Campbell & Wiernik (2015)

ANNUAL
REVIEWS Further The Modeling and

Click here for quick links to
Annual Reviews content online,
including:
Assessment of Work
Annu. Rev. Organ. Psychol. Organ. Behav. 2015.2:47-74. Downloaded from www.annualreviews.org
• Other articles in this volume

Performance
Access provided by University of Minnesota - Twin Cities on 11/24/15. For personal use only.
• Top cited articles

• Top downloaded articles
• Our comprehensive search
John P. Campbell and Brenton M. Wiernik
Department of Psychology, University of Minnesota, Minneapolis, Minnesota 55455;
email: campb006@umn.edu
Annu. Rev. Organ. Psychol. Organ. Behav. 2015. Keywords

2:47–74
performance models, contextual performance,
The Annual Review of Organizational Psychology
and Organizational Behavior is online at organizational citizenship, counterproductive work behavior,
orgpsych.annualreviews.org performance dynamics, performance appraisal
This article’s doi:
10.1146/annurev-orgpsych-032414-111427
Abstract
Copyright © 2015 by Annual Reviews. Individual work role performance drives the entire economy. It is or-
All rights reserved ganizational psychology and organizational behavior’s (OP/OB’s)
most crucial dependent variable. In this review, alternative specifica-
tions for the definition and latent structure of individual performance
are reviewed and summarized. Setting aside differences in terminol-
ogy, the alternatives are remarkably similar. The Campbell (2012)
model is offered as a synthesized description of the content of the la-
tent structure. Issues pertaining to performance dynamics are then
reviewed, along with the role played by individual adaptability to
changing performance requirements. Using the synthesized model of
the latent content structure and dynamics of performance as a back-
drop, issues pertaining to the assessment of performance are summa-
rized. The alternative goals of performance assessment, general
measurement issues, and the construct validity of specific methods
(e.g., ratings, simulations) are reviewed and described. Cross-cultural
issues and future research needs are noted.
47
INTRODUCTION
This article addresses the current state of the literature regarding the modeling and assessment of
performance in a work role. The focus is on individual performance. It is the basic building block
on which the entire economy is based (Kim & Ployhart 2014). Without individual performance
there is no team performance, no unit performance, no organizational performance, no economic
sector performance, no GDP. Despite its importance, research on performance does not compare
in size or scope to research on its antecedents and consequences. Of the 1,914 dependent variables
reported in primary empirical research articles in The Journal of Applied Psychology, Personnel
Psychology, and The Academy of Management Journal between 2008 and 2014, only 350 (18%)
are measures of individual performance at work. Certainly, other dependent variables are extremely
important, including individual work satisfaction, commitment, engagement, stress/health, and work/
family balance. However, without individual performance, there can be no job to be satisfied with, no
organization to be committed to, and no work to balance with family. We should strive to understand
individual performance to the fullest extent possible.
We focus on several central themes. In the remainder of this first section, we consider what
performance is and what it is not. In the second section, we review the similarity and dissimilarity
of alternative content models of performance and argue that the latent structure of performance is
invariant across levels, functional specialties, organizations, and cultures. In the third section, we
also consider the issues of performance dynamics and adaptability to distinguish between the
content of performance and its processes and context. In the fourth section, we consider issues
related to the assessment of performance in a work role. We highlight recent advances in the
measurement of performance, including the use of ratings, work simulations, and technology-
enhanced performance monitoring systems. We describe the potential opportunities and pitfalls
various assessment methods offer for providing meaningful performance information for different
assessment purposes. In the last section, we consider implications of research findings for practice
and areas for future research.
What Performance Is
Until the 1980s, there were virtually no attempts to model individual job performance as a con-
struct. There was only the “criterion problem” (Austin & Villanova 1992), and the objective was
to find performance indicators that approximate the “ultimate” criterion as closely as possible.
The ultimate criterion was defined as an indicator of an individual’s total contribution to the goals
of the organization. Unfortunately, no such indicator exists.
The situation began to change during the 1980s. For example, the Army Selection and Clas-
sification Project (Project A) was able to systematically select a sample of entry-level technical jobs
from a population of jobs, develop over 100 separate indicators of performance for each job,
and collect performance data on two cohorts of 10,000 enlisted personnel at three points in time:
at the end of training, at the end of their first tour of duty, and near the end of their second tour of
duty after they had assumed leadership responsibilities (see Campbell & Knapp 2001). This
permitted extensive applications of confirmatory factor analysis to test substantive models of
the latent structure of performance. Subsequently, multidimensional models of performance as
a construct were discussed by Borman & Motowidlo (1993), Campbell et al. (1993), and
Murphy (1989a).
From these sources, a consensus developed that individual job performance should be defined
as things that people actually do, actions they take, that contribute to the organization’s goals.
Someone must identify those actions that are relevant to the organization’s goals and those that are
48 Campbell Wiernik
not, regardless of whether they are in a written job description. For those that are relevant, the level
of proficiency with which the individual performs them must be scaled. Both the judgment of
relevance and the judgment of level of proficiency depend on a specification of the important
substantive content-based goals of the organization, not content-free goals such as “making
a profit,” and there may indeed be multiple goals, goal change, or goal conflict.
Nothing in this definition requires that a set of performance actions be circumscribed by the
term job or that they remain static over a significant length of time. Neither does it require that the
goals of an organization remained fixed or that a particular management cadre is responsible for
determining the organization’s goals (a.k.a. “vision”). Neither does it say that actions, or goals,
must be described at a certain level of specificity. Consequently, it is not a violation of this
definition of performance for individual organization members to decide themselves what actions
are most relevant for what they think the organization’s goals are, or should be. Individuals can be
quite active (Frese 2008) or proactive (Griffin et al. 2007). However, goal choices, and decisions
about what actions best serve them, must be legitimized by the stakeholders empowered to do so
by the organization’s charter. Otherwise, there is no organization. Perhaps the indictment of
“conventional” job analysis (see Pearlman & Sanchez 2010) should be that it does not validly
reflect current and future goals, and the actions that best serve them, because “job analysts” may
not be sufficiently knowledgeable about current and future organizational goals to determine the
appropriate performance actions for a particular work role.
What Performance Is Not

The above specification is intended to distinguish clearly between performance itself and (a) the
determinants of individual differences in performance and (b) the outcomes of performance (a.k.a.
results, goal achievement, the bottom line). It certainly seems the case that the determinants have
received the most research attention in our field. They include such things as individual trait
variables (e.g., cognitive abilities, personality, stable motivational dispositions, physical char-
acteristics and abilities), state variables (e.g., relevant knowledge and skill, attitudes, malleable
motivational states), and situational characteristics (e.g., the reward structure, managerial and
peer leadership), as well as the interactions among them. Campbell et al. (1993) have argued that
all of the above must affect performance by influencing three direct determinants operating in real
time: role-specific knowledge, skill, and choice behavior regarding the direction, intensity, and
duration of effort. The direct determinants totally mediate the effects of everything else. However,
knowledge, skill, and choice behavior are not to be confused with performance itself. Performance
itself is what directly facilitates achieving the organization’s goals. Motowidlo et al. (1997) make
a similar argument.
Performance must also be distinguished from the outcomes of performance if the variance in an
outcome (e.g., sales, stock price, salary) is due to other factors, in addition to the individual’s
performance level. As noted below, it may indeed be possible to develop outcome measures that
are virtually totally under the control of the individual, in which case the outcome measure is
a performance measure. We harp on these issues because the distinctions between performance, its
determinants, and its outcomes are often overlooked, particularly with regard to leadership
(Campbell 2013b). All three are important.
Performance should also not be confused with such indicators as efficiency or productivity,
although they are certainly important. Both imply a ratio of output to input; and while it may be
possible to think of the cost of achieving a certain level of performance, as defined above, that is not
our concern here. Finally, performance is not synonymous with development, attrition, or pro-
motion, although these things are certainly important as well.
www.annualreviews.org Modeling and Assessing Work Performance 49

MODELING THE LATENT STRUCTURE OF PERFORMANCE
The recent literature has produced a number of attempts to model the substantive content domains
of individual performance. These have been cataloged by Koopmans et al. (2011) and include
a variety of models that sometimes use different nomenclatures or represent variables other than
performance, as it is defined here. There are many references to productivity, work quality, work
quantity, in-role performance, extra-role performance, and a number of synonyms for effort,
management, leadership, interpersonal behavior, problem solving, teamwork, adaptability,
communication, emotional control, cooperation, development, creativity, etc. The lack of an
agreed upon nomenclature is striking, and it gives the impression that attempts to explicate the
latent structure of performance are quite helter-skelter. However, we assert that the opposite is
true. If this catalog is purged of terms that have no specific content specifications (e.g., problem
solving, creativity); that refer to metrics, not behavior (e.g., quantity, quality, productivity); or that
refer to prerequisite knowledge, skills, or personality traits, rather than performance itself, there is
considerable agreement, despite different terms being used for the same thing. The development of
this near consensus went something as follows.
Since the 1980s, a number of investigators have suggested models for the latent structure of
performance. Given the population of goal relevant actions or behaviors that an individual could
perform in a work role, can they be represented by a meaningful dimension structure that describes
the major distinguishable components of performance? The assumption here is that the construct
of performance is not unidimensional. Advancing the organization’s goals requires different
categories of individual actions that can be distinguished on the basis of the content of the behavior
that is involved, and it is possible to recover these categories.
Early attempts to identify performance categories were job analysis based (e.g., Brumback &
Vincent 1970, Fleishman & Quaintance 1984) and used various methods to cluster job tasks on
the basis of their content similarity. However, the task content was focused almost exclusively on
what is now termed the technical performance dimension. The major milestones along the way to
what we think is a near consensus about the broader structure of individual performance are
as follows.
Project A
The first major attempt to identify performance factors based on actual performance assessments
was the Project A effort during the late 1980s (see Campbell et al. 2001). That effort produced
a five-factor solution for entry-level Army enlisted personnel and a six-factor solution for non-
commissioned officers (NCOs). On the basis of cross-validation designs and confirmatory factor
analyses, the five-factor and six-factor models were shown to be quite robust across Army
occupations and across cohorts separated by three years. Both models contained one factor specific
to the Army (physical fitness and military bearing), but the remaining factors were more general,
dealing with technical performance, peer leadership, supervisory leadership, extra effort/initiative,
and personal discipline. The level of goodness-of-fit was remarkable, even when cross-validating
from one cohort to the other. The Project A factor structure was also similar to one specified by
Lance and colleagues (1992) using Air Force data.
Campbell et al. (1993) expanded the Project A model to make it more appropriate for non-
military jobs. Their model portrayed the latent structure of performance as composed of eight
factors: job-specific technical proficiency, non-job-specific technical proficiency, communication,
demonstrated effort and initiative, personal discipline, facilitating peer and team performance,
supervision/leadership, and management/administration. These dimensions were defined at a level
50 Campbell Wiernik
of specificity that was fairly general, but specific enough to be useful for descriptive purposes.
There could be more specific subfactors or higher-order factors.
Core Technical and Contextual Performance

Influenced by Project A and subsequent studies, Borman & Motowidlo (1993, 1997) proposed
a model of performance with two general factors, which they labeled core task performance and
contextual performance. Core task performance “consists of the activities that transform raw
materials into goods and services that are the organization’s products,” and contextual perfor-
mance “consists of activities that service and maintain the technical core by replenishing its supply
of raw materials, distributing its finished products, or providing important planning, super-
vising, or staff functions that enable it to function effectively and efficiently” (Motowidlo et al.
1997, p. 75). These definitions were not without ambiguity, much of which was resolved in the
Borman & Motowidlo (1997) paper that proposed five subfactors for contextual performance
described in behavioral terms. Subsequent research has supported both the distinctiveness of
core technical versus contextual performance and the nature of the contextual subfactors (see
Conway 1996).
Organizational Citizenship Behavior

