low-
structure
and
high-structure interviews,
and
they
allowed
a
more
direct comparison
of
validity.
These
constructs were applied men-
tal
skills, conscientiousness,
and
interpersonal skills.
The
mean
validities
for
structured ratings were higher
in all
three
cases,
although
the
difference
was
somewhat smaller
for
interpersonal
skills. This smaller difference
for
interpersonal skills could reflect
sampling error,
or it
could suggest that interpersonal skills
can be
rated
reasonably well
in
low-structure interviews.
Among constructs rated
in
high-structure interviews, there were
some clear differences
in
terms
of
validity. Emotional stability
and
organizational
fit had the
highest mean corrected validities (.56
and
.58, respectively), whereas interests
and
preferences
and
com-
munication
skills
had the
lowest corrected validities (.26
and
.31,
respectively). Differences among constructs rated
in low
structure
were
less pronounced, with interpersonal skills having
the
highest
mean
corrected
validity
(.31)
and
applied mental skills having
the
lowest corrected validity
(.13).
As we
noted earlier, some
of
these
estimates were based
on a
limited number
of
studies
and
should
be
viewed
as
tentative.
Correlation
Between
Construct
Ratings
and
Mental
Ability
Test
Scores
Mean correlations between interview construct ratings
and
men-
tal
ability test scores
for all
studies
are
shown
in the first
part
of
Table
4. As we
noted earlier,
all of the
mean correlations
in our
study, including
those
for low and
high structure, were computed
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

904
HUFFCUTT,
CONWAY,
ROTH,
AND
STONE
Table
1
Interview
Constructs
and
Associated Dimension Labels
Category
and
constructCommon interview dimension label
55
16.3 Mental
capability—ability
to
learn, organize,
process,
and
evaluate information
Major
theme: assessing
how
well candidates
can
think
General intelligence
20
Verbal
ability
1
Applied
mental skills
28
Creativity
and
innovation
6
33 9.8
Knowledge
and
skills—accumulated
knowledge,
skills,
and
abilities
Major
theme: assessing
what
candidates know
and
what
they
can do
Job
knowledge
and
skills
14
Education
and
training
6
Experience
and
general work history
13
118
34.9 Basic personality
tendencies—predispositions
to
act in
certain ways
Major
theme: assessing
how
candidates
are
likely
to act in the
workplace
Extroversion
21
Conscientiousness
55
Agreeableness
10
Openness
to
experience
6
Emotional
stability
21
Other
personality traits
5
94
27.8 Applied social
skills—ability
to
function
effectively
in
social
situations
Major
theme:
assessing
how
well
candidates
can
deal
with
other
people
Communication
skills
26
Interpersonal skills
43
Leadership
20
Persuading
and
negotiating
5
15
4.4
Interests
and
preferences—inclination
toward
certain
areas
or
activities
Major
theme: assessing what candidates like
to do
Occupational interests
13
Subject
and
topic interests
1
Hobbies
and
general
interests
1
Intellectual
capacity, intelligence, mental ability, ability
to
learn,
learning
the
work, analytical
ability,
mental
alertness, ability
to
think
quickly, perceptiveness
Support
for
arguments
Problem solving, problem assessment, judgment, decision
making,
critical thinking, planning, organizing
Creativity,
creativeness,
innovation
Knowledge,
technical
knowledge,
job
knowledge,
product
knowledge,
use of
tools, budgeting
Education,
academic
achievement,
grades
in
school
Experience, work history, exposure
Assertiveness,
dominance, ability
to
control situations,
drive,
energy,
decisiveness,
ambition,
positive
outlook
Dependability, responsibility, reliability, timeliness, sense
of
duty, need
for
achievement, motivation, willingness
to
work hard,
initiative,
persistence, time management,
moral
character, integrity, ethics, professionalism
Friendliness,
likability,
empathy, concern
for
others,
attitude, general attitude
Adaptability,
flexibility,
openness
to
change
Emotional stability, stress tolerance, performance under
stress, poise,
social
adjustment, self-control, coping,
maturity,
self-confidence,
ego
strength
Independence, self-reliance, self-understanding
Oral communication, communication skills, expression,
ability
to
present ideas, conversation ability, voice
and
speech,
listening
Interpersonal skills, interpersonal relations, social skills,
social sensitivity, working with others, ability
to
relate
to
people,
rapport, tact, ability
to
deal
with
people,
adapting
to
people,
teamwork, cooperation, team focus,
team building
Leadership, vision, coaching, developing people,
delegation, maintaining control, directing others,
activating
others, developing teamwork
in
others,
building
morale, discipline
Persuasiveness, ability
to
negotiate
Job
interest, interest
in
position, investment, commitment
to
a
career
Extracurricular activities
Hobbies
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

INTERVIEW CONSTRUCTS
905
Table
1
(continued)
Category
and
constructCommon interview dimension label
11
3.3
Organizational
fit—compatibility of
attitudes
and
beliefs with
those
of the
organization
Major
theme: assessing what things
the
candidates
really believe
in
Values
and
moral standards
11
12
3.6
Physical attributes
General physical attributes
8
Job-related
physical
skills
4
Quality orientation, safety orientation, appreciation
for
diversity,
acceptance
of
company mission, customer
service, customer focus, belief
in
product value, pride
in
the
organization
Health, appearance, attractiveness
Physical
requirements,
physical
ability,
stamina,
agility
Note.
A
total
of 338
dimensions were coded.
Nc
= the
number
of
dimensions
in
each
of the
major construct
categories;
% = the
percentage
of
dimensions
in
each
of the
major construct categories;
n = the
number
of
dimensions associated with each individual construct.
by
equally
weighting
the
respective correlations,
and
that practice
was
maintained
for the
ability correlations. Across construct cat-
egories,
the
mean corrected correlation
with
mental ability test
scores
was
.17.
The low
magnitude
of
this mean correlation
suggests
that
g
does
not
saturate interview ratings
at the
construct
level,
a
finding
that
is
consistent
with
previous
meta-analytic
research
involving
total
interview
scores
(Huffcutt
et
al.,
1996).
There
did
appear
to be at
least some
differences
across con-
structs
in
terms
of the
correlation
with
mental ability test scores.
In
particular,
ratings
of
general intelligence,
job
knowledge,
and
experience
had the
highest mean correlations, whereas ratings
of
applied
social skills (communication, interpersonal skills,
and
leadership),
applied mental skills,
and
interests
and
preferences
had
the
lowest mean correlations.
