Performance Flashcards
(Borman et al., 1993; Schmidt et al., 1986).
Previous research indicates that general ability measures are predictive of performance on all jobs, with general ability exerting its primary influence indirectly through job knowledge.
Coward and Sackett (1990)
conducted a definitive study involving 174 independent samples with a mean sample size of 210 (a database of 36,540 individuals). They found no evidence for nonlinear relationships between ability and performance
(Adler et al., 2016)
Supervisors and employees both dread performance
appraisals (as echoed by Murphy, 2019)
this article recaps the points made by the panelists who participated
in the debate. The arguments for eliminating ratings include these: (a) the disappointing interventions, (b) the disagreement when multiple raters evaluate the same
performance, (c) the failure to develop adequate criteria for evaluating ratings, (d) the weak relationship between the performance of ratees and the ratings they receive,
(e) the conficting purposes of performance ratings in organizations, (f) the inconsistent efects of performance feedback on subsequent performance,
The
arguments for retaining ratings include (a) the recognition that changing the rating
process is likely to have minimal efect on the performance management process as
a whole, (b) performance is always evaluated in some manner, (c) “too hard” is no
excuse for industrial–organizational (I-O) psychology, (d) ratings and diferentiated
evaluations have many merits for improving organizations, (e) artifcial tradeofs
are driving organizations to inappropriately abandon ratings, (f) the alternatives
to ratings may be worse, and (g) the better questions are these: How could performance ratings be improved
(Pulakos et al., 2015)
the great majority of appraisal systems in organisations are viewed as ineffective.
Performance management systems do not seem to fare much better; the conclusion that performance management is broken is shared among many researchers and practitioners.
For example, no solution
has been found to ameliorate the seemingly
intractable problem of leniency in ratings.
DeNisi & Murphy, 2017
Reviews of research on both performance appraisal and performance management noted that there is little if any evidence that these systems have any real impact on the performance or effectiveness of employees
Murphy (2019)
First, many stakeholders believe that PM is important and beneficial to measure
performance and to use that information to drive decisions. Second, it is widely believed that performance
feedback is valuable and that it helps to improve employee motivation and performance. Third, there does not
seem to be any clear alternative to the type of evaluation system most organizations use; virtually, every system
that has been proposed to replace traditional performance appraisal (e.g., performance management systems,
evaluation systems based on objective performance, and productivity measures) has fared as badly, if not worse.
RE: power law & why it makes appraisal systems lose their luster.
The stars
should be easy to spot without appraisal systems, and virtually, everyone else will be performing at such a lower level that differentiating
among these more average performers will be virtually pointless.
- WOW - Ufortunately, the people
who need and would benefit from performance feedback (e.g., poor performers) are often actively avoid feedback
(e.g., Moss et al., 2003; Moss, Sanchez, Brumbaugh, & Borkowski, 2009).
- WOW - Ufortunately, the people
There is considerable evidence that in order to work, performance feedback must
be accepted by recipients as fair and valid (Anseel & Lievens, 2009)
First, performance ratings, even when they are truly accurate, are often seen by recipients as unduly harsh. There
is extensive evidence (e.g., Campbell, Campbell, & Ho-Beng, 1998; Harris & Schaubroeck, 1988; Meyer, 1980; Thornton, 1980) that people view their own performance more favourably that do their supervisors, their peers, or other
external raters.
the tendency for people to view ratings they receive from others as unfairly low has been
identified by Murphy et al. (2018) as one of the principal structural sources of failure in performance appraisal systems.
Murphy et al. (2018)
discuss the “death spiral” of performance appraisal systems, describing ways in which disappointing experiences with
performance evaluation lead to higher levels of cynicism and disengagement, and show how these negative experiences feed upon themselves.
Yet, there is evidence that developmental feedback can be useful when the recipient is new to the
task or is a newcomer in the organisation (Li, Harris, Boswell & Xie, 201; Nurse, 2005; Reilly et al., 1996
Because the feedback coaches provide is
typically focused on learning and development rather than evaluation, rewards, and sanctions, employees may be
more receptive to this form of help and guidance than they are to traditional performance evaluations.
perf eval should not be done for more than one reason at a time!
Kluger and Nir (2010)
suggest using feedforward.
The feedforward interview focuses on (a) articulating what has gone well, by eliciting description of positive experiences from the target; (b) understanding how the strengths of the target and the context in which the event or
experience contributed to that positive experience; and (c) helping the target to apply those strengths and/or experiences to future challenges.
Bouskila-Yam and Kluger
(2011) proposed a fundamental reorientation of performance appraisal and feedback.
they
proposed that performance reviews should focus on what people do well and should work towards establishing goals
that are based on strengths rather than weaknesses.