Based on management theory, Organ (1988) proposed organizational citizenship behavior (OCB)
as a major component of performance and defined it as discretionary behavior, not necessarily part
of a job description, that promotes the effective functioning of the organization via being co-
operative, helping other people, tolerating less-than-ideal working conditions, going well beyond
minimal requirements, identifying with the organization’s goals, and participating voluntarily in
organizational governance and administration. There is now a relatively large literature on the
assessment, prediction, and consequences of OCB that is thoroughly reviewed by Organ et al.
(2011). As many as 30 different facets of OCB have been proposed, but six factors seem to reflect
their content (Podsakoff et al. 2000). Further, the relevant parties agree that there is virtually
complete overlap of the six OCB factors with the subfactors of contextual performance (see
Motowidlo 2000, Organ et al. 2011), although there is some mild argument over whose language
is the most interesting.
Counterproductive Work Behavior

The Project A performance model included a factor labeled maintaining personal discipline. It
involved such things as disciplinary actions, violation of regulations, and assessment of personal
misconduct. Its counterpart in civilian life is counterproductive work behavior (CWB), the content
of which has been specified by Bennett & Robinson (2000), Berry et al. (2007), Gruys & Sackett
(2003), and Dalal (2005). The behavioral definition of CWB is “scalable actions and behaviors
that employees engage in that detract from organizational goals or well-being. They include
behaviors that bring about undesirable consequences for the organization or its stakeholders”
(Ones & Dilchert 2013, p. 645). Two primary subfactors have emerged: deviance directed at the
organization (CWB-O; e.g., theft, absence) and deviance directed at other individuals (CWB-I; e.g.,
bullying, gossiping). Although substance abuse is seemingly directed at the self, factor analytically
it clusters with behaviors directed at the organization (Marcus et al. 2002), as the behavioral result
of drug use is typically to avoid work or perform poorly or unsafely. Spector and colleagues (2006)
went further and decomposed CWB into five specific facets, arguing that the facets have different

antecedents. For each of the two primary subfactors, it might also be possible to specify two sub-
facets corresponding to approach versus avoidance behaviors. The approach–avoidance dis-
tinction is a recurring one in motivation (e.g., Gable et al. 2003) and psychopathology (Markon
et al. 2005). For CWB, it is a distinction between actively acting against the organization (property
deviance: sabotage, theft, etc.) versus staying away (production deviance: unexcused absence, drug
abuse, etc.) and between actively acting against other individuals (personal aggression: bullying,
physical abuse, etc.) versus subtly undermining or excluding them (political deviance: gossiping,
showing favoritism, etc.). The CWB-I/CWB-O factors have received support in many factor
analytic studies, though a strong general CWB factor is present (Ones & Dilchert 2013). Evidence
also suggests that CWB are not simply the negative end of positively-stated dimensions such as the
OCB/contextual performance factors (Berry et al. 2007, Miles et al. 2002, Ones & Dilchert 2013,
Spector et al. 2010), but constitute distinguishable separate dimensions that have different
determinants. Measures should avoid including both OCB and CWB items on the same scale
(Spector & Cha 2014).
Proficiency, Adaptability, and Proactivity

Griffin and colleagues (2007) presented a model of work role performance that does not focus
directly on the latent structure of performance behavior. Instead, it posits a 3 3 3 classification of
work role behaviors in which one dimension represents organizational level (individual, team,
organization) and the second dimension progresses from proficiency on prescribed tasks; to
proficiency in adapting to changes in individual, team, or organizational requirements; to being
proactive in instituting new methods or solutions at the individual, team, or organizational level.
Three items (i.e., rating scales) assess proficiency within each of the nine cells. Given the item
content and definitions for each cell, the level dimension seems to represent (a) individual task
performance, (b) peer leadership and support in teams, and (c) certain aspects of the management
role. The proactivity column represents the extra effort/initiative component of several other
models. The adaptivity column represents a much more complex construct and will be discussed
subsequently. In essence, the Griffin et al. (2007) model represents much of the same content as the
models described above, but in a 3 3 3 format rather than a hierarchical one. Their argument is
that it makes the effects of context (i.e., level) clearer.
Competency Models
Competency modeling is an important area of practice in human resource (HR) management
(Shippmann 2010), and it has relevance for the specification and assessment of performance,
particularly with regard to management performance. Unfortunately, there remains some am-
biguity in specifying what a competency is. In our view, there are three choices: A competency
could refer to performance itself, to a direct determinant of performance (e.g., negotiating skill), or
to a more distal indirect determinant of performance (e.g., openness to experience). Shippmann
et al. (2000) seem to allow all three. In an attempt to clarify, Campion et al. (2011) characterize
competencies both as KSAOs (determinants of performance) and as “performance capabilities,”
which also seems to allow all three. Tett et al. (2000) did a content analysis of published com-
petency models, identified 53 competencies, and attempted to define each of the 53, which were
grouped into 10 categories. The 53 competencies did not uniformly represent performance itself.
Some seemed to represent a necessary skill, and others seemed to represent personality charac-
teristics. Also, the processes by which competencies were named and included in a model were not
always very clear.
52 Campbell Wiernik
Stevens (2013) discusses the current state of competency modeling and seems to conclude that
a competency model must include both the important determinants of performance and the
important factors of performance itself because both are necessary to inform selection, training
and development, promotions, job assignment, and compensation. That is, in the best of worlds,
all critical HR systems should be aligned with the same competency model. The real issue is how
expertly the competencies are specified and assessed.
Bartram (2005) tried to restrict a subset of competencies to performance itself and provided
a specification of the “great eight” competencies used by the consulting firm SHL. The great eight
competencies are higher-order factors representing 112 individual scales, some of which could be
construed to represent knowledge or skill, rather than performance. However, the great eight
probably came the closest to making competency synonymous with performance.
The Campbell Revision

Given what has transpired from the late 1980s to the present, Campbell (2012) revised Campbell
et al. (1993) to represent a consensus latent structure that is described as concretely as possible.
That is, the intent was to use as few difficult-to-define abstractions as possible, even though it
makes things sound less exciting.
The eight basic substantive factors of individual performance in a work role are asserted to be
the following:
1. Technical performance: All models acknowledge that virtually all jobs or work roles
have technical performance requirements. Such requirements can vary by substantive
area (driving a vehicle versus analyzing data) and by level of complexity or difficulty
within area (driving a taxi versus driving a jet liner, tabulating sales frequencies versus
modeling institutional investment strategies). As noted by Wisecarver et al. (2007), this
factor should also include, what they term, core interpersonal tasks such as those
involved when dealing with patients, vendors, customers, or community members. A
common term for these tasks is customer service. They are no less technical than
maintaining equipment. The subfactors for this dimension are obviously numerous,
and the domain could be parsed into wide or narrow slices.
2. Communication: The Campbell et al. (1993) model is the only one that isolated
communication as a separate dimension, but it appears as a subfactor in virtually all
others. It refers to the proficiency with which one conveys information that is clear,
understandable, compelling, and well organized. It is defined as being independent of
subject matter expertise, and thus a separate factor, and is not limited to formal
communication. The two major subfactors are oral and written communication, and
their importance can vary widely across work roles.
3. Initiative, persistence, and effort: This factor emerged from the contextual performance
and management performance literatures, as well as the OCB literature, where it was
referred to as conscientious initiative. It was also part of the Project A factor model. To
make this factor conform to the definition of performance used in this article, it must
be composed of observable actions. Consequently, it is typically specified in terms of
working extra hours, voluntarily taking on additional tasks, going beyond prescribed
responsibilities, or working under extreme or adverse conditions. Frese (2008) uses the
term active performance to describe these kinds of actions.
4. Counterproductive work behavior: Consistent with other models, CWB refers to
a category of individual actions or behaviors that are under individual control and

have a negative effect on the goals of the unit or organization. As noted above, there
seems to be general agreement that CWB has two major subfactors: deviance directed at
the organization and deviance directed at other individuals. Also, as explained above, it
seems reasonable to expect an approach–avoidance, or moving toward versus moving
away, distinction for both organizational deviance and individual deviance.
5. Supervisory, managerial, executive (i.e., hierarchical) leadership: This factor refers to
leadership performance in a hierarchical relationship. It also distinguishes between
leadership and management. Leadership refers to the interpersonal influence process.
The substantive content, as specified by the leadership research literature, is most
parsimoniously described by six subfactors (see sidebar titled Six Subfactors Comprising
Leadership Performance; also Campbell 2012), based on the remarkable convergence of
the literature from the Ohio State and Michigan studies through the contingency theories
of Fielder, House, Vroom, and Yetton to the current emphasis on being charismatic and
transformational, leading the team, and operating in highly complex and dynamic
environments. The subfactors describe what leaders do, not the outcomes of perfor-
mance (e.g., effective leader–member exchange, follower satisfaction, unit profitability)
or the determinants (e.g., cognitive ability, personality) of leadership performance or
the situational influences on leader performance. The subfactors are not “styles” or
ephemeral “perceptions.” In a given setting, the relative emphasis across subfactors may
be different, and different leadership models may hypothesize different paths from leader
performance to leader effectiveness (i.e., outcomes), which for some people may be the
interesting part, but the literature’s characterization of leader performance itself (in-
cluding transformational and charismatic leadership) seems always within the
boundaries of these six factors (see Campbell 2013b). Similarly, the six subfactors
circumscribe hierarchical leadership performance at all levels. However, the relative
emphasis on each subfactor may be different at different organizational levels, and the
specific actions within each subfactor may also receive differential emphases. It is also
the case that individuals may react differentially to interpersonal influence attempts by
the “leader.” Leadership is a series of reciprocal processes.
6. Hierarchical management performance: Within a hierarchical organization, this factor,
distinct from leadership as interpersonal influence, includes those actions that deal with
generating, preserving, and allocating the organization’s resources to best achieve its
goals. Given the existing literature, Campbell (2012) argues that there are eight
meaningful subfactors (see sidebar titled Eight Subfactors Comprising Management
Performance). As is true for the components of leadership, there may be considerably
different emphases on the management performance subfactors across work roles and
also as a function of the type of organization, organizational level, changes in the
situational context, or changes in organizational goals. Also, there can be very high
scorers and very low scorers on both the leadership and management subfactors. Very
high scorers on certain critical dimensions are transformational (e.g., Walumbwa &
Wernsing 2013). Very low scorers are dysfunctional (Hogan et al. 2011).
7. Peer/team member leadership performance: The content of this factor is parallel to the
actions that comprise hierarchical leadership (Factor 5 above). The defining character-
istic is that these actions are in the context of peer or team member interrelationships, and
the peer/team relationships in question can be at any organizational level (e.g., pro-
duction teams versus management teams). Many behaviors that comprise the OCB
dimension of personal support (e.g., helping, cooperating, courtesy, motivating) that are
not part of hierarchical leadership also belong here.
54 Campbell Wiernik
8. Peer/team member management performance: A defining characteristic of the high-
performance work team (e.g., Campbell & Kuncel 2001, Goodman et al. 1988) is that
team members perform many management functions, such as planning and problem
solving, determining within-team coordination requirements and workload balance,
and monitoring team performance. In addition, the contextual performance and OCB
literatures both strongly indicate that representing the unit or organization to external
stakeholders and exhibiting commitment and compliance to the policies and procedures
of the organization are critical performance factors at any organizational level. Con-
sequently, to a greater extent than most researchers realize or acknowledge, there are
important elements of management performance in the peer or team context as well as in
the hierarchical setting.
Again, these eight factors are intended to be an integrative synthesis of what the literature has
suggested are the principal content dimensions of performance in a work role. They are meant to
encompass all previous work on individual performance modeling, team member performance,
and leadership and management performance.
Because the different modeling efforts have had different starting points and relied on different
research streams, the degree of similarly across them is remarkable. Consequently, we assert that
at a given level of specificity, the eight factors represent a latent structure for performance that is
essentially invariant across organizational levels (including the team context), functional spe-
cialties, industry sectors, and types of organizations. This invariance does not preclude varying
degrees of importance for the factors, or subfactors, as a function of the specific work role, changes
in goals, or other properties of the context. However, a clear implication is that selection, training,
appraisal, and reward systems should be consistent with this latent structure. It is intended as
a universal competency model of performance.
This latent structure does not preclude higher-order factors (e.g., contextual performance/
OCB) or more specific subfactors (e.g., the taxonomy of detailed work activities in the O NET
SIX SUBFACTORS COMPRISING LEADERSHIP PERFORMANCE

1. Consideration, support, person-centeredness: Providing recognition and encouragement, being supportive when
people are under stress, giving constructive feedback, helping others with difficult tasks, and building networks
with and among others.
2. Initiating structure, guiding, directing: Providing task assignments, explaining work methods, clarifying work
roles, and providing tools, critical knowledge, and technical support.
3. Goal emphasis: Encouraging enthusiasm and commitment for the group/organization goals and emphasizing the
important missions to be accomplished.
4. Empowerment, facilitation: Delegating authority and responsibilities to others, encouraging participation, and
allowing discretion in decision making.
5. Training, coaching: Providing one-on-one coaching and instruction regarding how to accomplish job tasks, how
to interact with other people, and how to deal with obstacles and constraints.
6. Serving as a model: Modeling appropriate behavior regarding interacting with others, acting unselfishly,
working under adverse conditions, reacting to crisis or stress, working to achieve goals, showing confidence
and enthusiasm, and exhibiting principled and ethical behavior.
(Source: Campbell 2012)

EIGHT SUBFACTORS COMPRISING MANAGEMENT PERFORMANCE
1. Decision making, problem solving, and strategic innovation: Making sound and timely decisions about major
goals and strategies and forecasting future trends and formulating strategic and innovative goals (a.k.a. vision) to
take advantage of them.
2. Goal setting, planning, organizing, and budgeting: Formulating operative goals, determining how to use
personnel and resources to accomplish goals, anticipating potential problems, and estimating costs.
3. Coordination: Actively coordinating the work of two or more units or the work of several work groups within
a unit; this includes negotiating and cooperating with other units.
4. Monitoring unit effectiveness: Evaluating progress and effectiveness of units against goals, and monitoring costs
and resource consumption.
5. External representation: Representing the organization to those not in the organization (e.g., customers, clients,
government agencies, nongovernment organizations, the public).