Table
2
Differences
in
Construct
Frequency
Between Low-
and
High-Structure
Interviews
Category
and
construct
Low
structure
High
structure
Mental capability
General intelligence
Specific ability
Applied mental skills
Creativity
and
innovation
Knowledge
and
skills
Job
knowledge
and
skills
Education
and
training
Experience
and
general work history
Basic personality tendencies
Extroversion
Conscientiousness
Agreeableness
Openness
to
experience
Emotional
stability
Other
personality
traits
Applied social skills
Communication
skills
Interpersonal skills
Leadership
Persuading
and
negotiating
Interests
and
preferences
Occupational interests
Hobbies
and
extracurricular activities
Organizational fit
Values
and
moral standards
Physical attributes
General physical attributes
Job-related physical skills
25
15
1
6
3
15
3
4
8
48
8
19
7
0
12
2
23
7
11
3
2
7
5
2
3
3
9
6
3
19.2
11.5
0.8
4.6
2.3
11.5
2.3
3.1
6.2
36.9
6.2
14.6
5.4
0.0
9.2
1.5
17.7
5.4
8.5
2.3
1.5
5.4
3.8
1.5
2.3
2.3
6.9
4.6
2.3
30
5
0
22
3
18
11
2
5
70
13
36
3
6
9
3
71
19
32
17
3
8
8
0
8
8
3
2
1
14.4
2.4
0.0
10.6
1.4
8.7
5.3
1.0
2.4
33.7
6.3
17.3
1.4
2.9
4.3
1.4
34.1
9.1
15.4
8.2
1.4
3.8
3.8
0.0
3.8
3.8
1.4
1.0
0.5
Note. There were
19
studies
in the
low-structure category with
a
total
of 130
characteristics rated
and 28
studies
in
the
high-structure category with
a
total
of 208
characteristics rated.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

906
HUFFCUTT,
CONWAY,
ROTH,
AND
STONE
Table
3
Mean
Validity
Correlations
for
Ratings
of
Interview Constructs
Overall
Category
and
construct
Mental capability
General intelligence
Applied mental skills
Creativity
and
innovation
Knowledge
and
skills
Job
knowledge
and
skills
Education
and
training
Experience
and
general
work
history
Basic personality tendencies
Extroversion
Conscientiousness
Agreeableness
Openness
to
experience
Emotional stability
Other personality traits
Applied social
skills
Communication
skills
Interpersonal skills
Leadership
Persuading
and
negotiating
Interests
and
preferences
Occupational interests
Organizational
fit
Values
and
moral standards
Physical
attributes
General physical attributes
Job-related physical skills
k
8
13
4
6
1
3
8
22
4
2
6
1
9
19
8
3
9
5
1
2
75S
1,916
5,027
296
2,617
312
495
1,055
3,532
344
527
917
102
2,963
3,620
633
245
914
912
312
935
r,.v
.13
.15
.32
.23
.05
.27
.18
.18
.28
.16
.26
.08
.14
.21
.26
.13
.13
.27
-.18
.15
90% CI
.04-.22
.08-.22
.21-.33
.11-.35
—
.11-.43
.12-.24
.13-23
.26-.30
—
.20-.32
—
.06-.22
.16-.26
.18-.34
.10-.16
.05-21
.17-37
—
—
?,V(C)
.24
.28
.58
.42
—
.49
.33
.33
.51
—
.47
—
.26
.39
.47
.24
.24
.49
—
—
k
6
4
2
1
1
2
3
6
1
0
2
0
2
5
2
2
2
1
1
1
TSS
1,699
1,160
152
31
312
380
650
1,656
68
—
539
—
783
1,191
152
152
380
537
312
471
Low
structure
rxy
90% CI
r^c)
.14
.06-.22
.26
.07
-.02-.16
.13
.28 — —
.49 — —
.05 — —
.17
— —
.12
.09-.15
.22
.13
.05-.21
.24
.25 — —
— — —
.18 — —
— — —
.05 — —
.17
.08-.26
.31
.40 — —
j
i
.08 — —
.07 — —
-.18
— —
.10 — —
High structure
k
2
9
2
5
0
1
5
16
3
2
4
1
7
14
6
1
7
4
0
1
TSS
217
3,867
144
2,586
—
115
405
1,876
276
527
378
102
2,180
2,429
481
93
534
375
—
464
rv
.11
.19
.36
.18
—
.47
.22
.20
.29
.16
.31
.08
.17
.22
.22
.18
.14
.32
—
.19
90% CI
—
.11-27
—
.07-29
—
—
.14-30
.15-35
.27-31
—
.27-35
—
.07-.27
.16-.28
.13-31
—
.07-.21
.23-.41
—
—
^(0
—
35
—
.33
—
—
.40
37
.53
—
.56
—
31
.40
.40
—
.26
.58
—
—
Note. Confidence intervals
(CI)
and
corrected mean validity
correlations
are not
shown
for
constructs
with fewer than
three
coefficients, indicated
by
those
cells
with
dashes,
k - the
number
of
dimensions
in
each construct category that
had
provided
that information;
TSS = the
total number
of
subjects
associated with those dimensions;
~rxy
= the
uncorrected
correlation between
the
interview ratings
and job
performance evaluations;
r^Cc)
= the
correlation
corrected
for
range restriction
in the
interview
and
measurement error
in
performance evaluations.
Results
for
low-structure
and
high-structure interviews
are
shown
in the
remaining part
of
Table
4.
Across constructs,
the
mean
corrected correlation with mental ability test scores
was
.10
for
high-structure interviews
and
.31
for
low-structure interviews.
This analysis
sug-gests
greater saturation
of
mental ability
in
rat-
ings
for
low-structure interviews,
a finding
that
has
direct impli-
cation
for
racial group differences. Racial group differences
are
discussed
in the
next section.
On a
slightly
different
note, these results also provide
at
least
some
information
on the
construct
validity
of
interview ratings.
Ratings
of
general intelligence correspond directly
to
mental abil-
ity
test scores
at a
construct level
and
should have
had a
higher
correlation
with
the
test scores than
the
other constructs. Although
ratings
of
general intelligence
did in
fact
have
the
highest mean
correlation among
all of the
constructs,
the low
magnitude
of
this
mean
correlation (.32 uncorrected,
.44
corrected) suggests that
these
two
measurement
devices
are far
from equivalent.
More
than
that,
this
finding
casts doubt
on the
ability
to
evaluate general
intelligence
with
an
employment interview.
Racial
Group
Differences
in
Construct Ratings
Mean effect
sizes
for
racial
differences
in the
various construct
ratings
are
shown
in
Table
5.
Across
all
studies
and
constructs,
the
mean
d
value
was
.30. This value suggests
at
least some racial
group differences
in
interview construct ratings. These differences
are
considerably
less
than those
for
mental ability tests, which
can
approach
a
difference
of one
standard deviation (see Hunter
&
Hunter,
1984). However,
it is
important
to
note that
the
1.00 value
applies
to
ability tests
as a
whole, whereas
our
values apply
to the
average group differences resulting
from
a
single rating.
As
indicated
in
Table
5,
there appeared
to be
some racial group
differences
across constructs.
The
highest mean
effect
size
was for
ratings
of
general intelligence
and
ratings
of
experience
and
gen-
eral
work
history
(both
had a
mean
d
value
of
.49).
These
values
are
high, especially because they reflect racial group differences
for
a
single rating,
and may
suggest some caution
in the
evaluation
of
these characteristics with interviews.
The
lowest mean value
was
for
ratings
of
applied mental skills, which
had a
mean
d of
.13.
Racial group differences
for
low-structure
and
high-structure
interviews, respectively,
are
also shown
in
Table
5. As
indicated
in
Table
5,
racial
group differences were considerably higher overall
for
low-structure interviews.
In
particular,
the
average
d
value
across
all
constructs rated
in
low-structure interviews
was
.51,
whereas
the
average
d
value across
all
constructs rated
in
high-
structure
interviews
was
.13.
It
would
appear
that
less
formal
use
of
a job
analysis
and
giving interviewers more discretion
in
terms
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

INTERVIEW CONSTRUCTS
907
Table
4
Mean
Correlations Between Interview Construct Ratings
and
Mental Ability
Test
Scores
Overall
Low
structure
High
structure
Category
and
construct
TSS
90%
CI
r,(c)
k
TSS
CI
r,(c)
k TSS
90% CI
?K(c)
Mental capability
General intelligence
Applied mental skills
Creativity
and
innovation
Knowledge
and
skills
Job
knowledge
and
skills
Education
and
training
Experience
and
general
work
history
Basic
personality
tendencies
Extroversion
Conscientiousness
Agreeableness
Openness
to
experience
Emotional
stability
Other personality traits
Applied
social skills
Communication skills
Interpersonal
skills
Leadership
Persuading
and
negotiating
Interests
and
preferences
Occupational interests
Organizational
fit
Values
and
moral standards
Physical attributes
General physical attributes
Job-related physical skills
5
8
2
6
0
4
4
16
4
1
6
1
5
12
5
2
7
2
1
0
795
3,273
200
2,616
—
688
678
2,495
688
76
1,437
101
2,318
2,240
783
200
1,058
489
103
—
.32
.05
.20
.27
—
.26
.15
.11
.16
-.02
.15
-.03
.08
.00
.07
.05
.09
.05
.11
—
.22-.42
.00-.
10
—
.19-.35
—
.2S-.29
.02-.28
.05-.
17
.10-.22
—
.08-.22
—
.03-.