Several surveys have shown that performance evaluations are most commonly used in organisations for two
purposes—that is, to provide information that can be useful for making training and development and to serve as
input for decisions about salary, promotions, layoffs, and dismissals
Cleveland et al. (1989)
suggested that there were four broad purposes
for evaluating job performance: (a) to make distinctions between individuals, such as identifying the best candidates
for salary increases or promotions; (b) to make distinctions within individuals, such as identifying individual strengths
and weaknesses for the purpose of determining training and development needs and priorities; (c) to support HR systems in organisations, such as validating personnel tests, evaluating the success, or training programmes; and (d) documentation, such as providing a record to support decisions such as promotions or dismissal.
O’Boyle & Aguinis (2012)
conducted 5 studies involving 198 samples including 633,263 researchers, entertainers, politicians, and amateur and professional athletes. Results are remarkably consistent across industries, types of jobs,
types of performance measures, and time frames and indicate that individual performance is not normally distributed.
When performance data do not conform to the normal distribution, then the conclusion is that the error “must” lie within the sample
not the population. Subsequent adjustments are made (e.g., dropping outliers) in order to make the sample “better reflect” the “true” underlying
normal curve.
he normal distribution has been used to model a variety of phenomena including human traits such as height (Yule, 1912) and intelligence
(Galton, 1889), as well as probability distributions (Hull, 1928), economic
trends such as stock pricing (Bronzin, 1908), and the laws of thermodynamics (Reif, 1965).
Ferguson (1947) noted that “ratings for a large and representative
group of assistant managers should be distributed in accordance with the
percentages predicted for a normal distribution” (p. 308)
The normality
assumption persisted through the years, and researchers began to not only
assume job performance normality but forced it upon the observed distributions regardless of the actual observed distributional properties. For
example, in developing a performance appraisal system, Canter (1953)
used “a forced normal distribution of judgments” (p. 456) for evaluating
open-ended responses. Likewise, Schultz and Siegel (1961) “forced the
[performance] rater to respond on a seven-point scale and to normalize
approximately the distribution of his responses”
Specifically, when performance scores deviate from normality, the cause is attributed to leniency bias, severity bias, and/or a halo error (Aguinis, 2009; Schneier,
1977).
Rating systems where most employees occupy the same category
with only a few at the highest category are assumed to be indicative of
range restriction and other “statistical artifacts” (Motowidlo & Borman,
1977).
Jacobs (1974) argued
that in sales industries (automotive, insurance, stock) performance is not
normal because a small group of incumbents who possess the expertise and salesmanship dominate activity.
Whereas a value exceeding three standard deviations from the mean is often thought to be an outlier in the
context of a normal curve (e.g., Orr, Sackett, & Dubois, 1991), a Paretian distribution would predict that these values are far more common
and that their elimination or transformation is a questionable practice.
There are important differences between Gaussian and Paretian distributions. First, Gaussian distributions underpredict the likelihood of extreme events. For instance, when stock market performance is predicted
using the normal curve, a single-day 10% drop in the financial markets should occur once every 500 years (Buchanan, 2004). In reality,
it occurs about once every 5 years (Mandelbrot, Hudson, & Grunwald,
2005)
Third, a key difference between
normal and Paretian distributions is scale invariance. In OBHRM, scale
invariance usually refers to the extent to which a measurement instrument
generalizes across different cultures or populations.
Germane to OBHRM in particular is that if performance operates
under power laws, then the distribution should be the same regardless of
the level of analysis. That is, the distribution of individual performance
should closely mimic the distribution of firm performance. Researchers
who study performance at the firm level of analysis do not necessarily
assume that the underlying distribution is normal (e.g., Stanley et al.,
1995).
if individual performance is found to
also follow a power law distribution, as it is the case for firm performance
(Bonardi, 2004; Powell, 2003; Stanley et al., 1995)
Of a total of 198 samples
of performers, 186 (93.94%) follow a Paretian distribution more closely
than a Gaussian distribution.
*Our results suggest that the distribution of
individual performance is such that most performers are in the lowest
category. Based on Study 1, we discovered that approximately 66% to 80% of performers in domains are in the lowest category/band for their appropriate performance metric.
Extending to the second standard deviation, the difference in productivity between
the 97.73rd percentile and median researcher should be four, and this additional output is valued at $22,652. However, the difference between the
two points is actually seven. Thus, if SDy is two, then the additional output
of these workers is $39,645 more than the median worker.
And leadership/mentoring of stars;
Thus, greater attention
should be paid to the tremendous impact of the few vital individuals.
Despite their small numbers, slight percentage increases in the output of
top performers far outweigh moderate increases of the many.
In addition to the study of leadership, our results also affect research
on work teams.
Future research:
We
may expect the group productivity to increase in the presence of an elite
worker, but is the increase in group output negated by the loss of individual output of the elite worker being slowed by non-elites? It may also be
that elites only develop in interactive, dynamic environments, and the isolation of elite workers or grouping multiple elites together could hamper
their abnormal productivity.