6. Staffing: Procuring and providing for the development of human resources; this does not include one-on-one
coaching, training, or guidance.
7. Administration: Performing day-to-day administrative tasks, documenting actions, and making information
available in a timely manner.
8. Commitment and compliance: Complying with and showing commitment to the policies, procedures, and
directives of the organization, together with providing loyal constructive criticism.
(Source: Campbell 2012)
data model). However, the argument in Campbell (2012) is that aggregating above the eight factor
level loses information. For example, peer leadership and peer management are sufficiently distinct
that aggregating them into “citizenship performance” may introduce additional ambiguity into the
meaning of a particular score. Whether these distinctions can be captured with existing mea-
surement technologies is another matter. Two other issues with which such models must deal are
the existence of a general factor and compound factors.
The General Factor

A general factor does, in fact, exist in virtually all performance indicator covariance matrices,
especially those generated by performance ratings (Viswesvaran et al. 2005). The general factor
can be produced both by a subset of determinants that are common across subfactors (e.g.,
cognitive ability and conscientiousness, or real-time knowledge and skill) and by systematic
measurement errors (e.g., common method variance, common rater bias, implicit performance
models). However, the general factor is not a single latent variable and has never been defined
as such (Campbell 2013a). It is a formative construct, rather than a reflective construct
(Diamantopoulos et al. 2008). That is, the general factor does not reflect a single underlying latent
variable that can be specified. It must always be “formed” as a sum-score of different components.
For decision purposes (e.g., promote/not promote), an overall score must be obtained by weighting
and combining the components (see Borsboom et al. 2003 for further discussion of this issue).
Given an empirical general factor, there is still considerable differential performance and dif-
ferential predictability across subfactors, and the training and development considerations are
vastly different.
56 Campbell Wiernik
Compound Performance Domains
The eight factors are somewhat analogous to the Big Five dimensions of personality. In both
domains, higher-order factors with less informational content and specific facets with more in-
formation are present above and below the designated level of specificity. In the same manner, just
as there are compound traits in personality that represent combinations of basic traits, researchers
have also proposed compound performance dimensions that contain meaningful sets of work
behaviors from several of the eight factors that share a common goal (e.g., promoting environ-
mental sustainability, Ones & Dilchert 2012; or enhancing information technology, Taylor &
Todd 1995). These compound performance dimensions cut across several dimensions. For ex-
ample, environmental performance includes technical behaviors (e.g., installing solar panels), peer
support behaviors (e.g., encouraging others to recycle), and counterproductive behaviors (e.g.,
failing to follow waste disposal procedures). It is meaningful to study these behaviors as a set
because they are relevant for a common goal and share common antecedents and outcomes.
However, these compound dimensions should not be seen as somehow separate from the proposed
latent structure of performance.
CONTENT VERSUS PROCESS IN CONCEPTUALIZING PERFORMANCE

When attempting to summarize efforts to model work performance, it is important to distinguish
between specifications for performance content and specifications for the performance process
(i.e., the context in which performance takes places and the manner in which performance
develops and changes over time). So far, this article has centered on a model of performance
content, about which there is virtually a consensus. There is a parallel universe that addresses the
parameters of performance dynamics, including such concepts as active, proactive, and adaptive
performance. These two areas of inquiry, the latent structure of performance content and the
nature of performance dynamics, are not in competition. They address different issues, but each
has important implications for the other.
Performance Dynamics
No one seriously argues that individual work performance does not change over time, either
because the performance requirements change and individuals respond or because individuals
change even when performance requirements do not. Much of this literature is reviewed by
Sonnentag & Frese (2012). There are at least three aspects of performance requirements that could
change: (a) the behavioral and/or cognitive content of the requirements, (b) the level of perfor-
mance expected, and (c) the conditions under which a particular level of performance is expected
(or some combination of these). If there are interactive effects between individuals and the nature of
the work role content changes, then changes in the rank ordering of people over time result from
multiple sources. Given the current and future nature of employment, it is reasonable to expect that
such things will happen and are happening. Change is complicated.
Much of organizational psychology and organizational behavior (OP/OB) research and
practice deals with planned interventions designed to change performance by enhancing the in-
dividual knowledge, skill, and motivational determinants of performance, such as training and
development, goal setting, feedback, incentives of various kinds, supervision, and so on. Such
interventions, with performance requirements held constant, could increase the group mean, have
differential effects across people, or both. The aptitude–treatment interaction is always with us.
The performance changes produced can be sizable (e.g., Katzell & Guzzo 1983, Locke & Latham

2002). Interventions designed to enhance individual performance determinants can also be
implemented by the individual’s own processes of self-management and regulation (Lord et al.
2010). That is, as a result of an individual’s self-monitoring and self-evaluation of his or her own
performance against goals, additional training can be sought (perhaps from coworkers), different
performance goals can be self-set, feedback can be sought, and self-efficacy can change. The
effectiveness of these self-regulation processes can vary widely across individuals. In addition, if
they have the latitude to do so, people can conduct their own job redesign (i.e., change the be-
havioral content of their work roles) to better utilize their knowledge and skills and increase the
effort they are willing to spend. Academics are fond of doing that.
As noted by Sonnentag & Frese (2012), individual performance can also change simply as
a function of the passage of time. Of course, time is a surrogate for such things as practice and
experience, the aging process, and changes in affective or emotional states (Beal et al. 2005).
It is most likely the case that for any given individual over any given period of time, many of
these sources of performance change are operating simultaneously. Performance dynamics are
complex, and attempts to model the complexity have taken many forms. For example, there could
be characteristic growth curves for occupations (Murphy 1989b), differential growth curves
across individuals (Hofmann et al. 1992, Stewart & Nandkeolyar 2007, Zyphur et al. 2008), both
linear and nonlinear components for growth curves (Deadrick et al. 1997, Sturman 2003), and
cyclical changes resulting from a number of self-regulatory mechanisms (Lord et al. 2010).
Empirical demonstrations of each of these have been established.
A very recent, and very thorough, review of theory and research pertaining to within-person
dynamics, and their antecedents, is provided by Dalal et al. (2014). The authors outline the impli-
cations of within-person variability for both selection and performance assessment. Their message,
and the message here, is that performance dynamics are most likely dimension specific. That is, the
likelihood and nature of changes in the behavioral content, difficulty level, and situational parameters
of performance most likely differ across performance dimensions. Similarly, the determinants of
individuals being able to cope with (i.e., adapt to) such dynamics can also differ across performance
dimensions (Pulakos et al. 2006). Thinking of performance as one thing is counterproductive.
Performance Adaptability
The concept of adaptability, as a particular kind of performance dynamic, has taken on many
meanings in the literature. For a very broad and thorough review of adaptability as a construct, the
reader is directed to Baard et al. (2014) and Chan (2014). For the purposes of this article,
adaptability refers to being able to deal effectively with some combination of the following:
changes in organization goals, changes in individual performance requirements, and changes in the
performance environment, which have already been identified or are anticipated. Adaptability can
be viewed either as a component of performance itself or as a property of the individual (i.e., a
determinant of performance). Ployhart & Bliese (2006) present a discussion of this issue and argue
that it is probably more useful to identify the characteristics of the adaptive individual than it is to
propose adaptability as a distinct content dimension of performance. One reason is that the general
definition of adaptability is not content domain specific, and it has been difficult to provide
specifications for adaptability as a performance dimension. The best attempt to do so is by Pulakos
et al. (2000), who proposed eight adaptability performance factors, which they regard as part of the
latent structure of performance. The factors were obtained by mining a large database of critical
incidents of effective and ineffective performance and using systematic SME (subject matter
expert) judgments to identify and categorize the incidents that were reflective of adaptation.
Consequently, these authors disagree with the Ployhart & Bliese (2006) position.
58 Campbell Wiernik
However, the interpretation of the adaptability factors is not entirely straightforward. Some of
them seem to represent general skills (handling stress, solving problems creatively), whereas others
could be construed as specific subfactors of performance (handling particular kinds of emergen-
cies). Part of the difficulty is being clear about the distinction between the direct knowledge, skill,
and motivational determinants of performance (e.g., knowing how to handle stress), the context in
which performance takes place (e.g., stressful situations), and performance itself (e.g., managing
negative emotional displays). Research on each is of great value. For example, Ployhart & Bliese
(2006) developed a measure, based on the Pulakos et al. (2000) factors, of self-assessed skills,
interests, and response tendencies. Pulakos et al. (2002) also developed a self-report measure of
prior experiences, skill levels, and interests relative to the eight adaptability performance factors
they proposed. The two instruments both assess performance determinants and should show
considerable convergent validity.

It would also be useful to first sort the original sample of critical incidents used by Pulakos et al.
(2000) into performance dimensions and then identify those that reflect adaptability. Using this
framework, the adaptive incidents should reflect exceptionally high performance on each di-
mension. Again, such a framework regards adaptability as dimension specific and not as separate
components of the latent structure of performance.
In sum, work role performance requirements can change over time, sometimes over very short
periods of time, as a result of many factors. The message in this article is that the latent structure of
individual work performance is multidimensional, and the eight factors discussed above represent
a consensus developed over several decades. In terms of investigating such things as (a) the nature
of adaptive performance, (b) the determinants of adaptive performance, (c) the correlation of past
performance with future performance over time (and the reasons that it increases or decreases), (d)
characteristic performance growth curves for occupations or individuals, and (e) the nature of
performance changes across situations and contexts and over time, the research should be di-
mension specific. Performance is not one thing, and neither is adaptability. Consequently, any
procedure for assessing performance must be clear about the behavioral content of what is to be
assessed and also about what performance dynamics are to be accounted for.
PERFORMANCE ASSESSMENT
The structure and dynamics of performance are complex, which makes assessment of performance
a very difficult enterprise. Much of the difficulty results from the necessity of defining per-
formance as things that people actually do. This specification rules out using existing outcomes
(e.g., sales, defects, ROI) as performance criteria if a significant portion of their variance is not
controlled by the individual. Now, it is also true that work roles are designed or invented to
accomplish organizational goals and influence the bottom line (see Ployhart & Hale 2014), and
the causal path is a complicated and interesting one. However, the basic tenet here is that the
individual should not be held accountable for outcome determinants over which he or she has
no control.
Purposes for Performance Assessment

It is well recognized that performance can be assessed for one or more of several distinct reasons
and that the measurement purpose has a substantial influence on the measurement format, the
results of the assessment, and the effects of the assessment on subsequent performance (Murphy &
Cleveland 1995). Consider just the following potential reasons for performance assessment.