13
-.06-.06
.05-.09
—
.03-.
15
—
—
—
.44
.07
—
.37
—
.36
.21
.16
.23
—
.21
—
.11
.00
.10
—
.13
—
—
—
4
1
1
2
0
3
2
6
3
0
3
1
1
2
1
1
2
0
1
0
623
197
107
134
—
516
413
1,036
516
—
516
101
103
138
107
107
413
—
103
—
.34
.22-.46
.13 —
.29 —
34
— —
.27
.23-.31
.24 —
.17
.07-.27
.17
.10-.24
—
—
.19
.08-30
-.03
—
.11 —
-.13
—
.04 —
.06 —
.21
—
— —
.11
—
—
—
.46
1
—
7
— 1
—
4
—
0
.37 1
— 2
.24 10
.24 1
— 1
.27
3
—
0
4
— 10
—
4
— 1
—
5
—
2
—
0
—
0
172
3,076
93
2,482
—
172
265
1,449
172
76
921
—
2,035
2,101
676
93
645
489
—
—
.25
.04
.10
.24
—
.25
.07
.07
.13
-.02
.11
—
.07
.02
.08
.04
.05
.05
—
—
—
-.02-.
10
—
.22-.26
—
—
—
-.01-.15
—
—
.05-.
17
—
.01-.13
-.04-.08
.05-.
11
—
.01-.09
—
—
—
—
.06
—
.33
—
—
—
.10
—
—
.16
—
.10
.03
.11
—
.07
—
—
—
Note. Confidence intervals
(CI)
and
corrected
mean correlations
are not
shown
for
constructs with fewer than
three
coefficients, indicated
by
those cells
with
dashes,
k = the
number
of
dimensions
in
each construct category that
had
provided that information;
TSS = the
total number
of
subjects associated
with
those dimensions;
rg
= the
uncorrected
correlation between
the
interview ratings
and job
performance evaluations;
rg(c)
= the
correlation corrected
for
range restriction
in the
interview
and
measurement
error
in
performance evaluations.
of
the
interview
process
tended
to
result
in
higher racial group
differences
in the
subsequent ratings.
Finally,
it
should
be
noted that
racial
group differences
for
ratings
of
personality characteristics overall were
at
last three
times higher than those
found
for
paper-and-pencil
tests.
For
example,
Ones
and
Viswesvaran
(1998)
found
that racial group
differences
for
personality dimensions such
as
integrity tended
to
be
below
an
effect
size
of
.10.
However,
we
note that self-reports
of
personality
and
other ratings
of
personality
often
are
only
slightly correlated (Mount,
Barrick,
&
Strauss, 1994). Interviews
may
represent
a
situation
in
which
a
different
method
of
measure-
ment
(using others' ratings) provides
different
personality-related
information
that
is
associated with somewhat
larger
racial group
differences.
mean
effect
size (i.e., favoring female subjects)
for
ratings
of
applied
social
skills.
Mean
effect
sizes
for
low-structure
and
high-structure studies,
respectively,
are
also shown
in
Table
6. The
mean
d
value
for all
constructs rated
in
high-structure interviews
was
.00,
suggesting
that
these interviews overall
had
little
or no
impact
on
female
subjects.
The
mean
d
value across
all
constructs rated
in
low-
structure interviews
was
.23,
suggesting
that
these interviews
overall
had
some impact
on
female subjects.
There
was one
con-
struct
for
which
a
direct comparison could
be
made between
low-structure
and
high-structure interviews,
and
that construct
was
conscientiousness. Results
for
conscientiousness suggested higher
group differences
for
low-structure
interviews
than
for
high-
structure
interviews (mean
ds of .34 and
.12,
respectively).
Sex
Group
Differences
in
Construct
Ratings
Mean
effect
sizes
for sex
differences
in the
various construct
ratings
are
shown
in
Table
6.
Across
all
constructs,
the
mean
d
value
was
.06,
which suggests negligible
sex
differences
in
inter-
view
construct
ratings
overall.
There
were
some
differences across
constructs, such
as the
positive
.13
mean
effect
size
(i.e.,
favoring
male
subjects)
for
ratings
of
general intelligence
and the
—.13
Discussion
The
first
purpose
of
this investigation
was to
develop
a
taxon-
omy
of
possible constructs that employment interviews could
measure. Using literature
from
a
number
of
areas
in
psychology,
we
constructed
a
comprehensive taxonomy with seven
different
categories
of
constructs: mental capability, knowledge
and
skills,
basic personality tendencies, applied social skills, interests
and
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

908
HUFFCUTT,
CONWAY, ROTH,
AND
STONE
Table
5
Mean
Effect
Sizes
for
Racial Group
Differences
in
Interview Construct Ratings
Overall
Low
structureHigh structure
Category
and
construct
TSS
90%
CI
TSS
90% CI
TSS
90% CI
Mental capability
General intelligence
Applied
mental
skills
Creativity
and
innovation
Knowledge
and
skills
Job
knowledge
and
skills
Education
and
training
Experience
and
general work history
Basic personality tendencies
Extroversion
Conscientiousness
Agreeableness
Openness
to
experience
Emotional stability
Other personality traits
Applied social skills
Communication skills
Interpersonal skills
Leadership
Persuading
and
negotiating
Interests
and
preferences
Occupational interests
Organizational
fit
Values
and
moral standards
Physical attributes
General physical attributes
Job-related physical skills
6
6
0
1
0
3
3
15
4
1
6
1
7
6
2
0
2
3
1
3
2,331
1,566
—
103
—
1,208
1,333
5,443
1,313
267
2,359
103
1,864
1,733
463
—
1,105
568
103
919
.49
.13
—
1.07
—
.49
.18
.30
.33
.05
.34
.52
.39
.22
.25
—
.33
-.12
.48
.21
.2S-.73
.00-.26
—
—
—
.28-.70
.11-.25
.20-.40
.18-.48
—
.23-.4S
—
.22-.S6
.11-.33
—
—
—
-.15—
.09
—
.08-.34
5
1
0
1
0
2
1
8
2
0
3
1
2
2
0
0
1
0
1
2
1,564
471
—
103
—
441
338
2,554
441
—
912
103
574
652
—
—
338
—
103
652
.58
.49
—
1.07
—
.63
.28
.41
.52
—
.45
.52
.87
.37
—
—
.39
—
.48
.31
.35-.81
1
—
5
—
0
—
0
—
0
— 1
—
2
.26-.S6
7
— 2
— 1
.33-.S7
3
—
0
—
5
—
4
—
2
—
0
— 1
—
3
—
0
— 1
767
1,095
—
—
—
767
995
2,889
872
267
1,447
—
1,290
1,081
463
—
767
568
—
267
.03
.06
—
—
—
.21
.14
.17
.15
.05
.24
—
.21
.14
.25
—
.26
-.12
—
.02
—
-.02-.
14
—
—
—
—
—
.11-.23
—
—
.12-.36
—
.11-.31
-.03-.25
—
—
—
-.15—
.09
—
—
Note.
A
positive
effect
size
indicates that White
candidates
had
higher mean interview ratings than Black candidates. Confidence intervals
(CI)
are not
shown
for
constructs with fewer than
three
coefficients, indicated
by
those
cells
with
dashes,
k = the
number
of
dimensions
in
each
construct
category
that
provided
racial group information;
TSS = the
total number
of
subjects
associated
with those
dimensions;
rfWB = the
mean
effect
size
for
White-Black
group
differences.
preferences, organizational
fit, and
physical attributes.
The
devel-
opment
of
such
a
taxonomy
is
obviously
a
critical component
in
the
process
of
understanding interview constructs, because
it
guided
the
classification
and
interpretation
of
data
in
this study
and
may
influence future
research
in
this
area.
We
believe
that
our
taxonomy
provides
a
comprehensive
and
meaningful
basis
on
which
interview construct research
can
build and,
as
such, repre-
sents
a
contribution
to the
interview literature
in and of
itself.