CWB
(i.e., harmful behaviors targeted at the organization or its members) has
always been assumed to have a strong, negative relation with the other two
components, but it is unclear if this relationship remains strong, or even
negative, among elite performers. For example, the superstars of Study 4
often appeared as supervillains in Study 5. Do the most productive workers
also engage in the most destructive behavior? If so, future research should
examine if this is due to managers’ fear of reprimanding a superstar, the
superstar’s sense of entitlement, non-elites covering for the superstar’s
misbehavior out of hero worship, or some interaction of all three.
JUICY:
inally, going beyond any individual research domain, a Paretian distribution of performance may help explain why despite more than a century
of research on the antecedents of job performance and the countless theoretical models proposed, explained variance estimates (R2) rarely exceed
.50 (Cascio & Aguinis, 2008b). It is possible that research conducted over
the past century has not made important improvements in the ability to predict individual performance because prediction techniques rely on means
and variances assumed to derive from normal distributions, leading to
gross errors in the prediction of performance
Random assignment will only balance the groups when the distribution of the outcome is normally distributed (when the prevalence
of outliers is low). In the case of Paretian distributions, the prevalence
of outliers is much higher.
. As a result, a single high performer has an
important impact on the mean of the group and ultimately on the significance or nonsignificance of the test statistic. Likewise, the residual
created by extreme performers’ distance from a regression line widens
standard errors to create Type II errors. Interestingly, the wide standard
errors and unpredictable means caused by extreme performers should result in great variability in findings in terms of both statistical significance
and direction. This may explain so many “inconsistent findings” in the
OBHRM literature (Schmidt, 2008).
Techniques
exist that properly and accurately estimate models where the outcome is
Paretian. Poisson processes are one such solution, and although not well
established in OBHRM research, they do have a history in the natural
sciences (e.g., Eliazar & Klafter, 2008) and finance. In addition, Bayesian techniques are likely to provide the greatest
applicability to the study of superstars.
The Matthew effect (Ceci & Papierno, 2005; Merton, 1968) states that those already in an advantageous
position are able to leverage their position to gain disproportionate rewards
Likewise, compensation systems such as pay for performance and CEO compensation are an especially divisive issue, with
many claiming that disproportionate pay is an indicator of unfair practices
(Walsh, 2008).
As
we described earlier, a Pareto curve demonstrates scale invariance, and
thus whether looking at the entire population or just the top percentile, the
same distribution shape emerges. For selection, this means that there are
real and important differences between the best candidate and the second
best candidate.
And the challenge of keeping superstars as other orgs want them!
In Studies 1, 2, 4, and 5, we found only one sample (NCAA rushing)
for which individual performance was better modeled with a Gaussian
distribution than a Paretian distribution.
Consider two measurement-related reasons for the potential better
fit of a Gaussian distribution. First, a measure of performance may be
too coarse to capture differences between superstars and the “simply adequate” (Aguinis, Pierce, & Culpepper, 2009). Specifically, in Study 3,
performance was measured as whether an official was elected or not, and
the measure did not capture differences among performers such as by how
many votes an individual won or lost an election.
supervisory ratings, are one of the most popular ways to operationalize
performance - Normality is introduced by the scale or rater evaluation training.
Now, consider three situations and reasons why the underlying performance distribution, not just observed performance scores, may actually fit
a Gaussian as opposed to a Paretian model. First, it may be the case that, in
certain industries and certain job types, superstars simply do not emerge.
For example, the manufacturing economy of the 20th century strove not
only for uniformity of product but also uniformity of worker. Quotas,
union maximums, assembly lines, and situational and technological constraints all constrained performance to values close to the mean.
However, industries and organizations that rely on manual labor, have limited technology, and place
strict standards for both minimum and maximum production are likely
to lead to normal distributions of individual performance. As we move
into the 21st century, software engineers, consultants, healthcare workers,
and educators make up an increasingly large part of the economy; but, for
the foreseeable future, farmers, factory workers, and construction crews
will continue to play an important role, and these types of jobs may best
be modeled with a normal distribution (e.g., Hull, 1928; Tiffin, 1947).
Second, research is needed on the deleterious
effects of superstars. For example, does the presence of a superstar demotivate other workers to such an extent that total organizational output
decreases.
When and how do these individuals reach the elite group? What is
the precise composition of this elite group—do individuals rotate in and
out of this group, or once in the top group, they remain in the top for
most of their career? What individual, group, and cultural factors predict
an individual’s membership in the top-performing group over time?
Murphy & Cleveland, 1995)
Already argued performance was not normally distributed.
Borman & Bush (1993)
Taxonmoy of manageria performance consisting of 18 dimensions such as planning & organization
Campbell (1993)
Effectiveness
Productivity: ration of effectiveness (output) to the cost of achieving that level of effectiveness (input) (Mahoney 1988)
Pulakos et al. (2000)
Taxonomy of ADAPTIVE performance:
In Study 1, over 1,000 critical incidents
from 21 different jobs were content analyzed to identify an 8-dimension taxonomy of adaptive performance. Study 2 reports the development and administration of an instrument, the Job Adaptability
Inventory,
The goal of this research is to develop a taxonomy of adaptive
job performance along the lines of the job performance model
developed by Campbell et al. (1993).