For research purposes. The objective here is to estimate the individual’s “true score” on one or
more dimensions of performance to evaluate the validity of selection procedures, assess the efficacy
of interventions intended to improve performance, or test a variety of theoretically derived hy-
potheses that feature performance as a key variable. For these purposes, the construct validity of
the measure—that is, whether it assesses components of the model described above—is the prime
concern. Reliability is also important, but construct validity and reliability are often put at risk
because research is costly and not well supported in our field and because thorough performance
assessment does not have a high priority in many research studies.
For legal defense of personnel decisions. The objective here is to support the validity of specific
personnel decisions for meeting certain legal requirements. Similar to research purposes, the
primary concern for these performance assessments is construct validity and measurement re-
liability, with reliability being particularly important. Given the potential for serious legal and
financial consequences of poor validity and reliability documentation, appropriate measurement
design is more common in these situations than when assessment is purely for research purposes.
For high-stakes appraisal. The high stakes referred to here are promotion, dismissal, reassign-
ment, and compensation decisions. Both the individual and the organization have vital interests in
the results of the assessment and could be expected to pay considerable attention to their respective
interests. Considerations of fairness, transparency, accuracy, appraisal goals, and motivation (of
both the appraiser and appraisee) become paramount (DeNisi & Pritchard 2006, DeNisi &
Sonesh 2011, Murphy & Dechert 2013).
For performance feedback and development. Assuming that high-stakes decisions are not directly
involved, the emphasis here is on identifying specific, substantive performance behaviors that need
enhancement or improvement. Building on Kluger & DeNisi (1996), the feedback literature
stipulates that, for developmental purposes, feedback should be very concrete and very specific to
the individual’s performance. Assessment of an individual on the general factors or subfactors
discussed in the previous section would not be useful, unless accompanied by more within-factor
specifics. Also, to the greatest extent possible, the feedback process should avoid overall evalu-
ations of the appraisee or even overall evaluations on particular dimensions, such as technical,
communication, or peer leadership performance. Such evaluations risk diverting attention away
from specific performance improvements to affective self-evaluations.
For self-managed performance improvement. Here, the performance goals are self-set, but the
requirements for feedback on specific actions and avoidance of general evaluations are the same.
Individuals must conduct their self-appraisal accordingly.
Each of the above purposes are quite different, but all of them should be guided by the basic
requirement that the assessment must consider all of the eight performance factors, at some level of
specificity, to be a comprehensive measure of performance.
General Measurement Issues

Any attempt to assess work performance must also consider the following general measurement
issues.
Typical versus maximum performance. This distinction refers to the direction, amplitude, and
duration of effort focused on task accomplishment. Are they at the levels typically exhibited by
60 Campbell Wiernik
individuals in their work settings, or do they reflect specific conditions that keep attention highly
focused, increase effort levels, and maintain higher effort levels for some period of time (DuBois
et al. 1993)? Performance under such conditions is designated as maximum performance. If the
goal is to assess maximum performance, which may be appropriate for assessing performance
capabilities under certain critical conditions (Mangos et al. 2007), then the assessment method
must account for the appropriate motivational conditions. However, Dalal et al. (2014) argue that
besides greater effort, requirements for maximum performance can also bring additional abilities
into play. Assessment must also account for the fact that in any given work role, different per-
formance requirements may have different priorities (Mangos & Arnold 2008) and that requiring
greater effort may create unusual pressures that degrade performance on complex tasks (Klehe &
Anderson 2007). Also, some performance requirements simply might not be amenable to increases
in effort (Sackett 2007), such as those that incorporate the leadership and management factors
discussed above. Beus & Whitman (2012) carried out a meta-analysis of the typical/maximum
performance literature. The estimated correlation (corrected) between them was .42. Differential
prediction of typical versus maximum performance and variables that moderated the relationship
between them were also examined. Under certain conditions, the correlation between them can be
quite high (e.g., Deadrick & Gardner 2008).
Performance dynamics. As discussed above, performance is not static. It can change because of
changes in performance requirements; changes in the individual because of training, goal setting,
motivational interventions, affective states, aging, etc.; or changes in situational conditions, such as
constraints or opportunities created by coworkers or production practices (Stewart & Nandkeolyar
2007). The assessment method must take these dynamics into account, if necessary, to achieve the
measurement purpose. In general, this must be done either by repeated measurements (e.g., Stokes
et al. 2010) or by capturing summary judgments of performance change over time. Again, assessment
of performance change should be dimension specific. For example, the dynamics of technical
performance and team/peer leadership performance are most likely different.
Cross-cultural performance assessment. As described above, Campbell’s (2012) eight factors are
presented as a general model of the latent structure of job performance that is universal across jobs,
organizations, industries, and levels. In addition, we believe that it is universally applicable across
cultures. Although the relationships among factors and the relative importance of factors may
differ across jobs, we believe that these clusters of work tasks exist to some degree in all jobs around
the world. With this point in mind, it is important to understand how specific manifestations of
and relationships among performance factors differ across cultures, as well as how assessment
practices differ in various contexts. Empirical studies of the structure of performance in non-
American, non-European contexts have only recently been undertaken. For example, He (2012),
Rotundo & Xie (2013), and Xu et al. (2013) all examined the structure of CWB in Chinese
organizations. Although some dimensions manifest differently in China than in Western cultures
(e.g., interpersonal aggression was expressed primarily through indirect and political behaviors,
rather than through confrontation), in general, the observed structures were remarkably similar to
findings from American and European samples. More studies of this nature, especially studies
examining the lower-order structure of other dimensions of performance and the relationships
among the eight factors in new cultural contexts, are needed.
Research examining cultural differences in performance assessment processes is also relatively
new. Festing et al. (2012) provide an overview of comparative studies of performance appraisal
practices and considerations that must be made when designing performance evaluation systems
across cultural contexts, such as the influence of unique cultural values on evaluation practices. For

example, common Chinese values of modesty, leniency, and guanxi (social harmony and con-
sideration) may reduce the objectivity of performance ratings, as both the assessor and assessee use
the evaluation process to serve other goals (Barron & Sackett 2008). The most comprehensive
examination of international differences in performance appraisal practices was conducted by
Peretz & Fried (2012). These authors found substantial variability in the structure and formal-
ization of performance appraisal across the large number of organizations from 21 countries
studied. This variability was strongly related to cultural characteristics, such as power distance,
collectivism, and future orientation. Additionally, within a country, organizations that had
normative performance assessment practices experienced less absenteeism and turnover than did
organizations with culturally divergent practices. These results further reinforce the conclusion
that performance assessment systems designed for cross-cultural use need to accommodate local
practices and preferences to fulfill their purposes.

The distributional properties of performance. Recently, O’Boyle & Aguinis (2012) raised the
issue of the distributional properties of individual performance and their implications for per-
formance assessment, prediction, and management. Based on five examples (faculty publication
counts, entertainment industry awards, frequency of being elected to state legislatures, and both
positive and negative sports performance indicators), they argue that a Pareto distribution fits the
data far better than a normal distribution, thus calling a number of statistical estimation methods
(e.g., multiple regression) into question. Further, assessment via ratings is inappropriate, in their
opinion, because most applications of rating methodologies attempt to force a normal distri-
bution. Their prescriptions are to assess outcomes, not behavior, and to achieve greater differ-
entiation among the highest “elite” performers, if possible. It is the highest performers that
facilitate organizational success. These themes are amplified in Aguinis & O’Boyle (2014). Beck
et al. (2014) have taken issue with the O’Boyle & Aguinis assertions. They argue that seven
conditions must be satisfied before a data distribution can be called a performance distribution.
They then analyze several representative data sets, including indicators of sports performance, that
meet the requirements, and show that a normal model does fit the data. Their conclusion is that the
O’Boyle & Aguinis results are largely the result of statistical artifacts.
Given the distinction we make between performance and its outcomes, perhaps the most salient
point here is that a particular outcome distribution could be quite skewed, but the underlying
performance distribution can approach normality. For example, only a few golf professionals ever
win a major tournament. Most do not (a skewed distribution). However, the distribution of actual
scores (performance) is much more symmetrical. Parametric methods should probably not be used
with highly skewed outcome distributions. Also, the relationship of performance to the utility of
outcomes need not be linear and could take many forms. Finally, individual performance may not
be the only determinant of the outcome distribution, and the reliability of the outcome indicator
itself is an issue. For example, the outcomes of mutual fund managers appear to have zero year-to-
year reliability (Carhart 1997, Sauer 1997).
Performance Assessment Methods

There are a number of methods that attempt to assess individual work performance, and each
has its strengths and weaknesses. There is no ultimate criterion or even one best way. In this
section, we consider the applicability of different methods for different assessment purposes and
the ability of each method to assess performance itself, as well as deal with the measurement issues
described above.
62 Campbell Wiernik
Ratings. Performance ratings by supervisors, peers, subordinates, or by oneself are ubiquitous,
and this literature has been reviewed extensively (e.g., DeNisi & Sonesh 2011, Levy & Williams
2004, Murphy & Dechert 2013, Woehr & Roch 2012). We highlight only the major issues here.
The overriding issue is whether ratings have construct validity for the purpose(s) for which they
are to be used. That is, are individual differences in rating scores reflective of individual differences
in performance itself? Fundamental to judgments of construct validity are the specifications for
what is being measured. For any given situation, we must specify what performance is, and what it
is not, as concretely and completely as possible. This includes considerations of the dynamics to be
considered (e.g., being adaptive), the time interval of interest, and the typical versus maximum
performance requirements, in addition to the substantive content of the latent structure. Without
such specifications, construct valid assessment is problematic.
The kinds of evidence supporting construct validity are: the correspondence between the
performance specifications and the instructions given to raters (including the rating format); the
operative goals of the rater (i.e., do they correspond to the measurement purpose?); the rater’s level
of knowledge about what is to be rated; the level of interrater agreement and interrater reliability;
the correlations of ratings with performance assessments using alternative methods; the pattern of
correlations with other variables; the existence, or nonexistence, of rater biases (halo, leniency,
central tendency); and the degree to which ratings are contaminated, or not contaminated, by
particular variables (e.g., rater/ratee gender or race, opportunity to observe, the rater’s own
performance level, rater accountability, the rater’s interpersonal relationship with the ratee, rater
personality, level of rater effort, the rater’s own implicit performance model versus the one
prescribed by rating instructions and format, and the impression management strategies of the
ratee). All of these have received varying degrees of research attention.
One of the most critical factors is the goal of the rater during high-stakes appraisal (DeNisi &
Sonesh 2011, Kozlowski et al. 1998, Levy & Williams 2004, Spence & Keeping 2010). The goal of
assessing the ratee’s true performance has frequently been shown to be less important than al-
ternative rater goals, such as rewarding or punishing the ratee, conforming to organizational
expectations, or advancing the rater’s self-interests. Consequently, it could be argued that ratings
should not be used as high-stakes appraisals unless the goals and motivation of the raters can be
aligned with the goals of accuracy, fairness, and transparency (Pulakos & O’Leary 2011). In
general, high-stakes appraisals do not work very well as research criteria, and their construct
validity as assessments of performance itself is suspect (Murphy & Dechert 2013).
The use of 360 ratings for feedback and management development purposes is widespread.
Most 360 systems are based on competency models that specify the capabilities that a high-
functioning manager/executive should have (DeNisi & Kluger 2000). As already noted, com-
petencies can reflect personality characteristics, motivational tendencies, knowledge, skills,
dimensions of performance itself, or even outcome measures (e.g., “achieves results”). In practice,
they tend to lack substantive specifications, which makes using them for feedback purposes difficult
(Campion et al. 2011, Tett et al. 2000), and competency ratings tend to have low interrater
agreement. Sanchez & Levine (2009) argue that competency models function better as general
goals than as a means for individual assessment.
For ratings as research criteria, perhaps the most contentious argument concerns whether
interrater correlations represent the reliability of performance ratings. Murphy (2008) and
Murphy & DeShon (2000) argue that they do not, and these authors call the construct validity of
ratings into question. Ones et al. (2008) and Schmidt et al. (2000) contend that interrater cor-
relations are the appropriate estimator of performance ratings reliability in most samples used in
organizational research. For them, the construct validity of the ratings is a separate issue. Putka &
Hoffman (2014) agree with Schmidt and colleagues (2000) for a specific set of conditions but also