The
second purpose
of
this investigation
was to
evaluate which
constructs
in
this framework
are
actually rated
in
employment
interviews and,
perhaps
more important, which
are the
most
com-
monly
assessed.
To
make this assessment,
we
compiled
a
database
of
338
characteristics that were rated
in 47
actual interview studies.
These
47
studies included
a
diverse mixture
of
companies, prod-
ucts,
format
(e.g., level
of
structure, type
of
questions),
job
com-
plexity,
and
sources (e.g., journal articles, technical reports, con-
ference
papers,
and
dissertations).
We
believe
this
database
was
reasonably
representative
of
employment interviews
in
general
and
provided
a
sound basis
on
which
to
analyze interview constructs.
Our
results
suggest that personality traits
and
applied social
skills
are
rated more
often
in
employment interviews than
are any
other
type
of
construct.
These
constructs reflect behavioral ten-
dencies
and
provide employers with
an
idea
of how
potential
employees
are
likely
to act on the job and how
well they
can
interact
with
other
employees.
Given
the
frequency
with
which
they appeared
to be
rated (combined, they accounted
for
more than
60% of all the
rated characteristics),
it
would seem that many
employers
are
interested
in
behavioral tendencies
and
that they
are
an
important part
of
many
jobs.
Among
these
characteristics,
conscientiousness
was the
single
most
commonly
rated
construct,
as
it
accounted
for
more than
16% of all
ratings
and
appeared
under
labels such
as
responsibility, dependability, initiative,
and
persistence. Interpersonal skills
was the
next most frequently rated
construct,
as it
accounted
for
approximately
13%
for all
ratings
and
appeared under labels such
as
interpersonal relations, social skills,
team focus,
and the
ability
to
work with people.
Mental
capability
and
knowledge
and
skills were
the
next most
frequently
rated constructs
after
behavioral tendencies. These con-
structs reflect either what candidates already know
or how
well
they
can
process
new
information,
and
combined, they accounted
for
more than
25% of all the
characteristics rated.
In
fact, behav-
ioral
tendencies
and
mental-type
constructs together accounted
for
almost
90% of all
interview ratings, whereas interests, organiza-
tional
fit, and
physical attributes accounted
for the
remaining
ratings.
A
key finding
from
our
frequency analyses
was
that low-
structure
and
high-structure interviews
do not
tend
to
measure
the
same constructs.
In
particular, low-structure interviews
often
focus
more
on
constructs such
as
general intelligence, education
and
training,
experience,
and
interests, whereas high-structure
inter-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

INTERVIEW CONSTRUCTS
909
Table
6
Mean
Effect
Sizes
for Sex
Group
Differences
in
Interview Construct Ratings
Overall
Low
structureHigh
structure
Category
and
construct
TSS
90%
CI
TSS
90% CI
TSS
90% CI
Mental capability
General intelligence
Applied mental skills
Creativity
and
innovation
Knowledge
and
skills
Job
knowledge
and
skills
Education
and
training
Experience
and
general work history
Basic personality tendencies
Extroversion
Conscientiousness
Agreeableness
Openness
to
experience
Emotional stability
Other personality traits
Applied social skills
Communication
skills
Interpersonal skills
Leadership
Persuading
and
negotiating
Interests
and
preferences
Occupational interests
Organizational fit
Values
and
moral standards
Physical
attributes
General physical attributes
Job-related physical skills
4
6
0
1
0
2
4
12
4
1
3
0
4
8
2
0
2
3
1
2
1,250
1,598
—
89
—
1,168
1,496
4,041
1,458
464
1,490
—
911
1,490
219
—
1,168
325
171
645
.13
-.13
—
-.08
—
.59
.30
.19
.04
-.01
.16
—
-.26
-.12
-.02
—
.41
-.28
-.77
.67
-.17-.43
-.22—
.04
—
—
—
—
.16-.44
.06-.32
-.27-35
—
-.29-.61
—
-.43—
.09
-.23—
.01
—
—
—
-.43—
.13
—
—
3
0
0
0
0
1
1
4
2
0
1
0
0
1
0
0
1
0
1
1
697
—
—
—
—
345
345
1,052
516
—
345
—
—
181
—
—
345
—
171
181
.08
—
—
—
—
.49
.38
.34
.06
—
.54
—
—
-.41
—
—
.34
—
-.77
1.38
-.31-.47
1
—
5
—
0
—
0
—
0
— 1
—
3
.17-51
8
—
2
— 1
— 2
—
0
—
4
—
7
2
—
0
— 1
3
—
0
— 1
823
1,598
—
—
—
823
1,151
2,989
942
464
1,145
—
911
1,309
219
—
823
325
—
464
.28
-.13
—
—
—
.69
.27
.12
.02
-.01
-.03
—
-.26
-.08
-.02
—
.47
-.28
—
-.05
—
-.22—
.04
—
—
—
—
.09-.45
-.03-.27
—
—
—
—
-.43—
.09
-.18-.06
—
—
—
-.43—
.13
—
—
Note.
A
positive
effect
size indicates that male candidates
had
higher mean interview ratings than female candidates. Confidence intervals
(CI)
are not
shown
for
constructs with fewer than three coefficients, indicated
by
those
cells
with
dashes,
k = the
number
of
dimensions
in
each construct category that
provided
that
sex
group information;
TSS = the
total number
of
subjects associated
with
those dimensions;
rfMF = the
mean
effect
size
for
male-female
group differences.
views
often
focus
more
on
constructs such
as job
knowledge
and
skills,
organizational
fit,
interpersonal
and
social skills,
and ap-
plied mental skills (e.g., problem solving, decision making).
These
differences
are
most likely
due to the
more
frequent
(and stringent)
use of
formal
job
analysis methodology
in the
development
of
high-structure
interviews. Thus,
it
would appear that structure
influences
not
only
the
procedural
conduct
of the
interview
(e.g.,
consistency
in the
asking
of
questions
and the
manner
in
which
responses
are
scored)
but
also
what constructs
are
rated.
We are
unaware
of the
issue
of
different
constructs being empirically
substantiated
in any
other place
in the
literature
and
believe this
represents
a
contribution
in and of
itself.
Another purpose
of
this investigation
was to
begin
to
accumu-
late information
on the
properties
of
interview construct ratings,
including
validity
and
group differences.
We
were able
to
identify
a
number
of
constructs
that
had
high validity across jobs, including
job
knowledge, interpersonal skills, creativity, agreeableness,
job
knowledge, emotional stability, leadership,
and
organizational
fit.
We
were
also
able
to
identify
some constructs that
had
fairly
low
validity
across jobs, including interests
and
preferences
and
com-
munication
skills.
What
is
particularly interesting
is
that many
of the
high-validity
constructs
are
those that tend
to be
assessed
more often
in
high-
structure
interviews. Although
one
would expect higher validity
for
constructs rated more often
in
high-structure interviews, such
a
correspondence could suggest another explanation
for why
struc-
tured interviews have higher validity than low-structure inter-
views.
It is
generally accepted that structured interviews have
higher validity
in
part because they represent more reliable
assess-
ment
of
responses (Conway
et
al,
1995). Taken
as a
whole,
our
evidence suggests that structured interviews also have higher
va-
lidity
because
the
constructs
rated
more
frequently
in
them
(e.g.,
job
knowledge, interpersonal skills, organizational
fit)
tend
to be
better predictors
of
performance than
the
constructs rated more
frequently
in
low-structure interviews (e.g., interests, education,
experience).
What lends
further
support
to the
above notion
is
that
it is
largely
consistent
with
existing selection literature. Hunter
and
Hunter
(1984),
for
example,
found
that standard measures
of job
knowledge
had
high
validity
whereas standard measures
of
edu-
cation,
experience,
and
interests
had
much lower validity.
The one
exception
was
general mental ability,
as
existing selection litera-
ture
suggests
it is a
strong predictor
of
performance (Hunter
&
Hunter,
1984),
yet
interview ratings
of it did not
correlate strongly
with
job
performance. However,
as
noted
a
little later,
the
results
of
our
construct analysis suggest that interviewer ratings
of
general
intelligence
do not
tend
to
have
a
strong association with
the
mental ability construct.