The 8 components found were:
Handling emergencies or crisis situations - appropriate & proper urgency
Handling work stress
Solving problems creatively - turning probs upsidedown & insideout
Dealing with uncertain
and unpredictable
work situations - not needing things to be black and white; refusing to be
paralyzed by uncertainty or ambiguity.
Learning work tasks,
technologies, and
procedures
Interpersonal adaptability
Cultural adaptability
Physically oriented adaptability
Katz (1964)
Extra-role performance behaviors that contribute to org goals but are outside of the person’s job description.
Podsakoff et al. (1997) found wee related to effectiveness at work.
OCBs
Smith et al., 1983
Organ (1997)
taxonomy of OCBs:
Altruism
Civic virtue: things that support the org and being a good representative outside.
Conscientiousness (thinking ahead)
Courtesy - asking others how a big project is coming along.
Sportsmanship - not complaining, even is justified.
Befort & Hattrup (2003)
More experienced managers seem to appreciate the contribution of OCBs to org performance more.
Murphy & Cleveland (1995)
Practical suggestions for perf management:
Ployhart & Hakel (1998)
With more time at a job, peoples’ performance increases & eventually stabilizes.
THOUGH perf tends to ebb & flow more after stabilizing in more complex jobs (Sturman et al., 2005)
Daniels (1989_
Behaviorist appproach to perf management, focusing on triggers of high perf & operant conditioning
Bracken et al. (1997)
Primitive versions of 360 feedback have been found as far back as the 1950s.
Social Exchange Theory
Blau (1964) (confirmed)
Psychological contracts
Rousseu (1989) (confirmed)
Heilman & Chen (2005) and Kidder & Park & Parks (2001) both found
men doing OCBs percveived more positively - women perceived as just doing their job.
Gooden et al. (2018)
As some minority led organizations, particularly in nonprofit contexts already struggle
Aguinis (2013) on defining perf managment (PM)
Performance management: “continuous process of identifying, measuring, and developing the performance of individuals and teams and aligning performance with the strategic goals of the organization (Aguinis, 2013)”
Schleicher et al. (2019)
Organizations must use their own objectives to identify relevant criteria for evaluating their own PM.
Organizations must evaluate their own PM systems based on more than one criterion category.
Multiple sources should be used by organizations while collecting evaluation data.
Interventions meant to improve PM should utilize different levers at different levels when assessing criteria of interest.
Though recognized as important in past literature, positive reactions are plausibly not as necessary to PM as once thought, especially due to equifinality.
Bledow (2009) Innovation
Antecedents of innovation:
Cultural values re: innovation
Autonomy and control
Transformational leadership
Team Diversity
Trait openness & conscientiousness
Artistic/investigative interests
Positive mood
And - Negative mood
promotes creativity when support for creativity and positive moods are present
because negative mood signals a problematic state of affairs, leading individuals
to systematically address the problem at
hand.
Campbell (1999)
Re: Organ (1997),
among others. Campbell (1999) pointed out that the performance
factors suggested by these authors can be easily integrated as
subfactors into the eight-component taxonomy, forming a hierarchical description of the latent structure of performance. H
Landy & Farr (1980)
How the system is developed is more important than the method used.
In spite of the widespread use of judgmental indices of performance,
there has been a constant dissatisfaction with these measures on the part of
both researcher and practitioner. The source of this dissatisfaction has been
the vulnerability of these measures to both intentional and inadvertent bias.
Sex of the rater does not generally affect ratings, although female
raters may be more lenient. Raters usually give higher ratings to same-race
ratees, although this may be moderated by the degree of contact that members of each race have with each other.
Rater experience appears to positively affect the quality of performance
ratings, but the mechanism or mechanisms responsible (e.g., more training
or experience with the rating form, better observation skills, better knowledge of the job requirements, etc.) is not known.
sex stereotype of an
occupation interacts with the sex of the ratee, such that males receive more
favorable evaluations than do females in traditionally masculine occupations
but that no differences or smaller differences in favor of females occur in traditionally feminine occupations. Ratees tend to receive higher ratings from
raters of their same race, although this may not occur in highly integrated
Raters often give more favorable ratings to same-race ratees,
although situational factors may moderate this effect.
situations
Elmore and LaPointe (1975) reported that student
ratings of college instructors’ effectiveness were positively correlated with
student ratings of instructor warmth.
The number of response categories
available to the rater should not exceed nine.
There is some advantage to using behavioral
anchors rather than simple numerical or adjectival anchors. This advantage
is probably increased in the absence of good dimension definitions.
Ratings for administrative purposes will be more lenient than those for
research purposes..
Rater training has generally been shown to be effective in reducing rating
errors, especially if the training is extensive and allows for rater practice.