elaborate on how characteristics of the measurement design (e.g., fully crossed, nested, partially
nested, ill structured), the research questions being asked (e.g., predicting the rank ordering of
ratees versus assessing their actual performance level), and the kinds of generalizations desired
(e.g., generalizing across raters or generalizing across dimensions) influence the appropriate choice
of the reliability estimator. The choice of estimator can influence the value of the reliability es-
timate. In general, misspecifying the reliability estimator can underestimate reliabilities to varying
degrees, but the discrepancies become substantial only in very ill-structured designs. For most (but
certainly not all) research designs that have been reported, estimating reliability using the interrater
correlation, rather than proposed alternatives that explicitly estimate additional sources of
variance, is not likely to yield seriously biased estimates. However, given at least modest reli-
abilities, the basic question is still whether supervisor and peer ratings reflect individual differences
in performance or individual differences in ratees, raters, or their interactions, which are unrelated
to performance itself.
There is not space to review the construct validity evidence in detail, but at least the following
points are relevant. Ratings do tend to exhibit considerable halo and leniency effects. Whether this is
error or valid variance is another matter. However, ratings for research purposes tend to exhibit less
rater bias than do ratings for high-stakes purposes. As a cognitive decision-making process, ratings are
susceptible to primacy and recency effects and a strong tendency to make judgments quickly using
shortcut heuristics (Fisher 2008). Raters must process a lot of information, much of it from memory.
Consequently, interrater reliability is higher for nonmanagerial and less complex jobs, where per-
formance is more readily observable and concretely specified (Conway & Huffcutt 1997).
Rater source effects have been studied relatively extensively. In general, self-assessments ex-
hibit greater leniency, less accuracy, and lower correlations with other variables than do other
sources (Dunning et al. 2004, Heidemeier & Moser 2009). They seem to have little construct
validity. Source effects for supervisors, peers, and subordinates have been reviewed by Hoffman
et al. (2010), Hoffman & Woehr (2009), and Lance et al. (2008). The summary conclusions are
that although rater source differences are not large, they also do not represent error. Different
raters can have different perspectives. Consequently, rater intercorrelations, to some small degree,
can be underestimates of rater reliability. However, different sources do not seem to produce
different factor structures (Campbell & Knapp 2001, Facteau & Craig 2001). Several researchers
have also pointed out the shortcomings of the multitrait-multimethod matrix as a way to model
rating variance components when dealing with incomplete designs. Putka & Hoffman (2014)
suggest alternatives.
Finally, a great deal of evidence shows that ratings have consistent correlations with other
variables (e.g., cognitive ability and personality) and that there are meaningful differential cor-
relations of such variables with ratings of different performance dimensions (e.g., Organ et al.
2011). The research on rater training, particularly frame-of-reference training, shows that such
training significantly improves the construct validity of ratings (Noonan & Sulsky 2001,
Schleicher et al. 2002), which would not be expected if ratings did not assess performance itself. A
recent study by Hoffman et al. (2012) evaluated the use of frame-of-reference scales (FORS), which
attempt to provide more complete specifications for the dimensions to be rated, and showed them
to have greater construct validity and accuracy than traditional scales. Also, a meta-analysis by
Bommer et al. (1995) suggests that although the overall correlations between ratings and alternative
“objective” measures of performance are relatively low, when the performance components being
assessed are similar, the intercorrelations are higher. This was supported in a comprehensive
multimethod study of jet engine mechanic performance (Vance et al. 1988). There is also a modest
literature on the relationship of assessment center ratings and performance ratings obtained later
(Hermelin et al. 2007). The correlations are reasonably high, even though assessment center
64 Campbell Wiernik
ratings and later performance ratings reflect the maximum versus typical performance distinction
(i.e., assessment centers are designed to elicit maximum performance).
In our judgment, the construct validity of performance ratings is relatively substantial, even
though the performance construct is poorly specified in many studies and rater biases do exist.
Construct validity would be enhanced further to the extent that performance is concretely spec-
ified; the specifications are incorporated in rater training and the rating instruments; and raters (a)
have observed the ratee extensively, (b) accept the rating goal and the performance specifications,
(c) understand the rating instrumentation, (d) know they are accountable for rating accuracy, (e)
have ample time, and (f) are sensitive to such contaminants as liking for the ratee. These conditions
are probably not met in many data collections.
Samples, simulations, and proxies. The use of work samples and simulations as criterion mea-
sures has a long history in applied psychology. For example, performance on work samples con-
stituted a large amount of the data used in Project A to develop the enlisted and NCO performance
models. Hunter (1983) discussed the relative construct validity of ratings and work samples as
measures of job performance, and Howard (1983) proposed the use of work samples and sim-
ulations to evaluate training outcomes. Distinctions are frequently drawn between work samples,
in which an individual performs an actual job task using real job materials (e.g., fixing a real
engine, processing real client emails), and simulations, in which individuals perform tasks in
fabricated situations or with facsimiles of task materials (e.g., driving using a video simulator, role
playing a conflict negotiation). Although there are conceptual differences between these forms of
assessment, their use as measures of performance is based on the same logic, and, in most cases, the
choice of one over the other is one of practicality, rather than conceptual choice. Also in this
category of measures are behavioral proxies that attempt to elicit the same performance responses
as actual or simulated work tasks, but that do not closely mimic actual job tasks or situations. The
primary examples of this kind are assessment center exercises, which are increasingly being used
for developmental (Rupp et al. 2006) and even performance evaluation purposes (Riggio et al.
2003). These three methods share most of the same strengths and weaknesses. Consequently, for
the purposes of this article, we refer to all of them as simulations.
One of the key advantages of simulations over other measures of job performance is their ability
toassess employees’ capabilities for performing critical tasks that are otherwise difficult, unethical, or
impossible to assess with any frequency. For this purpose, the most sophisticated simulations have
been developed for training and evaluating individuals in medical and related professions (Kunkler
2006) and military contexts (e.g., Colegrove & Bennett 2006). These simulations have seen the most
use in training situations, where the level of detail facilitates providing feedback on specific behaviors.
Simulations are also useful for measuring potential performance in emergency situations
(i.e., performing adaptively), such as emergency landings for pilots or crash avoidance for drivers.
Although simulations for other forms of performance have been developed (e.g., for management,
Halpin & Biggs 2009; teamwork, Heinrichs et al. 2008; communication, O’Neil et al. 1997;
leadership performance, Thomas et al. 2001), they are not currently as sophisticated.
Compared with ratings, simulations have the advantage of being potentially more valid
assessments of employees’ ability to perform at a particular level of proficiency, and they are
assumed to be free from the contamination issues of performance ratings. However, these
measures can suffer from construct validity issues that are as serious as those faced by ratings. The
primary threat is one of construct deficiency. Simulations can be expensive and time consuming to
develop and administer. As a result, they typically assess only a few (or one) critical job tasks,
typically representing some form of technical performance. To the extent that other factors of
performance are important (as they are to some degree for all jobs), simulations lack construct

validity. Simulations can also suffer from criterion contamination if the elicited behaviors do not
accurately reflect real performance on the job with fidelity (Lievens & Patterson 2011). For
example, if the controls of a driving simulation device respond differently than an actual vehicle,
simulator performance is not an accurate measure of actual job performance. Finally, if simu-
lations are scored using observer ratings, they can suffer from the same perceptual and evaluative
biases (though not the sampling biases) as other ratings-based measures of job performance.
The issue of typical versus maximum performance is especially salient for the use of simulations
as performance measures. Individuals completing a simulation are likely to perform at their
maximum capacity because simulations measure performance within a short period of time,
leaving little opportunity for the motivational and self-control processes that decrease perfor-
mance from maximum to typical levels to take place. Even when individuals perform at satis-
factory levels in a simulation evaluation, they may not demonstrate this level of performance
consistently on the job. As a result, simulations may be more useful for developmental purposes
than for high-stakes decision making, unless the behaviors being assessed are always likely to elicit
maximum effort (e.g., emergency responses). On the plus side, requirements for adaptive
responses can be built into a simulation.
Technology-enhanced assessment. At an increasing rate, popular press and management and

business practice publications are drawing attention to the potential uses of advanced technologies
and large amounts of diverse, rapidly generated data (so-called “big data”) to improve business
practices (Lohr 2013, McAfee & Brynjolfsson 2012). Of particular relevance to this article are
suggestions that such technologies can enhance or replace other forms of performance assessment
(Hunt 2011). Despite the enthusiasm for these new technologies, caution is warranted, as all too
often these systems measure outcomes, such as sales volume, rather than performance itself
(Cravens et al. 1993). Technology-based performance assessments can be useful, but only if they
measure performance that is under individual control. For example, onboard computer systems
have been used to track delivery trucks and to assess safe driving behaviors and compliance with
delivery protocols (Kargupta et al. 2010, Şimşek et al. 2013). Similarly, electronic recordings of
emails and phone conversations can be used to assess call center, customer service, and com-
munication performance, especially when these techniques are combined with audio transcription
and text mining software to reduce the need for supervisors to evaluate each communication
(Miller 2003). Rapidly delivered data also present new opportunities to provide immediate
feedback to employees, such as by presenting employees with a daily or hourly scorecard of a
relevant performance metric (e.g., error rate, change in customer numbers) or using wearable
technologies that sound alerts when unsafe movements are made. Again, so long as the data
provided by these tools are under employee control, they constitute a potentially useful measure of
individual performance. However, big data can suffer from the same criterion deficiency and
construct validity problems as simulations. On the plus side, with big data it is potentially possible
to capture performance dynamics as they occur naturally.
Additional potential problems with using technology-based monitoring systems to provide
feedback are concerns about invasions of privacy and feelings of dehumanization (Hunt 2011,
Miller 2003). Care must be taken to ensure employee acceptance of the technology as a legitimate
source of performance information.
Goal Achievement and Performance Outcomes

Throughout this article, we have emphasized that measures of performance must assess behavior
that is under individual control, not more distal performance outcomes. However, in practice,
66 Campbell Wiernik
many organizations simply want to use outcome measures as indicators of performance. Also,
describing performance in terms of attaining mutually set or accepted goals can increase goal
achievement and the perceived value of the evaluation process (Locke & Latham 2002). As we
have stated before, these indicators constitute performance measures only so long as factors
outside of the individual’s control are substantially removed from consideration. Pulakos &
O’Leary (2010) discuss ways in which that can be done. Sales figures may be an appropriate
performance indicator when they reflect only differential levels of effort or skill, such as for call
center employees requesting donations from telephone numbers assigned at random, or when
environmental and task difficulty factors can be controlled for, such as by referencing a particular
employee’s sales against the norms for economically similar areas. Similar considerations apply
for other jobs, including manufacturing, management, and executive jobs. However, there is again
the problem of criterion deficiency, and critical parts of nontechnical performance dimensions
(e.g., peer leadership and management) may go unassessed.

An unfortunate example of a poorly chosen outcome measure, which has received substantial
media and policy attention, is the attempt to use changes in standardized achievement test scores of
public school students to assess individual teacher performance. These are the so-called value-
added models, which attempt to assess student test score gains as indicators of learning in a spe-
cific teacher’s classroom and hold the individual teacher responsible, even though the year-to-year
assignment of students to teachers is not random and controlling for classroom differences via
statistical covariates is highly problematic. These issues have been thoroughly discussed by Haertel
(2013), who concludes that the gain scores are saturated with error and irrelevant variance and are
not assessments of teacher performance. We agree, as do other measurement professionals and
teachers themselves (e.g., Mueller 2011).
In sum, when appropriately chosen, outcome measures are useful primarily for high-stakes
decision making. They offer value for feedback and developmental purposes only so far as they
provide information on the behavioral changes necessary to improve performance.
CONCLUSIONS AND RECOMMENDATIONS

Between 1980 and today, a near consensus about what performance is has emerged. Performance
is not unidimensional and, strictly speaking, should not be used in the singular (e.g., is he/she a high
performer?). It is composed of all the individual actions that support or detract from the organi-
zation’s goals to varying degrees. At a particular level of generality/specificity there is also general
agreement about what the major components of job performance content are, although many of
us will be reluctant to give up our own labels for them. For example, the romance of OCB is
a powerful thing for many. There is also a consensus that individual performance is dynamic, for
many different reasons that can be specified, and its dynamic features are most likely different for
different performance components. As OP/OB research continues, researchers should situate their
work within the well-understood space of the latent structure of performance, rather than attempt
to declare that each new construct is wholly distinct from what has come before. Cumulative
science demands that future research build upon previously generated knowledge, not disregard it
in favor of exciting new terminology.
Despite widespread acceptance of the definition of performance as what the individual actually
does, not determinants or outcomes, all too often researchers continue to conflate the three, es-
pecially in the areas of teamwork (e.g., DeChurch & Mesmer-Magnus 2010) and leadership (e.g.,
Lord & Dinh 2014). For meaningful scientific communication to take place, clarity and precision
in language are necessary. Performance should be specified in behavioral terms as things that
people do. This is not a behaviorist statement.