On
a
more practical note,
we
were able
to
identify
several
constructs that appear
to
have
a
desirable combination
of
proper-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

910
HUFFCUTT,
CONWAY, ROTH,
AND
STONE
ties
for
structured
interviews.
In
particular, applied mental skills
(e.g., decision making, problem solving, judgment), interpersonal
skills
(e.g., interpersonal relations, rapport, tact, cooperation),
and
conscientiousness
(e.g., reliability,
dependability,
persistence)
all
appear
to
provide reasonable validity
with
low
racial differences
and
very
low sex
differences
(which
often
favor
women).
So,
when
there
is a
choice
of
which constructs
to
focus
on
with
a
structured
interview
(based
on the job
analysis),
our
results suggest that
the
above
constructs
are a
better choice than other constructs like
interests
and
communication skills. These recommendations
should
be
viewed
as
tentative
of
course because several
of the
findings
are
based
on a
relatively small number
of
studies.
In
addition,
we
should note that several other constructs, particularly
creativity,
organizational
fit
(e.g.,
appreciation
for
diversity, qual-
ity
orientation, pride
in the
organization)
and
emotional stability
(e.g., stress tolerance, poise, self-control), look very promising
in
terms
of
validity,
but
there were
insufficient
data available
to
assess their group differences.
In
contrast,
identifying
desirable constructs among less struc-
tured
interviews
was
more
difficult.
The
ratings made
from
these
interviews
were generally less valid
and had
higher race
and sex
differences.
Among
the
constructs
for
which
we had
sufficient
data
to
analyze, conscientiousness-related characteristics were probably
the
most desirable because they provided
at
least some validity
with
only moderate levels
of
race
and sex
group
differences.
In
comparison, ratings
of
general intelligence
had
similar validity
but
higher
levels
of
race
and sex
differences. Ratings
of
interpersonal
skills
actually
had
higher
validity
than either ratings
of
conscien-
tiousness
or
ratings
of
general
intelligence,
but we did not
have
sufficient
data
to
analyze their group differences. Again,
we do not
advocate
use of
low-structure interviews
but
want
to
provide
at
least
some guidance
for
them
in
case
an
organization insists
on
using
them.
One
other purpose
of
this investigation
was to
begin
the
process
of
determining
the
degree
to
which interview ratings actually
reflect
the
intended constructs. This
is an
important issue
and one
that
should
not be
taken
for
granted, especially because
it has
been
raised
in
other areas
of
selection.
For
example,
a
common
finding
in
the
assessment center literature (Fleenor, 1996; Sackett
&
Dre-
her,
1982)
is
that individual ratings tend
to
align more with
the
exercise
from
which they came (e.g., in-basket) than
with
the
specific
characteristic being rated (e.g., leadership).
Although
data
on the
correspondence between interview ratings
and
paper-and-pencil
scores
of the
same constructs were generally
sparse,
we did find
sufficient
information
to
make
a
partial
assess-
ment
with general intelligence.
Our
results
did not
provide strong
support
for the
construct
validity
of
these ratings,
at
least
not for
low-structure
interviews.
Rather,
these
results
suggest
that
assess-
ment
of
general intelligence attributes (e.g., ability
to
learn, intel-
lectual
capacity)
is
probably best
left
to
more traditional paper-
and-pencil instruments.
As
more data become available,
it may be
possible
to
determine
if the
relationship between ratings
of
general
intelligence
and
mental ability test scores
is
stronger
for
structured
interviews.
An
interesting question
is why
ratings
of
general intelligence
do
not
correspond very well
to
mental
ability
test
scores
in
low-
structure
interviews.
It
could
be a
matter
of
reliability, because
the
reliability
of
low-structure interviews
in
general
is not
very good
(Conway
et
al.,
1995).
Alternatively,
it may be
that ratings
of
general intelligence
are
picking
up on
other constructs, such
as
general impressions
or
characteristics that
are
salient during
the
interview
process. Extroversion,
for
example,
has
been
found
to
influence
interviewer evaluations
in at
least
one
study (Caldwell
&
Burger,
1998).
So
what have
we
learned
about employment interview con-
structs? Four major conclusions
emerge
from
the
above discussion.
First,
a
variety
of
constructs
can be and are
rated
in
employment
interviews. Although personality, social skills, mental capability,
and
knowledge
and
skills
are the
most commonly
rated,
other
constructs
such
as
organizational
fit,
interests,
and
physical
at-
tributes
are
rated
as
well.
Second,
high-structure interviews tend
to
focus
on
different
constructs than low-structure interviews, most
likely
due to the
more stringent
use of job
analysis methodology
in
the
former. Third, part
of the
reason
for the
higher validity
of
structured
interviews
may be
that they focus more
on
constructs
that
have
a
stronger relationship with performance (e.g.,
job
knowledge
vs.
education). Finally, group differences (both race
and
sex) should
be
taken into account when employment inter-
views
are
being developed because
the
extent
of
these differences
varies
across
constructs (even
for
constructs with comparable
validity).
Where
do we go
from
here? Probably
one of the
most important
needs
for
future
research
is to
assess
the
degree
to
which inter-
viewers
can
accurately
assess
personality
and
social
characteris-
tics.
Specifically,
we
need correlations between interviewer ratings
and
independent assessments
of the
same
characteristics.
In
addi-
tion,
we
call
for
more research
on the
relative contribution
of
reliability
and
specific construct measurement
to
high-structure
interview
validity. Finally,
we
need
a lot
more
data
to
flesh
out our
understanding
of
many
of
these
constructs and,
in
particular,
to
allow
better control
for
moderators such
as the
degree
of
structure.
There does appear
to be a
tendency
for
researchers
to
report
information
only
on
total interview
scores,
including relationships
with
ability, personality
tests,
or
both.
We
call
for
interview
researchers
to
thoroughly report their results
for
each dimension
and
not
just
for the
total interview
scores.
We
also call
for
more
job-specific interview
research
to
tease
out
which constructs
are
more universal
and
which
are
more specific
to
given classes
of
jobs
(e.g.,
sales,
managerial).
As
always, limitations
of
this investigation should
be
noted.
First, although
the
total number
of
dimensions
collected
for
con-
struct
identification
and
mapping
was
reasonable
(TV
=
338),
the
number
of
coefficients
for
some
of the
analyses
was
low. Small
sample sizes made
these
results
more
tentative
and did not
allow
for
analysis
of
other potential moderator variables
in
addition
to
structure.
Second,
we did not find
data
on the
correlation between
interviewer
ratings
of
personality
traits
and
social
skills
and
cor-
responding paper-and-pencil measures
of the
same traits. This
information
would allow determination
of the
degree
to
which
interviewers
are
successful
in
assessing these characteristics.
Third,
we did not find
enough
interrater
reliability
data
to
analyze,
because
again
most
of
these data were reported only
for
total
interview
scores.
Fourth,
we
could
not
correct
the
validity
and
ability
correlations
individually
for
artifacts, which
may
have
led
to
some inconsistencies
in the
corrected
mean values. Finally,
constructing
the
construct taxonomy required judgment
on our
part,
and it is
quite
possible
that other
researchers
might have
organized
it in a
different
way.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

INTERVIEW CONSTRUCTS
911
References
References marked with
an
asterisk
indicate
studies
included
in the
meta-analysis.
*Adorno,
A. J.,
Binning,
J. F.,
Srinivasagam,
N., &
Williams,
K. B.
(1997,
April). Incremental validity
of
structured panel interview ratings
and
ability
measures. Paper presented
at the
12th Annual Conference
of the
Society
for
Industrial
and
Organizational Psychology,
St.
Louis,
MO.
*Albrecht,
P.
A.,
Glaser,
E. M., &
Marks,
J.
(1964).
Validation
of a
multiple-assessment procedure
for
managerial personnel. Journal
of
Applied
Psychology,
48,
351-360.