Questions still remain about the longitudinal effects of such training,
There appears to be no reduction in halo error when all ratees
are evaluated on one trait, then all ratees are evaluated on the next trait,
and so forth
(Rotundo & Sackett, 2002).
*Typically, job performance is conceptualized as
consisting of three dimensions: in-role or task behavior, organizational
citizenship behavior (OCB), and CWB)
f
O’Boyle & Kroska (2017)
Stars r frequency is alleged to be increasing, owing in part to
the decline of the manufacturing sector and the rise of the knowledge economy (Powell
and Snellman, 2004). This is because knowledge-economy jobs tend to possess features
(e.g., autonomy and multiplicity of production) that lend themselves to star emergence
(Aguinis and O’Boyle, 2014).
e, in high-complexity jobs, the top
percentile of workers more than doubles the productivity of average workers (Hunter et al., 1990)
This possibly indicates that stars provide multiplicative value as
opposed to a simple additive model in contributing to team outcomes.
Adler (1985), another renowned labor economist, contends star status is not
a function, at least not directly, of any internal attribute of the person (i.e., talent). Rather,
star status is bestowed purely by market response. Put differently, market response is
output and vice versa.
. Further supporting normality is that supervisor ratings,
the most commonly used job performance assessment tool (Aguinis, 2013), tend to be
normally distributed.
On the surface there appears to be ample support for the “norm of normality,” but with
regard to stars, two considerations are worth mentioning. First, normality should not be
confused with prevalent or natural. Many things are normally distributed, but most are
not. In practice, normality appears to be the exception rather than the rule (Limpert,
Stahel, and Abbt, 2001), and this is not a particularly recent finding either in the
natural sciences (Groth, 1914; Powers, 1936; Sinnot, 1937) or in the social sciences
(Geary, 1947; Pearson, 1895). Even when limiting the scope to management and applied
psychology research, normality is by no means certain. For example, in a Psychological
Bulletin article, Micceri (1989) reviewed distributions of scores from 440 achievement
and psychometric measures and found that everyone diverged from normality. Micceri
concluded that normal curves were about as common as unicorns.
The second consideration regarding a Gaussian distribution of job performance is that, as previously stated, the compendium of evidence for its normality largely rests on supervisor ratings (Aguinis, 2013). Putting aside the amount of faith one can have in something where nearly half of its measurement is noise (LeBreton, Scherer, and James, 2014; Viswesvaran, Ones, and Schmidt, 1996),
the bigger issue is that this may be an example of reverse causality.
Researchers instruct supervisors to place employees onto normal distributions (e.g.,
Motowidlo and Borman, 1977; Schneier, 1977), and researchers throw out any
performance-evaluation items that fail to generate normal distributions (e.g., Rotundo
and Sackett, 2002). Then, with circular logic, these same data are used to verify that job
performance is normally distributed
First, as expected in a non-normal distribution, the measures of central tendency in a star distribution are always drastically different from one another. The mean is considerably larger than the median, and thus the majority of workers in the distribution are below the arithmetic mean. -The measures of central tendency are distanced from each other—the mean is much larger than the median, and the median is much larger than the mode.
Third, star
distributions exhibit a property known as scale invariance (Eliazar and Klafter, 2007).
This means that the shape of the full star distribution shown in Figure 3.2 will be retained
even if zoomed in on the top quartile, top decile, or even the top percentile—highfrequency peak to the left; long tail to the right.
power-law distribution. The term power is a reference to the
exponential relation between one quantity (e.g., talent) and another quantity (e.g.,
performance and compensation).
The variance is quasiinfinite, making traditional inferential statistics that rely on stable estimates of variance
(e.g., ANOVA, OLS regression, and structural equation modeling) inappropriate and
prone to both Type I and Type II errors (O’Boyle and Aguinis, 2012).
historians typically oscillate between
great-person theory and the trends-and-forces perspective.
there must come a point where small increases in intelligence yield massive
differences in output.
where does the relation between GMA and job performance take on an exponential function (i.e., a tipping point)?
Do certain variables interact differently with cog ability when cog ability is at sig. higher levels?
it is entirely possible that weak or even nonexistent interactions observed
at normal levels of GMA become quite strong when the population is limited to geniuses.
For example, CEOs tend to be high in GMA (Wei and Rindermann, 2015). CEOs also tend
to exhibit higher levels of narcissism and psychopathy compared with the general
population (Babiak, Neumann, and Hare, 2010; Resick, Whitman, Weingarden, and Hiller,
2009). Do these maladaptive personality traits typified by grandiosity and callousness
provide a competitive advantage when coupled with a high IQ.