Performance in a work role is a complex phenomenon, which makes assessment difficult. There
is no way to make it simple. Regardless of whether the measurement method consists of ratings,
simulations, outcomes under the control of the individual, or big-data capture, the information
obtained must correspond to the specifications for what performance is. The consensus model
described above is intended to serve as a basic starting point for all performance assessment, and
this includes considerations of individual performance trends and responses to changing require-
ments and goals.
In this regard, each of the named methods has strengths and weaknesses. In our view, the
primary needs for future research and development are the following:
1. How can we build on existing rater training methods to better teach raters what
performance is, what goals they must have, what they must know about the person being
rated, what dynamics (e.g., specific adaptive responses) they should try to account for,
what time interval is of interest, and what potential contaminants of their ratings they
must manage (not just halo and leniency effects)? Going further, a widely available
MOOC (massively open online course) dealing with these issues could potentially
benefit many sectors, perhaps even society at large. The need for judgments of one
person’s performance by others will not go away. Big data will not replace it. We simply
must do it better.
2. How, and for what reasons, do raters actually make rating judgments? What informa-
tion do they use? How do they combine it? What are their operative goals? We need
many more protocol analysis studies examining such questions for each of the major
rating purposes. This means sitting beside someone, perhaps virtually, and having them
talk through what they are doing. It is different than studying ratings as a cognitive
process, valuable though that is. Protocol analysis is used extensively in the study of
expertise (e.g., Hoffman & Militello 2009), why not here?
3. How can online performance rating forms for research purposes be structured and
delivered to avoid careless ratings and to instill the same feelings of value, re-
sponsibility, and attention that can be obtained when ratings are completed in the
physical presence of a researcher or supervisor? Collecting ratings in person is time
consuming and expensive. How can these drawbacks be avoided without substantial
loss in data quality?
4. What are the best ways to promote transparency and procedural justice in the perfor-
mance evaluation process, for both assessors and assessees?
5. How can simulations be used to assess performance on other dimensions of perfor-
mance, in addition to technical performance? This is happening to some degree, but new
technologies make it possible to do much more.
6. How can simulations be used to assess adaptive responses to changing requirements, and
not just on the technical dimension?
7. What are the best ways to display and present behavioral performance data to facilitate
effective evaluation and decision making? How can research on dashboards and other
methods from computer and data science be effectively combined with meaningful
performance metrics (Yigitbasioglu & Velcu 2012)?
In sum, the last 100 years have seen a great deal of research and development regarding the
determinants of performance, including a wide variety of contextual variables, and OP/OB has
made much progress. It is our hope that over the next few years (fewer than 100), more effort
will be devoted to explicating and understanding the dependent variable side of the equation—
performance itself.
68 Campbell Wiernik
DISCLOSURE STATEMENT
The authors are not aware of any affiliations, memberships, funding, or financial holdings that
might be perceived as affecting the objectivity of this review.
ACKNOWLEDGMENTS
We thank Kylee Bolt, Jeffrey Braun, Marissa Clark, and Shannon Evans for their invaluable assis-
tance in preparing this article.
LITERATURE CITED
Aguinis H, O’Boyle E. 2014. Star performers in twenty-first century organizations. Pers. Psychol. 67:313–50
Austin JT, Villanova P. 1992. The criterion problem: 1917–1992. J. Appl. Psychol. 77:836–74
Baard SK, Rench TA, Kozlowski SWJ. 2014. Performance adaptation: a theoretical integration and review.
J. Manag. 40:48–99
Barron LG, Sackett PR. 2008. Asian variability in performance rating modesty and leniency bias. Hum.
Perform. 21:277–90
Bartram D. 2005. The great eight competencies: a criterion centric approach to construct validation. J. Appl.
Psychol. 90:1185–203
Beal DJ, Weiss HM, Barros E, MacDermid SM. 2005. An episodic process model of affective influences on
performance. J. Appl. Psychol. 90:1054–68
Beck JW, Beatty AS, Sackett PR. 2014. On the distribution of job performance: the role of measurement
characteristics in observed departures from normality. Pers. Psychol. 67:531–66
Bennett RJ, Robinson SL. 2000. Development of a measure of workplace deviance. J. Appl. Psychol.
85:349–60
Berry CM, Ones DS, Sackett PR. 2007. Interpersonal deviance, organizational deviance, and their common
correlates: a review and meta-analysis. J. Appl. Psychol. 92:410–24
Beus JM, Whitman DS. 2012. The relationship between typical and maximum performance: a meta-analytic
examination. Hum. Perform. 25:355–76
Bommer WH, Johnson JL, Rich GA, Podsakoff PM, MacKenzie SB. 1995. On the interchangeability of
objective and subjective measures of employee performance: a meta-analysis. Pers. Psychol. 48:587–605
Borman WC, Motowidlo SJ. 1993. Expanding the criterion domain to include elements of contextual per-
formance. See Schmitt & Borman 1993, pp. 71–98
Borman WC, Motowidlo SJ. 1997. Task performance and contextual performance: the meaning for personnel
selection research. Hum. Perform. 10:99–109
Borsboom D, Mellenberg GJ, van Heerden J. 2003. The theoretical status of latent variables. Psychol. Rev.
110:203–19
Brumback GB, Vincent JW. 1970. Factor analysis of work-performed data for a sample of administrative,
professional, and scientific positions. Pers. Psychol. 23:101–7
Campbell JP. 2012. Behavior, performance, and effectiveness in the twenty-first century. See Kozlowski 2012,
pp. 159–96
Campbell JP. 2013a. Assessment in I/O psychology: an overview. See Geisinger et al. 2013, pp. 355–95
Campbell JP. 2013b. Leadership, the old, the new, and the timeless: a commentary. See Rumsey 2013,
pp. 401–22
Campbell JP, Hanson MA, Oppler SH. 2001. Modeling performance in a population of jobs. See Campbell &
Knapp 2001, pp. 307–34
Campbell JP, Knapp DJ, eds. 2001. Exploring the Limits in Personnel Selection and Classification. Mahwah,
NJ: Erlbaum
Campbell JP, Kuncel NR. 2001. Individual and team training. In Handbook of Industrial, Work &
Organizational Psychology, Vol. 1: Personnel Psychology, ed. N Anderson, DS Ones, HK Sinangil,
C Viswesvaran, pp. 278–313. Thousand Oaks, CA: Sage

Campbell JP, McCloy RA, Oppler SH, Sager CE. 1993. A theory of performance. See Schmitt & Borman 1993,
pp. 35–70
Campion MA, Fink AA, Ruggeberg BJ, Carr L, Phillips GM, Odman RB. 2011. Doing competencies well: best
practices in competency modeling. Pers. Psychol. 64:225–62
Carhart MM. 1997. On persistence in mutual fund performance. J. Financ. 52:57–82
Chan D, ed. 2014. Individual Adaptability to Changes at Work: New Directions in Research. New York:
Routledge
Colegrove CM, Bennett W Jr. 2006. Competency-based training: adapting to warfighter needs. Mesa, AZ: Air
Force Res. Lab.
Conway JM. 1996. Additional construct validity evidence for the task/contextual performance distinction.
Hum. Perform. 9:309–29
Conway JM, Huffcutt AI. 1997. Psychometric properties of multisource performance ratings: a meta-analysis
of subordinate, supervisor, peer, and self-ratings. Hum. Perform. 10:331–60