Allport,
G. W.
(1937). Personality.
New
York: Holt.
Arvey,
R. D., &
Campion,
J. E.
(1982).
The
employment interview:
A
summary
and
review
of
recent research. Personnel Psychology,
35,
281-322.
Barrick,
M.
R.,
&
Mount,
M. K.
(1991).
The Big
Five
personality
dimen-
sions
and job
performance:
A
meta-analysis. Personnel Psychology,
44,
1-26.
*Berkley,
S.
(1984).
VII.
Validation report corrections
officer
trainee.
Commonwealth
of
Pennsylvania, State Civil Service Commission.
Bobko,
P.,
Roth,
P. L., &
Potosky,
D.
(1999). Derivation
and
implications
of
a
meta-analytic
matrix incorporating cognitive ability, alternative
predictors
and job
performance. Personnel Psychology,
52,
561-589.
*Bolanvich,
D. J.
(1994).
Selection
of
female engineering trainees. Journal
of
Educational Psychology,
35,
545-553.
Borman,
W. C.,
White,
L.
A.,
Pulakos,
E.
D.,
&
Oppler,
S. H.
(1991).
Models
of
supervisory
job
performance ratings. Journal
of
Applied
Psychology,
76,
863-872.
*Bosshardt,
M. J.
(1992).
Situational interviews versus behavior descrip-
tion interviews:
A
comparative validity study. Unpublished doctoral
dissertation, University
of
Minnesota, Twin Cities Campus.
*Bradley,
K.
M.,
Bernthal,
P., &
Thomas,
J.
(1998,
April). Factor structure
and
predictive
validity
of
structured employment interviews:
An
appli-
cation
of
industrial psychology
in a
Kuwaiti petrochemical corporation.
Paper presented
at the
13th Annual Conference
of the
Society
for
Industrial
and
Organizational
Psychology,
Dallas,
TX.
Cable,
D.
M.,
&
Judge,
T. A.
(1997).
Interviewers' perceptions
of
person-
organization
fit and
organizational selection decisions. Journal
of
Ap-
plied Psychology,
82,
546-561.
Caldwell,
D.
F.,
&
Burger,
J. M.
(1998). Personality characteristics
of job
applicants
and
success
in
screening interviews. Personnel Psychology,
51,
119-136.
Campion,
M.
A.,
Campion,
J.
E.,
&
Hudson,
J. P., Jr.
(1994).
Structured
interviewing:
A
note
on
incremental validity
and
alternative question
types.
Journal
of
Applied Psychology,
79,
998-1002.
Campion,
M.
A.,
Palmer,
D.
K.,
&
Campion,
J. E.
(1997).
A
review
of
structure
in the
selection interview. Personnel Psychology,
50,
655-702.
Campion,
M.
A.,
Pursell,
E.
D.,
&
Brown,
B. K.
(1988). Structured
interviewing:
Raising
the
psychometric properties
of the
employment
interview.
Personnel Psychology,
41,
25-42.
Cardall,
A. J.
(1942). Preliminary manual
for the
Test
of
Practical
Judg-
ment. Chicago: Science Research Associates.
*Chapman,
D.
S.,
&
Rowe,
P. M.
(1998).
The
impact
of
videoconference
media, interview structure,
and
interviewer
sex on
interviewer evalua-
tions
in the
employment interview:
A
field experiment. Unpublished
manuscript.
Cohen,
R.
J.,
&
Swerdlik,
M. E.
(1999). Psychological testing
and
assess-
ment:
An
introduction
to
tests
and
measurement (4th ed.). Mountain
View,
CA:
Mayfield.
*Conard,
M. A.
(1988).
The
contribution
of the
interview
to the
prediction
of
job
performance. (Doctoral dissertation, University
of
Connecticut,
1988).
Dissertation Abstracts International,
49,
5555.
Conway,
J.
M.,
Jako,
R.
A.,
&
Goodman,
D. F.
(1995).
A
meta-analysis
of
interrater
and
internal consistency reliability
of
selection interviews.
Journal
of
Applied Psychology,
80,
565-579.
*Conway,
J.
M.,
&
Peneno,
G. M.
(1999).
Comparing
structured
interview
question types: Construct validity
and
applicant reactions. Journal
of
Business
and
Psychology,
13,
485-506.
Costa,
P.
T.,
Jr.,
&
McCrae,
R. R.
(1992). Revised
NEO
Personality
Inventory
(NEO-PI-R)
and NEO
Five-Factor Inventory
(NEO-FFI)
pro-
fessional manual. Odessa,
FL:
Psychological Assessment Resources.
*Dalessio,
A.
T.,
&
Silverhart,
T. A.
(1994).
Combining biodata, test,
and
interview information: Predicting decisions
and
performance
criteria.
Personnel Psychology,
47,
303-315.
*DeGroot,
T.
(1997).
The
impact
of
managerial nonverbal cues
on the
reactions
of
subordinates. Unpublished doctoral dissertation, University
of
Florida.
*Delaney,
E. C.
(1954).
Teacher selection
and
evaluation:
With
special
attention
to the
validity
of the
personal interview
and the
National
Teacher
Examinations
as
used
in one
selected community. Unpublished
doctoral dissertation, Columbia University.
Digman,
J. M.
(1990). Personality structure: Emergence
of the
five-factor
model. Annual Review
of
Psychology,
41,
417-440.
Dipboye,
R. L.
(1992).
Selection interviews: Process perspectives. Cincin-
nati,
OH:
South-Western.
*Dipboye,
R.
L.,
Gaugler,
B.
B.,
Hayes,
T.
L.,
&
Parker,
D. S.
(1992).
Individual
differences
in the
incremental validity
of
interviewers'
judg-
ments. Unpublished manuscript.
*Dougherty,
T.
W.,
Ebert,
R.
J.,
&
Callender,
J. C.
(1986). Policy capturing
in
the
employment interview. Journal
of
Applied
Psychology,
71,
9-15.
Erez,
A.,
Bloom,
M.
C.,
&
Wells,
M. T.
(1996).
Using random rather than
fixed
effects
models
in a
meta-analysis: Implications
for
Situational
specificity
and
validity
generalization.
Personnel Psychology,
49,
275-
306.
Festinger,
L.
(1957).
A
theory
of
cognitive dissonance. Stanford,
CA:
Stanford
University Press.
Fleenor,
J. W.
(1996). Constructs
and
developmental assessment centers:
Further troubling empirical
findings.
Journal
of
Business
and
Psychol-
ogy,
10,
319-335.
Fuller,
J.
B.,
&
Hester,
K.
(1999).
Comparing
the
sample-weighted
and
unweighted
meta-analysis:
An
applied perspective. Journal
of
Manage-
ment,
25,
803-828.
*Gaines,
L.
K.,
&
Lewis,
B. R.
(1982). Reliability
and
validity
of the
oral
interview
board
in
police promotions:
A
research note. Journal
of
Criminal
Justice,
10,
403-419.
Gardner,
H.
(1983). Frames
of
mind:
The
theory
of
multiple intelligences.
New
York: Basic Books.
*Gordon,
M.
J.,
&
Lincoln,
J. A.
(1976). Family practice resident selection:
Value
of the
interview. Journal
of
Family Practice,
3,
175-177.
*Green,
P.
C.,
Alter,
P., &
Carr,
A. F.
(1993). Development
of
standard
anchors
for
scoring generic past-behavior questions
in
structured inter-
views. International Journal
of
Selection
and
Assessment, 7(4), 203-
212.
*Grove,
D. A.
(1981).
A
behavioral consistency approach
to
decision
making
in
employment selection. Personnel Psychology,
34,
55-64.
Harris,
M. M.
(1999). What
is
being measured?
In R. W.
Eder
& M. M.
Harris
(Eds.),
The
employment interview handbook (pp.
143-157).
Thousand Oaks,
CA:
Sage.
Hedges,
L.
V.,
&
Vevea,
J. L.
(1998).
Fixed-
and
random-effects models
in
meta-analysis.
Psychological Methods,
3,
486-504.