Ready, Conger, and Hill (2010) posited that high-potential
stars possess such X-factors as a drive to excel, catalytic learning capability (i.e., the
ability to scan for new ideas; the cognitive functioning to integrate new ideas into existing
information and then translate ideas into action), enterprising spirit,
Context antecedents:
Collectively, these insulators and
conductors can be viewed from the perspective of cumulative advantage. Cumulative
advantage can help to explain why even in very efficient markets a widening gap can
emerge between those with resources and those without. Sometimes referred to as the
Matthew Effect (Merton, 1968) or 80-20 Rule, cumulative advantage posits that
small initial advantages in wealth, skills, talent, or just plain luck will build over time to
yield large differences in output and rewards.
two
particularly strong conductors: multiplicity of production and monopolistic productivity.
Multiplicity of production is the extent to which creating additional productivity requires
less effort, resources, etc. than previous productivity.
a star worker is able to capitalize on his or her past popularity
to generate more productivity in the future with less individual effort.
When to be careful about stars growing through monopolistic productivity:
Whether the conductor of star productivity is primarily through multiplicity of production
or primarily through monopolistic production has sweeping implications for an
organization. Consider compensation. If stars are maintaining their status in large part by
siphoning off the production of non-stars (i.e., monopolistic production), then resources
are being misallocated. Star production should be symbiotic, not parasitic. Consider also
selection and retention. If new potential stars recognize that the old, established stars
dominate production through monopolistic practices, then the new stars will seek out
better opportunities—potentially with a competitor. In essence, organizations will lose
their future in order to preserve their past.
Productivity ceiling as a risk-factor to stardom:
In an 8-hour day, the productivity ceiling is eight deals. For nonstars who may only close one or two deals a day, the paperwork is a minor constraint to
their performance because their KSAs limit their production more so than the absolute
ceiling. However, for a star broker closing deals at a fast rate, the one-deal-an-hour limit
is considerably more detrimental. If this broker were to receive an administrative
assistant to do the half-hour of paperwork, then this would double the broker’s
productivity ceiling.
One challenge is whether jobs consist of just one unique task, though, as most don’t.
-*Beck et al. (2014) argued that for most jobs,
performance is multifaceted, and when performance is aggregated across these facets,
the skewed individual facets will become normally distributed.
the issue of facet weight. Are all job facets equally important to the
overarching job performance construct? If not, which are more integral to performance?
To what extent do the weights support or refute star existence?
There is also a need to investigate the degree of overlap between star performers and
dark stars.
How much leeway do
organizations give their stars, and what effect does their differential treatment have on
non-stars?
Bill Gates allegedly said, “A great lathe operator commands
several times the wage of an average lathe operator, but a great writer of software code is worth 10,000 times the price of an average software writer” (Veksler, 2010).
Google’s Senior Vice President of People Operation (i.e., Head of HR), Laszlo
Bock, reflecting on Google’s compensation policy, stated: “two people doing the same
work can have a hundred times difference in their impact, and in their rewards … there
have been situations where one person received a stock award of $10,000, and another
working in the same area received $1,000,000” (Bock, 2015: 241)
how acceptable to an organization is extensive variance in pay for the same
position? An issue well known in professional sports, but largely unheard of in the private
sector, is that stars create the possibility, even the inevitability, that certain employees
will earn more than their supervisors do.
and how much is stars’ star power due to their network and support structure as it is to any internal attributes? (e.g., are they still equally star powered whenw oring in a new team in a new org?)
O’Boyle and Aguinis (2012) found star effects in narrow operationalizations
of CWB in sports (e.g., flagrant fouls, red cards, and ejections).
meta-analytic evidence also
supports a highly skewed distribution of CWB, with the majority of individuals engaged in
virtually no CWB, while a small minority dominates CWB engagement (Greco, O’Boyle,
and Walter, 2015).
Greco et al. (2015)
Despite the alleged frequency of counterproductive work behavior (CWB) in the population, most samples exhibit exceedingly low base rates.
One potential explanation for this incongruence is non-response bias, which leads to range restriction in the CWB distribution.
We investigated this possibility by determining if response rates within CWB research were lower than that found across management research and whether range restriction could explain reduced CWB engagement.
We also examined whether range restriction attenuated CWB’s relations to other variables.
Our primary findings are that reported response rates for studies containing CWB measures are substantially lower (37%) than response rates reported within the general management literature (52.7%), and that range restriction is likely present in the CWB literature resulting in low base rates and attenuated relations to CWB.
In addition, tests of publication bias indicated that low response rates are less likely to be reported, and the true response rate within the CWB literature may be considerably lower.
Wildman et al. (2010)
The PM literature can generally be categorized into three distinct perspectives:
individual-level PM, team-level PM, and organizational-level PM. Very little research has
simultaneously examined multiple levels. This is problematic given that actual
performance in organizations takes place at all three levels simultaneously, and perhaps
more important, all three levels of performance are intertwined.