Cravens DW, Ingram TN, LaForge RW, Young CE. 1993. Behavior-based and outcome-based salesforce
control systems. J. Mark. 57:47–59
Dalal RS. 2005. A meta-analysis of the relationship between organizational citizenship behavior and coun-
terproductive work behavior. J. Appl. Psychol. 90:1241–55
Dalal RS, Bhave DP, Fiset J. 2014. Within-person variability in job performance: a theoretical review and
research agenda. J. Manag. 40(5):1396–436
Deadrick DL, Bennett N, Russell CJ. 1997. Using hierarchical linear modeling to examine dynamic perfor-
mance criteria over time. J. Manag. 23:745–57
Deadrick DL, Gardner DG. 2008. Maximal and typical measures of job performance: an analysis of per-
formance variability over time. Hum. Resour. Manag. Rev. 18:133–45
DeChurch LA, Mesmer-Magnus JR. 2010. The cognitive underpinnings of effective teamwork: a meta-
analysis. J. Appl. Psychol. 95:32–53
DeNisi AS, Kluger AN. 2000. Feedback effectiveness: Can 360-appraisals be improved? Acad. Manag.
Perspect. 14:129–39
DeNisi AS, Pritchard RD. 2006. Performance appraisal, performance management and improving individual
performance: a motivational framework. Manag. Organ. Rev. 2:253–77
DeNisi AS, Sonesh S. 2011. The appraisal and management of performance at work. See Zedeck 2011, pp. 255–79
Diamantopoulos A, Riefler P, Roth KP. 2008. Advancing formative measurement models. J. Bus. Res.
61(12):1203–18
DuBois CL, Sackett PR, Zedeck S, Fogli L. 1993. Further exploration of typical and maximum performance
criteria: definitional issues, prediction, and White-Black differences. J. Appl. Psychol. 78:205–11
Dunning D, Heath C, Suls JM. 2004. Flawed self-assessment implications for health, education, and the
workplace. Psychol. Sci. Public Interest 5:69–106
Facteau JD, Craig SB. 2001. Are performance ratings from different rater sources comparable? J. Appl.
Psychol. 86:215–27
Farr JL, Tippins NT, eds. 2010. Handbook of Employee Selection. New York: Routledge
Festing M, Knappert L, Dowling PJ, Engle AD. 2012. Global performance management in MNEs: concep-
tualization and profiles of country-specific characteristics in China, Germany, and the United States.
Thunderbird Int. Bus. Rev. 54:825–43
Fisher CD. 2008. What if we took within-person performance variability seriously? Ind. Organ. Psychol. 1:185–89
Fleishman EA, Quaintance MK. 1984. Taxonomies of Human Performance: The Description of Human
Tasks. Orlando, FL: Academic
Frese M. 2008. The word is out: We need an active performance concept for modern workplaces. Ind. Organ.
Psychol. 1:67–69
Gable SL, Reis HT, Ward AJ. 2003. Evidence for bivariate systems: an empirical test of appetition and aversion
across domains. J. Res. Personal. 37:349–72
Geisinger KF, Bracken BA, Carlson JF, Hansen JIC, Kuncel NR, et al., eds. 2013. APA Handbook of Testing
and Assessment in Psychology, Vol. 1: Test Theory and Testing and Assessment in Industrial and
Organizational Psychology. Washington, DC: Am. Psychol. Assoc.
70 Campbell Wiernik
Goodman PS, Devadas R, Griffith-Hughson TL. 1988. Groups and productivity: analyzing the effectiveness
of self-management teams. In Productivity in Organizations: New Perspectives from Industrial and
Organizational Psychology, ed. JP Campbell, RJ Campbell, pp. 295–327. San Francisco: Jossey-Bass
Griffin MA, Neal A, Parker SK. 2007. A new model of work role performance: positive behavior in uncertain
and interdependent contexts. Acad. Manag. J. 50:327–47
Gruys ML, Sackett PR. 2003. Investigating the dimensionality of counterproductive work behavior. Int. J. Sel.
Assess. 11:30–42
Haertel EH. 2013. Reliability and validity of inferences about teachers based on student test scores. William H.
Angoff Meml. Lect. Ser. 14, Cent. Res. Hum. Cap. Educ., ETS Res. Dev., Mar. 22, Washington, DC
Halpin AL, Biggs WD. 2009. Evaluating business plans in a simulation environment. Dev. Bus. Simul. Exp.
Learn. 36:149–54
He P. 2012. Counterproductive work behavior among Chinese knowledge workers. Int. J. Sel. Assess.
20:119–38
Heidemeier H, Moser K. 2009. Self–other agreement in job performance ratings: a meta-analytic test of
a process model. J. Appl. Psychol. 94:353–70
Heinrichs WL, Youngblood P, Harter PM, Dev P. 2008. Simulation for team training and assessment: case
studies of online training with virtual worlds. World J. Surg. 32:161–70
Hermelin E, Lievens F, Robertson IT. 2007. The validity of assessment centres for the prediction of supervisory
performance ratings: a meta-analysis. Int. J. Sel. Assess. 15:405–11
Hoffman BJ, Gorman CA, Blair CA, Meriac JP, Overstreet B, Atchley EK. 2012. Evidence for the effectiveness
of an alternative multisource performance rating methodology. Pers. Psychol. 65:531–63
Hoffman BJ, Lance CE, Bynum B, Gentry WA. 2010. Rater source effects are alive and well after all. Pers.
Psychol. 63:119–51
Hoffman BJ, Woehr DJ. 2009. Disentangling the meaning of multisource performance rating source and
dimension factors. Pers. Psychol. 62:735–65
Hoffman RR, Militello LG. 2009. Perspectives on Cognitive Task Analysis. New York: Psychol. Press,
Taylor & Francis
Hofmann DA, Jacobs R, Gerras SJ. 1992. Mapping individual performance over time. J. Appl. Psychol. 77:185–95
Hogan J, Hogan R, Kaiser RB. 2011. Management derailment. See Zedeck 2011, pp. 555–76
Howard A. 1983. Work samples and simulations in competency evaluation. Prof. Psychol. Res. Pract. 14:780–96
Hunt ST. 2011. Technology is transforming the nature of performance management. Ind. Organ. Psychol. 4:188–89
Hunter JE. 1983. A causal analysis of cognitive ability, job knowledge, job performance, and supervisor
ratings. In Performance Measurement and Theory, ed. FJ Landy, S Zedeck, JN Cleveland, pp. 257–66.
Mahwah, NJ: Erlbaum
Kargupta H, Sarkar K, Gilligan M. 2010. MineFleet: an overview of a widely adopted distributed vehicle
performance data mining system. Proc. ACM SIGKDD Int. Conf. Knowl. Discov. Data Min., 16th,
Washington, DC, July 25–29, pp. 37–46. New York: ACM
Katzell RA, Guzzo RA. 1983. Psychological approaches to productivity improvement. Am. Psychol. 38:468–72
Kim Y, Ployhart RE. 2014. The effects of staffing and training on firm productivity and profit growth before,
during, and after the Great Recession. J. Appl. Psychol. 99:361–89
Klehe U-C, Anderson N. 2007. Working hard and working smart: motivation and ability during typical and
maximum performance. J. Appl. Psychol. 92:978–92
Kluger AN, DeNisi A. 1996. The effects of feedback interventions on performance: a historical review, a meta-
analysis, and a preliminary feedback intervention theory. Psychol. Bull. 119:254–84
Koopmans L, Bernaards CM, Hildebrandt VH, Schaufeli WB, de Vet Henrica CW, van der Beek AJ. 2011.
Conceptual frameworks of individual work performance: a systematic review. J. Occup. Environ. Med.
53:856–66
Kozlowski SWJ, ed. 2012. The Oxford Handbook of Organizational Psychology, Vol. 1. New York: Oxford
Univ. Press
Kozlowski SWJ, Chao GT, Morrison RF. 1998. Games raters play: politics, strategies, and impression
management in performance appraisal. In Performance Appraisal: State of the Art in Practice, ed.
JW Smither, pp. 163–205. San Francisco: Jossey-Bass

Kunkler K. 2006. The role of medical simulation: an overview. Int. J. Med. Robotics Comput. Assist. Surg. 2:203–10
Lance CE, Hoffman BJ, Gentry WA, Baranik LE. 2008. Rater source factors represent important sub-
components of the criterion construct space, not rater bias. Hum. Resour. Manag. Rev. 18:223–32
Lance CE, Teachout MS, Donnelly TM. 1992. Specification of the criterion construct space: an application of
hierarchical confirmatory factor analysis. J. Appl. Psychol. 77:437–52
Levy PE, Williams JR. 2004. The social context of performance appraisal: a review and framework for the
future. J. Manag. 30:881–905
Lievens F, Patterson F. 2011. The validity and incremental validity of knowledge tests, low-fidelity simulations,
and high-fidelity simulations for predicting job performance in advanced-level high-stakes selection.
J. Appl. Psychol. 96:927–40
Locke EA, Latham GP. 2002. Building a practically useful theory of goal setting and task motivation: a 35-year
odyssey. Am. Psychol. 57:705–17
Lohr S. 2013. Big data, trying to build better workers. New York Times, Apr. 21, p. BU4
Lord RG, Diefendorff JM, Schmidt AM, Hall RJ. 2010. Self-regulation at work. Annu. Rev. Psychol. 61:543–68
Lord RG, Dinh JE. 2014. What have we learned that is critical in understanding leadership perceptions and
leader-performance relations? Ind. Organ. Psychol. 7:158–77
Mangos PM, Arnold RD. 2008. Enhancing military training through the application of maximum and typical
performance measurement principles. Perform. Improv. 47:29–35
Mangos PM, Steele-Johnson D, LaHuis D, White ED. 2007. A multiple-task measurement framework for
assessing maximum-typical performance. Hum. Perform. 20:241–58
Marcus B, Schuler H, Quell P, Hümpfner G. 2002. Measuring counterproductivity: development and initial
validation of a German self-report questionnaire. Int. J. Sel. Assess. 10:18–35
Markon KE, Krueger RF, Watson D. 2005. Delineating the structure of normal and abnormal personality: an
integrative hierarchical approach. J. Personal. Soc. Psychol. 88:139–57
McAfee A, Brynjolfsson E. 2012. Big data: the management revolution. Harv. Bus. Rev. 90:60–68
Miles DE, Borman WE, Spector PE, Fox S. 2002. Building an integrative model of extra role work behaviors:
a comparison of counterproductive work behavior with organizational citizenship behavior. Int. J. Sel.
Assess. 10(1/2):51–57
Miller JS. 2003. High tech and high performance: managing appraisal in the information age. J. Labor Res.
24:409–24
Motowidlo SJ. 2000. Some basic issues related to contextual performance and organizational citizenship
behavior in human resource management. Hum. Resour. Manag. Rev. 10:115–26
Motowidlo SJ, Borman WC, Schmit MJ. 1997. A theory of individual differences in task and contextual
performance. Hum. Perform. 10:71–83
Mueller L. 2011. How I-O can contribute to the teacher evaluation debate: a response to Lefkowitz. TIP 49:17
Murphy KR. 1989a. Dimensions of job performance. In Testing: Theoretical and Applied Perspectives, ed.
RF Dillon, JW Pellegrino, pp. 218–47. New York: Praeger
Murphy KR. 1989b. Is the relationship between cognitive ability and job performance stable over time? Hum.
Perform. 2:183–200
Murphy KR. 2008. Explaining the weak relationship between job performance and ratings of job performance.
Ind. Organ. Psychol. 1:148–60
Murphy KR, Cleveland JN. 1995. Understanding Performance Appraisal: Social, Organizational, and Goal-
Based Perspectives. Thousand Oaks, CA: Sage
Murphy KR, Dechert PJ. 2013. 2013 performance appraisal. See Geisinger et al. 2013, pp. 611–27
Murphy KR, DeShon R. 2000. Interrater correlations do not estimate the reliability of job performance ratings.
Pers. Psychol. 53:873–900
Noonan LE, Sulsky LM. 2001. Impact of frame-of-reference and behavioral observation training on alternative
training effectiveness criteria in a Canadian military sample. Hum. Perform. 14:3–26
O’Boyle E, Aguinis H. 2012. The best and the rest: revisiting the norm of normality of individual performance.
Pers. Psychol. 65:79–119
O’Neil HF, Allred K, Dennis RA. 1997. Validation of computer simulation for assessment of interpersonal skills. In
Workforce Readiness: Competencies and Assessments, ed. HF O’Neil, pp. 229–54. Mahwah, NJ: Erlbaum
72 Campbell Wiernik
Ones DS, Dilchert S. 2012. Employee green behaviors. In Managing Human Resources for Environmental
Sustainability, ed. SE Jackson, DS Ones, S Dilchert, pp. 85–116. San Francisco: Jossey-Bass/Wiley
Ones DS, Dilchert S. 2013. Counterproductive work behaviors: concepts, measurement, and nomological
network. See Geisinger et al. 2013, pp. 643–59
Ones DS, Viswesvaran C, Schmidt FL. 2008. No new terrain: reliability and construct validity of job per-
formance ratings. Ind. Organ. Psychol. 1:174–79
Organ DW. 1988. Organizational Citizenship Behavior: The Good Soldier Syndrome. Lexington, MA:
Lexington Books
Organ DW, Podsakoff PM, Podsakoff NP. 2011. Expanding the criterion domain to include organizational
citizenship behavior: implications for employee selection. See Zedeck 2011, pp. 281–323
Pearlman K, Sanchez JI. 2010. Work analysis. See Farr & Tippins 2010, pp. 73–98
Peretz H, Fried Y. 2012. National cultures, performance appraisal practices, and organizational absenteeism
and turnover: a study across 21 countries. J. Appl. Psychol. 97:448–59

Ployhart RE, Bliese PD. 2006. Individual adaptability (I-ADAPT) theory: conceptualizing the antecedents,
consequences, and measurement of individual differences in adaptability. See Salas 2006, pp. 3–39
Ployhart RE, Hale D. 2014. The fascinating psychological microfoundations of strategy and competitive
advantage. Annu. Rev. Organ. Psychol. Organ. Behav. 1:145–72
Podsakoff PM, MacKenzie SB, Paine JB, Bachrach DG. 2000. Organizational citizenship behaviors: a critical
review of the theoretical and empirical literature and suggestions for future research. J. Manag. 26:513–63
Pulakos ED, Arad S, Donovan MA, Plamondon KE. 2000. Adaptability in the workplace: development of
a taxonomy of adaptive performance. J. Appl. Psychol. 85:612–24
Pulakos ED, Dorsey DW, White SS. 2006. Adaptability in the workplace: selecting an adaptive workforce. See
Salas 2006, pp. 41–71
Pulakos ED, O’Leary RS. 2010. Defining and measuring results of workplace behavior. See Farr & Tippins
2010, pp. 513–29
Pulakos ED, O’Leary RS. 2011. Why is performance management broken? Ind. Organ. Psychol. 4:146–64
Pulakos ED, Schmitt N, Dorsey DW, Arad S, Borman WC, Hedge JW. 2002. Predicting adaptive performance:
further tests of a model of adaptability. Hum. Perform. 15:299–323
Putka DJ, Hoffman BJ. 2014. “The” reliability of job performance ratings equals 0.52. In More Statistical and
Methodological Myths and Urban Legends, ed. CE Lance, RJ Vandenberg, pp. 247–75. New York:
Taylor & Francis
Riggio RE, Mayes BT, Schleicher DJ. 2003. Using assessment center methods for measuring undergraduate
business student outcomes. J. Manag. Inq. 12:68–78
Rotundo M, Xie JL. 2013. Understanding the domain of counterproductive work behaviour in China. In
Human Resource Management “With Chinese Characteristics”: Facing the Challenges of Globalization,
ed. M Warner, pp. 86–107. New York: Routledge
Rumsey MG, ed. 2013. The Oxford Handbook of Leadership. New York: Oxford Univ. Press
Rupp DE, Gibbons AM, Baldwin AM, Snyder LA, Spain SM, et al. 2006. An initial validation of developmental
assessment centers as accurate assessments and effective training interventions. Psychol. Manag. J.
9:171–200
Sackett PR. 2007. Revisiting the origins of the typical-maximum performance distinction. Hum. Perform.
20:179–85
Salas E, ed. 2006. Advances in Human Performance and Cognitive Engineering Research, Vol. 6. Bingley, UK:
Emerald
Sanchez JI, Levine EL. 2009. What is (or should be) the difference between competency modeling and tra-
ditional job analysis? Hum. Resour. Manag. Rev. 19:53–63
Sauer DA. 1997. Information content of prior period mutual fund performance rankings. J. Econ. Bus.
49:549–67
Schleicher DJ, Day DV, Mayes BT, Riggio RE. 2002. A new frame for frame-of-reference training: enhancing
the construct validity of assessment centers. J. Appl. Psychol. 87:735–46
Schmidt FL, Viswesvaran C, Ones DS. 2000. Reliability is not validity and validity is not reliability. Pers.
Psychol. 53:901–12

