ArticlePDF Available

Working Hard and Working Smart: Motivation and Ability During Typical and Maximum Performance

Authors:

Abstract

The distinction between what people can do (maximum performance) and what they will do (typical performance) has received considerable theoretical but scant empirical attention in industrial-organizational psychology. This study of 138 participants performing an Internet-search task offers an initial test and verification of P. R. Sackett, S. Zedeck, and L. Fogli's (1988) model of typical versus maximum performance: Motivation--in the form of direction, level, and persistence of effort exerted--rose significantly under the maximum performance condition. Consequently, the correlation between motivation--in the form of direction and level of effort--and performance diminished, whereas the correlation between ability--in the form of declarative knowledge and procedural skills--and performance increased under the maximum performance condition. Overall, results confirm the general propositions of the model. Implications for the generalizability of these findings, theory, practice, and directions for future studies of typical and maximum performance are discussed.
Working Hard and Working Smart: Motivation and Ability During Typical
and Maximum Performance
Ute-Christine Klehe and Neil Anderson
University of Amsterdam
The distinction between what people can do (maximum performance) and what they will do (typical
performance) has received considerable theoretical but scant empirical attention in industrial–
organizational psychology. This study of 138 participants performing an Internet-search task offers an
initial test and verification of P. R. Sackett, S. Zedeck, and L. Fogli’s (1988) model of typical versus
maximum performance: Motivation—in the form of direction, level, and persistence of effort exerted—
rose significantly under the maximum performance condition. Consequently, the correlation between
motivation—in the form of direction and level of effort—and performance diminished, whereas the
correlation between ability—in the form of declarative knowledge and procedural skills—and perfor-
mance increased under the maximum performance condition. Overall, results confirm the general
propositions of the model. Implications for the generalizability of these findings, theory, practice, and
directions for future studies of typical and maximum performance are discussed.
Keywords: typical performance, maximum performance, motivation, ability
The distinction between typical and maximum performance
holds wide-reaching practical and theoretical implications for per-
formance assessment (e.g., Guion, 1991, 1998), personnel selec-
tion (e.g., Ackerman & Humphreys, 1990; Arvey & Murphy,
1998; Borman, 1991; Boudreau, 1991; Campbell, 1990), and train-
ing (e.g., Dewberry, 2001; Smith-Jentsch, Salas, & Brannick,
2001). Yet empirical research on this distinction is still scarce
(Klehe & Anderson, 2005; Klehe, Anderson, & Viswesvaran,
2007; Sackett, 2007), and research validating fundamental as-
sumptions of the typical–maximum performance distinction is as
yet unavailable.
When applying Cronbach’s (1960) distinction between typical
and maximal predictors of performance to the criterion domain,
Sackett, Zedeck, and Fogli (1988) argued that the primary differ-
ence between performance during short, evaluative situations,
when people show what they can do (maximum performance), and
performance during nonevaluated, day-to-day situations, which
reveal what people will do (typical performance), lies in the role of
motivation. Whereas motivation varied during typical performance
situations, it was constrained to be high during situations of max-
imum performance. The purpose of this study is to present an
initial detailed examination of Sackett et al.’s propositions regard-
ing the roles of motivation and ability during periods of typical and
maximum performance conditions and thus to provide an empiri-
cally solidified foundation for research on this important distinc-
tion.
Typical and Maximum Performance
Job performance, the behaviors relevant to the goals of the
organization (Campbell, 1990; McCloy, Campbell, & Cudeck,
1994), is a function of the performer’s ability and motivation
(Locke, Mento, & Katcher, 1978; Maier, 1955). Campbell (1990)
classified ability into (a) declarative knowledge, the ability to state
the relevant facts and things, and (b) procedural knowledge and
skills, the knowledge attained when knowing what to do (i.e.,
declarative knowledge) has been successfully combined with
knowing how to do it. Motivation, Campbell argued, is the com-
bined effect of three choices: (a) the choice to expend effort
(direction), (b) the choice of which level of effort to expend
(level), and (c) the choice to persist in the expenditure of that effort
(persistence). Campbell further argued that performance on any
task requires a minimum of both ability and motivation. The
impact of ability versus motivation on performance, however, can
vary across situations.