Simply stated, performance criteria represent whatever aspects of performance a
certain set of stakeholders have identified as critical.
o Errors can be divided into three common categories
▪ Distributional errors: incorrect representations of performance
distributions across employees being evaluated (can occur in rating means
– severity or leniency, or variance – range restriction and central tendency)
▪ Illusory halo: correlations between ratings of two different dimensions
being higher (or lower) than the correlation between the actual behaviors
reflecting those dimensions. Essentially, raters are either overestimating or
underestimating the relationship between dimensions
▪ Other types of errors: perceptual errors (similar-to-me or first impression
Multiple measures or single composite criterion measure
▪ Multiple best if purpose is to diagnose performance issues
▪ Composite best for comparing across units that may not do the same type
of work.
PM useful for research, feedback, training, and evaluation, and these are all at
individual level
o PM also useful for organizational planning by assessing performance at individual
and team levels to inform organizational strategy, strategic workforce planning,
and talent management.
What: Content of measurement
o Based on what level your purpose is at
o Task based needs individual-level data
o Team based needs team-level data, etc.
Authors suggest repeated measures design to capture dynamic nature of
performance.
Do you want day-to-day performance? If so, will getting this data interrupt the
actual performance? Can you create a high-fidelity simulation to measure
performance?
o Do you want maximal or typical performance?
o Beware that your observation itself will likely elicit maximal performance
Campbell’s model limited because focuses solely on task performance and
doesn’t include OCBs
▪ Contextual performance
▪ Adaptive performance
Strategies for measuring individual performance
▪ Performance appraisal
● Dominant view is that supervisor-based PA systems are costly in
time and money and usually only result in negative emotions with
very little demonstrable value (Nickols, 2007). This is because if
PA controlled by supervisor, PA will only serve supervisor’s
opinions.
Nickols (2007) found that PAs often accompanied by periods of
reduced productivity, heightened negative emotional states such as
anxiety, depression, and stress; and lowered morale and
motivation. Also foster a focus on short-term goals at the sacrifice
of long-term goals.
Marks et al. (2001) 10 core team processes that occur throughout the
performance cycle of a team: mission analysis formulation and planning,
goal specification, strategy formulation, monitoring progress toward goals,
systems monitoring, team monitoring and backup behavior, coordination,
conflict management, motivation and confidence building, and affect
management.
Organizational perspective
o Organizational performance: financial outcomes are of most interest (e.g. ROI,
productivity, sales), but customer service and organizational learning are also
important
Future research in performance measurement
o Distribution: globalization, how to measure performance from a distance
o Diversity: how to evaluate fairly across demographic groups
Morgeson et al. (2005)
A review of Morgeson et al. (2005) summarized benefits of 360
feedback, such as increase in information and formal feedback
between employees, increase in management learning,
encouragement of goal setting and skill development, a change in
corporate culture, and improved managerial effectiveness
Waldman et al.
1998
found that bidirectional nature of system encouraged
employees to deliberately sabotage the system with deals on giving
each other high ratings.
Salas et al. (2007)
Team performance: while individuals are the ones behaving, teams have
behavioral interactions and processes beyond those individual behaviors that must
be aggregated and considered; team performance is a bottom-up emergent
process, beginning with individuals and progressing towards teams (Salas et al.,
2007)
Salas et al. (2005)
Big Five in Teamwork: team leadership,
team orientation, mutual performance monitoring, backup behavior, and
adaptability. Also important are shared mental models, closed-loop
communication, and mutual trust
Team adaptability (Burke et al., 2006):
change in team performance
resulting from an identified cue or cue pattern, leading to effective and
efficient outcomes for the entire team; developed the most comprehensive
conceptualization of team adaptation
Review by Ahmed (1999)
● Goal model (Georgopoulos & Tannenbaum, 1971): defines org
effectiveness in terms of org’s achievements of its stated official
goals
o Directly attempts to align PM with org strategy
● System model: views organizations as open systems that work in a
close relationship with the external environment, effectiveness
seen as org’s ability to survive and adapt in a changing
environment
● Internal process model: effectiveness is the extent to which the
internal business processes of an org are smooth, orderly,
continuous, predictable, and minimally conflicting
● Human relations model: person-centered, maintains employee
needs as most important aspect of org effectiveness
● Political approach: uses criteria such as responsiveness,
accountability, representativeness, and adherence to democratic
valuesstrategy
Strategies for measuring org performance
▪ Financial data – but be careful not to engage in short-sighted decisions to
maximize short-term return and sacrifice long-term well-being of firm
Ghalayini & Noble (1996) outline limitations of financial measures
o Missing link between measuring performance in financial
terms and the improvement efforts needed to improve that
financial performance (i.e. how to address individual and
team level performance issues with org-level financial data)
o Do not take into account strategy or goals of org as an
entity (i.e. ignores qualitative goals such as reputation for
customer service)
(Pandey, 2005)
Kaplan & Norton (2005)
Balanced scorecard: system of combining financial and nonfinancial
measures of performance in one single scorecard (Pandey, 2005)
● Kaplan & Norton (2005) found that over 60% of orgs use a
scorecard
Measure financial, customer (customer satisfaction, customer
retention, market share, customer profitability), internal business
processes (quality of business processes in org, key objectives are
process improvement and suppliers’ relations), and learning and
growth (innovation, creativity, capability)
Six Sigma
Six Sigma
● Developed at Motorola as a response to a large loss in productivity
due to a lack of quality (Raisinghani et al., 2005)
● Historically predominantly manufacturing-based, but spreading
rapidly
● Relies on planned changed, team-based collaboration, a focus on
performance improvement, systems perspective, and reliance on
the scientific method and statistical methodology (Jeffery, 2005)
● At first was a statistical method for tracking and reducing
variability in processes, but now more complex operations
improvement and problem solving methodology
Salas et al. (2003)
Multilevel approach to performance measurement is best! (Salas et al., 2003
Aguinis et al. (2016) - CUMULATIVE ADVANTAGE: CONDUCTORS AND
INSULATORS
Results indicate
that higher levels of multiplicity of productivity, monopolistic productivity, job autonomy, and job complexity (i.e., conductors of cumulative
advantage) are associated with a higher probability of an underlying
power law distribution, whereas lower productivity ceilings (i.e., insulator of cumulative.