Schmitt N, Borman WC, eds. 1993. Personnel Selection in Organizations. San Francisco: Jossey-Bass
Shippmann JS. 2010. Competencies, job analysis, and the next generation of modeling. In Handbook of
Workplace Assessment, ed. JC Scott, DH Reynolds, pp. 197–232. San Francisco: Jossey-Bass/Wiley
Shippmann JS, Ash RA, Batjtsta M, Carr L, Eyde LD, et al. 2000. The practice of competency modeling. Pers.
Psychol. 53:703–40
Şimşek B, Pakdil F, Dengiz B, Testik MC. 2013. Driver performance appraisal using GPS terminal mea-
surements: a conceptual framework. Transp. Res. Part C Emerg. Technol. 26:49–60
Sonnentag S, Frese M. 2012. Dynamic performance. See Kozlowski 2012, pp. 548–78
Spector PE, Bauer JA, Fox S. 2010. Measurement artifacts in the assessment of counterproductive work
behavior and organizational citizenship behavior: Do we know what we think we know? J. Appl. Psychol.
95:781–90
Spector PE, Cha XX. 2014. Re-examining citizenship: how the control of measurement artifacts affects
observed relationships of organizational citizenship behavior and organizational variables. Hum. Per-
form. 27:165–82
Spector PE, Fox S, Penney LM, Bruursema K, Goh A, Kessler S. 2006. The dimensionality of counter-
productivity: Are all counterproductive behaviors created equal? J. Vocat. Behav. 68:446–60
Spence JR, Keeping LM. 2010. The impact of non-performance information on ratings of job performance:
a policy-capturing approach. J. Organ. Behav. 31:587–608
Stevens GW. 2013. A critical review of the science and practice of competency modeling. Hum. Resour. Dev.
Rev. 12:86–107
Stewart GL, Nandkeolyar AK. 2007. Exploring how constraints created by other people influence intra-
individual variation in objective performance measures. J. Appl. Psychol. 92:1149–58
Stokes CK, Schneider TR, Lyons JB. 2010. Adaptive performance: a criterion problem. Team Perform. Manag.
16(3/4):212–30
Sturman MC. 2003. Searching for the inverted U-shaped relationship between time and performance:
meta-analyses of the experience/performance, tenure/performance, and age/performance relationships.
J. Manag. 29:609–40
Taylor S, Todd PA. 1995. Understanding information technology usage: a test of competing models. Inf. Syst.
Res. 6:144–76
Tett RP, Guterman HA, Bleier A, Murphy PJ. 2000. Development and content validation of a “hyper-
dimensional” taxonomy of managerial competence. Hum. Perform. 13:205–51
Thomas JL, Dickson MW, Bliese PD. 2001. Values predicting leader performance in the U.S. Army Reserve
Officer Training Corps Assessment Center: evidence for a personality-mediated model. Leadersh. Q.
12:181–96
Vance RJ, MacCallum RC, Coovert MD, Hedge JW. 1988. Construct validity of multiple job performance
measures using confirmatory factor analysis. J. Appl. Psychol. 73:74–80
Viswesvaran C, Schmidt FL, Ones DS. 2005. Is there a general factor in job performance ratings? A meta-
analytic framework for disentangling substantive and error influences. J. Appl. Psychol. 90:108–31
Walumbwa FO, Wernsing T. 2013. From transactional and transformational leadership to authentic lead-
ership. See Rumsey 2013, pp. 392–400
Wisecarver MM, Carpenter TD, Kilcullen RN. 2007. Capturing interpersonal performance in a latent per-
formance model. Mil. Psychol. 19:83–101
Woehr DJ, Roch S. 2012. Supervisory performance ratings. In The Oxford Handbook of Personnel Selection
and Assessment, ed. N Schmitt, pp. 517–31. New York: Oxford Univ. Press
Xu S, Wang Q, Liu C, Li Y, Ouyang K. 2013. Content and construct of counterproductive work behavior in
a Chinese context. Soc. Behav. Personal. Int. J. 41:921–32
Yigitbasioglu OM, Velcu O. 2012. A review of dashboards in performance management: implications for
design and research. Int. J. Account. Inf. Syst. 13:41–59
Zedeck S, ed. 2011. APA Handbook of Industrial and Organizational Psychology, Vol. 2: Selecting and
Developing Members for the Organization. Washington, DC: Am. Psychol. Assoc.
Zyphur MJ, Chaturvedi S, Arvey RD. 2008. Job performance over time is a function of latent trajectories and
previous performance. J. Appl. Psychol. 93:217–24
74 Campbell Wiernik
Annual Review of
Organizational
Psychology and
Organizational Behavior
Volume 2, 2015 Contents

Organizational Psychology Then and Now: Some Observations

Edgar H. Schein . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Group Affect
Sigal G. Barsade and Andrew P. Knight . . . . . . . . . . . . . . . . . . . . . . . . . . 21
The Modeling and Assessment of Work Performance
John P. Campbell and Brenton M. Wiernik . . . . . . . . . . . . . . . . . . . . . . . . 47
Justice, Fairness, and Employee Reactions
Jason A. Colquitt and Kate P. Zipay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Methodological and Substantive Issues in Conducting Multinational and
Cross-Cultural Research
Paul E. Spector, Cong Liu, and Juan I. Sanchez . . . . . . . . . . . . . . . . . . . . 101
Leadership Development: An Outcome-Oriented Review Based on Time and
Levels of Analyses
David V. Day and Lisa Dragoni . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Beyond Lewin: Toward a Temporal Approximation of Organization
Development and Change
Jean M. Bartunek and Richard W. Woodman . . . . . . . . . . . . . . . . . . . . . 157
Beyond the Big Five: New Directions for Personality Research and Practice in
Organizations
Leaetta M. Hough, Frederick L. Oswald, and Jisoo Ock . . . . . . . . . . . . . 183
Corporate Social Responsibility: Psychological, Person-Centric, and
Progressing
Deborah E. Rupp and Drew B. Mallory . . . . . . . . . . . . . . . . . . . . . . . . . 211
Time in Individual-Level Organizational Studies: What Is It, How Is It Used,
and Why Isn’t It Exploited More Often?
Abbie J. Shipp and Michael S. Cole . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
vi
Dynamics of Well-Being
Sabine Sonnentag . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Low-Fidelity Simulations
Jeff A. Weekley, Ben Hawkes, Nigel Guenole, and Robert E. Ployhart . . . 295
Emotional Labor at a Crossroads: Where Do We Go from Here?
Alicia A. Grandey and Allison S. Gabriel . . . . . . . . . . . . . . . . . . . . . . . . . 323
Supporting the Aging Workforce: A Review and Recommendations for
Workplace Intervention Research
Donald M. Truxillo, David M. Cadiz, and Leslie B. Hammer . . . . . . . . . 351
ESM 2.0: State of the Art and Future Potential of Experience Sampling
Methods in Organizational Research
Daniel J. Beal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
Ethical Leadership
Deanne N. Den Hartog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
Differential Validity and Differential Prediction of Cognitive Ability Tests:
Understanding Test Bias in the Employment Context
Christopher M. Berry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
Organizational Routines as Patterns of Action: Implications for Organizational
Behavior
Brian T. Pentland and Thorvald Hærem . . . . . . . . . . . . . . . . . . . . . . . . . 465
Pay, Intrinsic Motivation, Extrinsic Motivation, Performance, and Creativity
in the Workplace: Revisiting Long-Held Beliefs
Barry Gerhart and Meiyu Fang . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489
Stereotype Threat in Organizations: Implications for Equity and Performance
Gregory M. Walton, Mary C. Murphy, and Ann Marie Ryan . . . . . . . . . . 523
Technology and Assessment in Selection
Nancy T. Tippins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
Workplace Stress Management Interventions and Health Promotion
Lois E. Tetrick and Carolyn J. Winslow . . . . . . . . . . . . . . . . . . . . . . . . . 583
Errata
An online log of corrections to Annual Review of Organizational Psychology and
Organizational Behavior articles may be found at http://www.annualreviews.org/
errata/orgpsych.
Contents vii

Campbell & Wiernik (2015)

Uploaded by

Document Informationclick to expand document information

Document Informationclick to expand document information

Copyright:

Available Formats

Campbell & Wiernik (2015)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Campbell & Wiernik (2015)

Uploaded by

Copyright:

Available Formats

ANNUAL

REVIEWS Further The Modeling and

• Other articles in this volume

• Top cited articles

Annu. Rev. Organ. Psychol. Organ. Behav. 2015. Keywords

What Performance Is Not

www.annualreviews.org  Modeling and Assessing Work Performance 49

Core Technical and Contextual Performance

Organizational Citizenship Behavior

Counterproductive Work Behavior

www.annualreviews.org  Modeling and Assessing Work Performance 51

Proficiency, Adaptability, and Proactivity

The Campbell Revision

www.annualreviews.org  Modeling and Assessing Work Performance 53

SIX SUBFACTORS COMPRISING LEADERSHIP PERFORMANCE

www.annualreviews.org  Modeling and Assessing Work Performance 55

government agencies, nongovernment organizations, the public).

The General Factor

CONTENT VERSUS PROCESS IN CONCEPTUALIZING PERFORMANCE

www.annualreviews.org  Modeling and Assessing Work Performance 57

considerable convergent validity.

Purposes for Performance Assessment

www.annualreviews.org  Modeling and Assessing Work Performance 59

General Measurement Issues

www.annualreviews.org  Modeling and Assessing Work Performance 61

practices and preferences to fulfill their purposes.

Performance Assessment Methods

www.annualreviews.org  Modeling and Assessing Work Performance 63

www.annualreviews.org  Modeling and Assessing Work Performance 65

Technology-enhanced assessment. At an increasing rate, popular press and management and

Goal Achievement and Performance Outcomes

(e.g., peer leadership and management) may go unassessed.

CONCLUSIONS AND RECOMMENDATIONS

www.annualreviews.org  Modeling and Assessing Work Performance 67

www.annualreviews.org  Modeling and Assessing Work Performance 69

of subordinate, supervisor, peer, and self-ratings. Hum. Perform. 10:331–60

www.annualreviews.org  Modeling and Assessing Work Performance 71

and turnover: a study across 21 countries. J. Appl. Psychol. 97:448–59

www.annualreviews.org  Modeling and Assessing Work Performance 73

Volume 2, 2015 Contents

Organizational Psychology Then and Now: Some Observations

You might also like

www.annualreviews.org Modeling and Assessing Work Performance 49

www.annualreviews.org Modeling and Assessing Work Performance 51

www.annualreviews.org Modeling and Assessing Work Performance 53

www.annualreviews.org Modeling and Assessing Work Performance 55

www.annualreviews.org Modeling and Assessing Work Performance 57

www.annualreviews.org Modeling and Assessing Work Performance 59

www.annualreviews.org Modeling and Assessing Work Performance 61

www.annualreviews.org Modeling and Assessing Work Performance 63

www.annualreviews.org Modeling and Assessing Work Performance 65

www.annualreviews.org Modeling and Assessing Work Performance 67

www.annualreviews.org Modeling and Assessing Work Performance 69

www.annualreviews.org Modeling and Assessing Work Performance 71

www.annualreviews.org Modeling and Assessing Work Performance 73