Sackett et al. (1988) introduced a continuum spanning from
typical to maximum performance situations as one instance of such
variation. During situations of typical performance, performers (a)
are relatively unaware that their performance might be observed
and evaluated, (b) are not instructed to perform their best, and (c)
have their mean performance observed over an extended time
period. Consequently, performers may choose to focus on their
work or not (direction), to invest their full or just some partial
amount of effort (level), and, over time, to maintain that level of
effort or to reduce it (persistence). “In a typical performance
Ute-Christine Klehe, Department of Work and Organizational Psychol-
ogy, University of Amsterdam, Amsterdam, the Netherlands; Neil Ander-
son, Amsterdam Business School, University of Amsterdam.
This research was funded by German Academic Exchange Service
Grant D/02/000857 to Ute-Christine Klehe during her postdoctoral studies
at the University of Amsterdam. We thank Wolfgang Grund (affiliated with
H.A.S.E. GmbH) for his generous contribution of the necessary software
and support during data analysis, Wouter Boendermaker for his assistance
during data collection, and Martin Kleinmann for his valuable comments
on an earlier version of this article.
Correspondence concerning this article should be addressed to Ute-
Christine Klehe, Arbeids- & Organisatiepsychologie, Universiteit van Am-
sterdam, Roetersstraat 15, 1018 WB Amsterdam, the Netherlands. E-mail:
u.klehe@uva.nl
Journal of Applied Psychology Copyright 2007 by the American Psychological Association
2007, Vol. 92, No. 4, 978–992 0021-9010/07/$12.00 DOI: 10.1037/0021-9010.92.4.978
978
setting, choices about time on task, level of effort, and persistence
of effort are less constrained” (DuBois, Sackett, Zedeck, & Fogli,
1993, p. 206).
In contrast, maximum performance situations are characterized
by (a) performers’ explicit awareness of being evaluated, (b) their
awareness and acceptance of instructions to maximize effort, and
(c) a short enough time duration to enable performers to keep their
attention focused on the task. Sackett et al. (1988) and DuBois et
al. (1993) argued that these characteristics force motivation to be
high in maximum performance situations: Direction of effort is
constrained by individuals’ knowledge of being monitored.
DuBois et al. argued, “Unless one is inviting disciplinary action,
one has little choice but to expend effort on the task in question”
(p. 206). The level of effort in these experiments was high, because
individuals were aware of and accepted the instruction to expend
effort. Persistence was not demanded during maximum perfor-
mance situations, as performance was only observed for a period
brief enough that remaining focused on the task should not have
been difficult (see also Sackett, 2007).
Of course, typical and maximum performance represent a con-
tinuum (Sackett et al. 1988), making any comparison between
assessments relative rather than categorical. Consequently, any
reference to a typical performance situation needs to be understood
as being relative in comparison with a situation that is located
closer toward the maximum end of the continuum, and vice versa.
This given, Sackett et al.’s assumptions have considerable conse-
quences for the impact of knowledge and skills on performance.
Sackett et al. argued that, given high motivation, performance
under maximum performance conditions primarily reflects per-
formers’ abilities. Consequently, work-related abilities should cor-
relate higher with performance under maximum than under typical
performance conditions. Accordingly, DuBois et al. (1993) found
general mental ability to be a slightly better predictor of the
maximum than of the typical speed with which supermarket cash-
iers processed goods. Klehe and Latham (2006) found structured
selected interviews to be better predictors of typical than of max-
imum team playing performance among master’s of business ad-
ministration students. Ployhart, Lim, and Chan (2001) found that
openness to experience was primarily linked to maximum and not
to typical transformational leadership among Singaporean military
recruits. More recent studies by Marcus, Goffin, Johnston, and
Rothstein (2007), ForsterLee (2007), Ones and Viswesvaran
(2007), as well as Witt and Spitzmu¨ller (2007) largely replicated
and extended these findings to other types of predictors.
None of these studies, however, tested the underlying
assumption that motivation rises under maximum performance
conditions or presented a comparison of performance scores
across typical and maximum performance conditions. Campbell
(1990), for example, questioned whether the difference between
typical and maximum performance was solely a function of
motivation. Kirk and Brown (2003) found significant relation-
ships between performance on a (maximum performance) walk-
through performance test and post hoc measures of the moti-
vational predictors work domain self-efficacy and need for
achievement, thus questioning the held-true but never tested
assumption that motivation is constrained to be high during
maximum performance situations.
Hypotheses
Past research has measured either motivation or ability prior to
(e.g., Klehe & Latham, 2006; Ployhart et al., 2001) or after (Kirk
& Brown, 2003) task performance. Yet one must assess motivation
and procedural skills simultaneously with performance to test
Sackett et al.’s (1988) underlying assumptions concerning the
development of direction, level, and persistence of effort across
both typical and maximum performance periods. We therefore
hypothesized the following:
Hypothesis 1: Participants’ average motivation will be higher
under maximum than under typical performance conditions:
(a) Direction of effort will be more task related, (b) level of
effort will be higher, and (c) level of effort will sink less
throughout the maximum performance period, compared with
times of typical performance.