Herrnstein,
R.
J.,
&
Murray,
C.
(1994).
The
bell curve: Intelligence
and
class structure
in
American
life.
New
York:
Free
Press.
Hilton,
A.
C.,
Bolin,
S.
F.,
Parker,
J.
W.,
Jr., Taylor,
E.
K.,
&
Walker,
W.
B.
(1955).
The
validity
of
personnel
assessments
by
professional
psychologists. Journal
of
Applied Psychology,
39,
287-293.
*Hoffman,
C.
C.,
&
Holden,
L. M.
(1993). Dissecting
the
interview:
An
application
of
generalizability
analysis.
In D. L.
Denning
(Chair),
Psy-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

912
HUFFCUTT,
CONWAY,
ROTH,
AND
STONE
chometric
analysis
of
the
structured interview. Symposium conducted
at
the 8th
Annual
Conference
of the
Society
for
Industrial
and
Organiza-
tional
Psychology,
San
Francisco,
CA.
Huffcutt,
A.,
&
Arthur,
W., Jr.
(1994). Hunter
and
Hunter (1984) revisited:
Interview
validity
for
entry-level jobs. Journal
of
Applied Psychol-
ogy,
79,
184-190.
Huffcutt,
A.,
Roth,
P., &
McDaniel,
M.
(1996).
A
meta-analytic
investi-
gation
of
cognitive ability
in
employment interview evaluations: Mod-
erating
characteristics
and
implications
for
incremental
validity.
Journal
of
Applied Psychology,
81,
459-473.
*Huffcutt,
A.,
Weekley,
J.,
Wiesner,
W.,
DeGroot,
T., &
Jones,
C. (in
press). Evaluation
and
comparison
of
situational
and
behavior
descrip-
tion
interview
questions. Personnel Psychology.
Hunter,
J. E., &
Hunter,
R. F.
(1984).
Validity
and
utility
of
alternative
predictors
of job
performance. Psychological Bulletin,
96,
72-98.
Hunter,
J.
E.,
&
Schmidt,
F. L.
(1990). Methods
of
meta-analysis:
Cor-
recting
error
and
bias
in
research
findings.
Newbury
Park,
CA:
Sage.
*Huse,
E. F.
(1962). Assessments
of
higher-level personnel:
IV. The
validity
of the
assessment techniques based
on
systematically varied
information.
Personnel Psychology,
15,
195-205.
Janz,
T.
(1982).
Initial
comparisons
of
patterned behavior description
interviews
versus
unstructured
interviews. Journal
of
Applied Psychol-
ogy,
67,
577-580.
*Johnson,
E.
K.
(1991).
The
structured interview: Manipulating structuring
criteria
and the
effects
on
validity,
reliability,
and
practicality. (Doctoral
dissertation,
Tulane
University,
1990). Dissertation Abstracts Interna-
tional,
51,
5622.
*Jones,
D.
E.,
Jr.
(1978). Predicting teacher processes with
the
Teacher
Perceiver
Interview.
Unpublished doctoral dissertation,
Virginia
Poly-
technic
Institute
and
State
University.
Karren,
R.
J.,
&
Graves,
L. M.
(1994). Assessing person-organization
fit
in
personnel
selection:
Guidelines
for
future
research.
International
Journal
of
Selection
and
Assessment, 2(3),
146-156.
*Kinicki,
A.
J.,
Lockwood,
C.
A.,
Horn,
P.
W.,
&
Griffeth,
R. W.
(1990).
Interviewer
predictions
of
applicant
qualifications
and
interviewer
va-
lidity:
Aggregate
and
individual
analyses. Journal
of
Applied Psychol-
ogy,
75,
477-486.
Klimoski,
R. J.
(1993).
Predictor constructs
and
their measurement.
In N.
Schmitt
& W. C.
Borman
(Eds.),
Personnel selection
in
organizations
(pp.
99-134).
San
Francisco: Jossey-Bass.
Kristof,
A. L.
(1996).
Person-organization
fit:
An
integrative review
of its
conceptualizations, measurement,
and
implications. Personnel Psychol-
ogy,
49,
1-49.
*Landy,
F. J.
(1976).
The
validity
of the
interview
in
police
officer
selection. Journal
of
Applied Psychology,
61,
193-198.
Latham,
G. P.,
Saari,
L.
M.,
Pursell,
E.
D.,
&
Campion,
M. A.
(1980).
The
situational
interview. Journal
of
Applied Psychology,
65,
422-427.
*Little,
J. P.,
Shoenfelt,
E.
L.,
&
Brown,
R. D.
(2000,
April).
The
situa-
tional versus
the
patterned-behavior-description interview
for
predicting
customer-service performance. Paper presented
at the
15th
Annual
Con-
ference
of the
Society
for
Industrial
and
Organizational Psychology,
New
Orleans,
LA.
*Martin,
M. A.
(1972).
Reliability
and
validity
of
Canadian Forces selec-
tion interview procedures (Report
No.
72-4).
CFB
Toronto, Downsview,
Ontario, Canada: Canadian
Forces
Personnel Applied Research Unit.
McDaniel,
M.
A.,
Whetzel,
D.
L.,
Schmidt,
F.
L.,
&
Maurer,
S.
(1994).
The
validity
of
employment interviews:
A
comprehensive review
and
meta-
analysis.
Journal
of
Applied
Psychology,
79,
599-616.
*Moreano,
A.
G.,
&
Sproule,
C. F.
(1976). Reliability
and
other data
on
structured oral examinations. Unpublished manuscript, Commonwealth
of
Pennsylvania,
State
Civil Service Commission.
Mount,
M.
K.,
&
Barrick,
M. R.
(1995).
The Big
Five personality dimen-
sions: Implications
for
research
and
practice
in
human
resources man-
agement.
Research
in
Personnel
and
Human Resources Manage-
ment,
13,
153-200.
Mount,
M.
K.,
Barrick,
M.
R.,
&
Strauss,
J. P.
(1994).
Validity
of
observer
ratings
of the Big
Five personality factors. Journal
of
Applied Psychol-
ogy,
79,
272-280.
Murphy,
K.
R.,
&
Davidshofer,
C. O.
(1998).
Psychological testing:
Principles
and
applications (4th ed.). Upper Saddle River,
NJ:
Prentice
Hall.
Murphy,
K.
R.,
&
DeShon,
R.
(2000).
Inter-rater
correlations
do not
estimate
the
reliability
of job
performance ratings. Personnel Psychol-
ogy,
53,
873-900.
Oliver,
P. J.
(1989).
Towards
a
taxonomy
of
personality descriptors.
In D.
Buss
& N.
Cantor
(Eds.),
Personality psychology: Recent trends
and
emerging directions (pp.
261-271).
New
York:
Springer-Verlag.
Ones,
D.,
&
Viswesvaran,
C.
(1998). Gender, age,
and
race differences
on
overt integrity tests: Results
across
four
large-scale
job
applicant data
sets. Journal
of
Applied Psychology,
83,
35-42.
Orpen,
C.
(1985).
Patterned behavior description interviews versus
un-
structured interviews:
A
comparative validity study. Journal
of
Applied
Psychology,
70,
774-776.
Osburn,
H.
G.,
&
Callender,
J.
(1992).
A
note
on the
sampling variance
of
the
mean
unconnected
correlation
in
meta-analysis
and
validity general-
ization. Journal
of
Applied Psychology,
77,
115-122.
*Prewett-Livingston,
A.
J.,
Feild,
H.
S.,
Veres,
J.
G.,
Ill,
&
Lewis,
P. M.
(1994).
Effects
of
face
on
interview ratings
in a
situational panel inter-
view.
Journal
of
Applied Psychology,
81,
178-186.
*Prien,
E. P.
(1962). Assessments
of
higher-level personnel:
V. An
analysis
of
interviewers' predictions
of job
performance. Personnel Psychology,
15,
319-334.
*Pulakos,
E.
D.,
&
Schmitt,
N.
(1995).