In addition, higher levels of multiplicity of productivity, monopolistic
productivity, and job autonomy were associated with a greater proportion of productivity stars (i.e., productivity distributions with heavier tails), whereas lower productivity ceilings were associated with a smaller proportion of productivity stars (i.e., productivity distributions
with lighter tails).
Results were the same even while controlling for past experience.
Implications:
Compensation
pay dispersion may be
seen as more acceptable and fair to employees if they are aware that the
distribution has a heavy tail (i.e., a large proportion of productivity stars).
Thus, it may be beneficial to share information on the shape of the productivity distribution with various organizational members.
-but if no extra pay, learning about power curves may turn off stars about their pay.
which types of roles you want to prioritize star performers (power distribution in both best actor and best screenwriter)
Limitations:
There is an empirically documented nontrivial
relationship between performance defined as behavior and performance
defined as results for individuals (Bommer et al., 1995) as well as for
teams (e.g., Beal et al., 2003). Nevertheless, our results are not directly
informative regarding the shape of the distribution when performance is
not defined in terms of results.
The rest below here is mostly for personal growth:
The most direct analog of cumulative advantage
is compound interest, something Albert Einstein is alleged to have once
quipped as “the most powerful force in the universe” (Kay, 2008).
Cumulative advantage is also the result of path dependent change
where a specific sequence of events “creates unequal propensities for
future events” (Gluckler, 2007, p. 620). Crawford (2012) argued that, ¨
over time, positive feedback from the environment (i.e., success) allows
individuals to accumulate intangible resources such as knowledge and
absorptive capacity, which can then be leveraged in later interactions.
Note that past productivity does not necessarily need to increase KSAs
for cumulative advantage to occur. For example, in American football, the
best wide receiver on a team does not need to improve his speed, catching ability, or route accuracy to increase the number of receptions and
touchdowns. His past results will make the quarterback more likely to
pass the ball to him, which will lead to more receptions and touchdowns
independent of any increased KSAs
Similarly, early productivity in an
academic’s career may make other high producing academics more willing to collab with her in future research. These collaborations increase
the likelihood of publication not only because the quality of work itself
may be better but also because the reputation of the researcher might
make a journal editor more inclined to accept the paper for publication
(Peters & Ceci, 1982).
Multiplicity of productivity (past success helping facilitate future success:
, due to increased opportunities to
perform (e.g., more time to devote to research due to decreased teaching
demands), positive feedback from the environment (e.g., accepted publications), and an increased network of collaborators and resources due
to past successes (e.g., better access to data collection opportunities and
computing equipment),
*Job characteristics
We focus on three such job characteristics: job autonomy and job complexity as hypothesized conductors and productivity ceiling as a hypothesized
insulator.
Empirically, job autonomy generally has a positive relation with productivity (Humphrey, Nahrgang, & Morgeson, 2007). Job autonomy is an
especially salient conductor because it offers high-productivity individuals the flexibility and control over processes that may lead to stratification
of individuals’ output levels (Kohn & Schooler, 1983).
and because stars are
better able to leverage available resources (Aguinis & O’Boyle, 2014)
and autonomy allows stars to build stronger and larger networks, which are known
to enhance success and generate extreme productivity levels (Crawford
& LePine, 2013; Oliver & Liebeskind, 1998; Zucker, Darby, & Brewer,
1998)
One reason is that, similar to the rationale for job autonomy, job complexity introduces more variance in worker output (Gerhart, 1988; Hunter, Schmidt, & Judiesch, 1990).
This productivity ceiling is particularly noticeable
for jobs that have a physical and/or time limit component
Even different parts of a job have different productivity ceilings: Example:
# of sales calls made per day vs $ made per day - If, for example, a $1,000 sale takes about the same amount of
time as a $100 sale, then revenue exhibits a higher ceiling to productivity