Experience-based
and
situational
interview questions: Studies
of
validity. Personnel Psychology,
48,
289-308.
*Raza,
S.
M.,
&
Carpenter,
B. N.
(1987).
A
model
of
hiring decisions
in
real employment interviews. Journal
of
Applied Psychology,
72,
596-
603.
*Reeb,
M.
(1969).
A
structured interview
for
predicting military
adjust-
ment. Occupational Psychology,
43,
193-199.
Riggio,
R.
(1996).
Introduction
to
industrial/organizational
psychology
(2nd
ed.).
New
York: Harper Collins.
*Robertson,
I.
T.,
Gratton,
L.,
&
Rout,
U.
(1990).
The
validity
of
situa-
tional
interviews
for
administrative
jobs.
Journal
of
Organizational
Behavior,
11,
69-76.
*Roth,
P.
L.,
&
Campion,
J. E.
(1992).
An
analysis
of the
predictive power
of
the
panel interview
and
pre-employment
tests. Journal
of
Occupa-
tional
and
Organizational Psychology,
65,
51-60.
Rotter,
J. B.
(1966).
Generalized
expectancies
for
internal
vs.
external
reinforcement.
Psychological Monographs,
80(1,
Whole
No.
609).
Rynes,
S.
L.,
&
Gerhart,
B.
(1990).
Interviewer assessments
of
applicant
"fit":
An
exploratory investigation. Personnel Psychology,
43,
13-35.
Sackett,
P.
R.,
&
Dreher,
G. F.
(1982).
Constructs
and
assessment center
dimensions: Some troubling empirical
findings.
Journal
of
Applied
Psychology,
67,
401-410.
Schmidt,
F.
L.,
&
Hunter,
J. E.
(1992).
Development
of a
causal model
of
processes
determining
job
performance. Current Directions
in
Psycho-
logical Science,
1,
89-92.
Schmidt,
F.
L.,
&
Hunter,
J. E.
(1998).
The
validity
of
selection
methods
in
personnel psychology: Practical
and
theoretical implications
of 85
years
of
research
findings.
Psychological Bulletin, 124,
262-274.
Schmidt,
F.
L.,
&
Rader,
M.
(1999). Exploring
the
boundary conditions
for
interview validity: Meta-analytic validity
findings for a new
interview
type. Personnel Psychology,
52,
445-464.
*Sparks,
C. P.
(1974).
Validity
of a
selection
program
for
commercial
drivers
(PR
74-9).
Unpublished manuscript.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

INTERVIEW CONSTRUCTS
913
*Sparks,
C. P.
(1977). Selecting process operator
and
laboratory techni-
cian trainees
(PR
77-15).
Unpublished manuscript.
*Sparks,
C. P., &
Manese,
W. R.
(1970). Interview ratings
with
and
without knowledge
of
pre-employment
test
scores.
The
Experimental
Publication System,
4,
142.
Spearman,
C.
(1927).
The
abilities
of
man: Their
nature
and
measurement.
New
York:
Macmillan.
*SpychaIski,
A.
(1994).
A
test
of a
model
of
employment interview
infor-
mation gathering. Unpublished master's
thesis,
Rice University.
Sternberg,
R. J.,
Conway,
B. E.,
Ketron,
J. L., &
Bernstein,
M.
(1981).
People's
conceptions
of
intelligence. Journal
of
Personality
and
Social
Psychology,
41,
37-55.
Sternberg,
R.
J.,
&
Kaufman,
J. C.
(1998).
Human
abilities.
Annual Review
of
Psychology,
49,
479-502.
Stevens,
M.
J.,
&
Campion,
M. A.
(1999).
Staffing
work teams: Develop-
ment
and
validation
of a
selection test
for
teamwork settings. Journal
of
Management,
2,
207-228.
Taylor,
H. R.
(1949).
[Review
of the
Cardall
Test
of
Practical
Judgment].
In
O.
K.
Euros
(Ed.),
The
third mental measurements yearbook (pp.
735-736).
New
Brunswick,
NJ:
Rutgers University
Press.
Terman,
L. M.
(1918).
The use of
intelligence tests
in the
army. Psycho-
logical Bulletin,
15,
177-187.
Tett,
R. P.,
Jackson,
D.
N.,
&
Mitchell,
M.
(1991). Personality measures
as
predictors
of job
performance:
A
meta-analytic
review. Personnel Psy-
chology,
44,
703-742.
Thorndike,
E. L.
(1920). Intelligence
and its
uses. Harper's Magazine,
CXL,
227-235.
Thorndike,
E. L.
(1921). Educational psychology:
The
original nature
of
man
(Vol.
1). New
York: Columbia
Teacher's
College.
Thurstone,
L. L.
(1938). Primary mental abilities.
In
Psychometric mono-
graphs
(No.
1).
Chicago: University
of
Chicago Press.
Ulrich,
L.,
&
Trumbo,
D.
(1965).
The
selection interview since 1949.
Psychological Bulletin,
63,
100-116.
*U.S.
Office
of
Personnel Management.
(1987).
The
structured interview.
Washington,
DC:
Office
of
Examination Development, Alternative
Ex-
amining
Procedures Division.
*Van Iddekinge,
C.
H.,
Roth,
P.
L.,
Huffcutt,
A.
I.,
&
Eidson,
C.
E.,
Jr.
(2000).
Structured interview ethnic group
differences:
Greater than
we
thought? Manuscript submitted
for
publication.
Viswesvaran,
C.,
Ones,
D. S., &
Schmidt,
F. L.
(1996).
Comparative
analysis
of the
reliability
of job
performance ratings. Journal
of
Applied
Psychology,
81,
557-574.
Walsh,
U. R.
(1975).
A
test
of the
construct
and
predictive validity
of a
structured interview. Unpublished
doctoral
dissertation, University
of
Nebraska—Lincoln.
Walters,
L.
C.,
Miller,
M.
R.,
&
Ree,
M. J.
(1993). Structured interviews
for
pilot selection:
No
incremental validity. International Journal
of
Aviation Psychology,
3(1),
25-38.
Wernimont,
P.
P.,
&
Campbell,
J. P.
(1968). Signs, samples,
and
criteria.
Journal
of
Applied Psychology,
52,
372-376.
Wiesner,
W.,
&
Cronshaw,
S.
(1988).
A
meta-analytic investigation
of the
impact
of
interview format
and
degree
of
structure
on the
validity
of the
employment interview. Journal
of
Occupational Psychology,
61,
275-
290.
*Wiesner,
W.
H.,
Latham,
G. P.,
Bradley,
P.
J.,
&
Okros,
A. C.
(1992,
June).
A
comparison
of the
situational
and
behavior description inter-
views
in the
selection
of
naval
officers:
Preliminary results. Paper
presented
at the
annual convention
of the
Canadian Psychological
As-
sociation, Quebec City, Quebec, Canada.
Winograd,
T.
(1975).
Frame representations
and the
declarative-procedural
controversy.
In D.
Bobrow
& A.
Collins
(Eds.),
Representation
and
understanding: Studies
in
cognitive science (pp.
185-210).
New
York:
Academic
Press.
*Woodworth,
D.
G.,
Barren,
F.,
&
MacKinnon,
D. W.
(1957).
An
analysis
of
life
history interviewer's ratings
for 100 Air
Force captains (Research
Rep.
No.
AFPTRC-TN-57-129).
Lackland
Air
Force Base,
TX: Air
Force Personnel
and
Training Research Center.
Wright,
P.
M.,
Lichtenfels,
P.
A.,
&
Pursell,
E. D.
(1989).
The
structured
interview:
Additional studies
and a
meta-analysis.
Journal
of
Occupa-
tional
Psychology,
62,
191-199.
*Zedeck,
S.,
Tziner,
A.,
&
Middlestadt,
S. E.
(1983). Interviewer validity
and
reliability:
An
individual analysis approach. Personnel Psychol-
ogy,
36,
355-370.
Received
July
7,
1999
Revision received October
26,
2000
Accepted October
26,
2000
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.