Given the reasonable assumption that knowledge and procedural
skills remain stable across performance situations and given that
performance results from knowledge, skills, and motivation
(Campbell, 1990), an increase in motivation during a maximum
performance period should also lead to an increase in the resulting
performance. As with Hypothesis 1, hardly any of the past studies
comparing typical and maximum performance (Klehe & Latham,
2006; Ployhart et al., 2001; Sackett et al., 1988) reported any
findings regarding this fundamental proposition.
Hypothesis 2: Participants’ performance will be higher under
maximum than under typical performance conditions.
Furthermore, it has to be tested whether the difference between
typical and maximum performance is truly a function of changed
motivation and not due to factors such as changes in participants’
procedural skills. More precisely put, a drop in motivation from
the maximum to the typical performance condition should account
for the respective drop in performance.
Hypothesis 3: Varying performance under typical versus
maximum performance conditions will be primarily due to
changes in performers’ motivation.
Given that maximum performance conditions force motivation to
be high, whereas typical performance conditions place fewer con-
straints on direction, level, and persistence of effort, Sackett et al.
(1988) argued that measures of motivation should be better pre-
dictors of performance under typical than under maximum perfor-
mance conditions.
Hypothesis 4 (a, b, and c): Motivation will correlate higher
with performance under typical performance conditions than
under maximum performance conditions: Under typical per-
formance conditions, (a) direction, (b) level, and (c) persis-
tence of effort will correlate higher with participants’ perfor-
mance than they will under maximum performance
conditions.
Besides these very proximal indicators of motivation, the same
should hold true for more distal motivational constructs that influ-
ence the direction, level, and persistence of effort. Two variables
that appear to be of particular interest in this respect are partici-
979
WORKING HARD AND WORKING SMART
pants’ self-efficacy in handling the task at hand and their intrinsic
enjoyment of the task.
Self-efficacy, the perceived ability to master a particular task, is
one of the most established motivational predictors of direction,
level, and persistence of effort studied in industrial and organiza-
tional psychology (Bandura, 1997). Self-efficacy is of particular
interest for the study of typical versus maximum performance
given Kirk and Brown’s (2003) countertheoretical finding of a
significant relationship between self-efficacy and performance un-
der maximum performance conditions. However, given that per-
formers’ self-efficacy is based in part on individuals’ self-
assessments of their relevant abilities, such a correlation may be of
little surprise. Also, given the continuous nature of typical versus
maximum performance, a single correlation between the motiva-
tional construct self-efficacy and performance under either typical
or maximum performance conditions is no adequate basis for
challenging the theoretical propositions underlying the typical–
maximum performance distinction. Consequently, we propose that
although participants’ task-related self-efficacy may correlate sig-
nificantly with their performance under a maximum performance
condition, this correlation is likely to be smaller under maximum
performance than under typical performance conditions.
A second motivational predictor that appears particularly rele-
vant in the context of the current study is task valence, the degree
to which individuals are interested in the task and in engaging
further in it (e.g., Freitas, Liberman, Salovey, & Higgins, 2002).
This concept is relevant for the study of typical versus maximum
performance situations insofar as it represents participants’ intrin-
sic motivation to engage in the task, as opposed to the extrinsic
motivation initiated by a maximum performance situation.
1
It is
likely that performers will differ in their default interest and
engagement in the task. Under typical working conditions, this
differing interest should then influence performers’ subsequent
performance. However, if Sackett et al.’s (1988) assumptions are
correct and motivation is high across individuals during maximum
performance situations irrespective of whether they enjoy the task,
the impact of task valence on performance should diminish as soon
as performers enter a maximum performance situation.
Hypothesis 4 (d and e): (d) Task-related self-efficacy and (e)
task valence will correlate higher with participants’ perfor-
mance under typical than under maximum performance con-
ditions.
Given that performance is a function of (a) motivation, (b) declar-
ative knowledge, and (c) procedural skills (Campbell, 1990) and
that the relative impact of variance in motivation on performance
declines during maximum performance situations, the relative
impact of ability on performance should grow under maximum
performance conditions, in the form of both (a) declarative knowl-
edge in the content and task domain and (b) the procedural skills
used to accomplish the task—that is, the degree to which people
“work smart.” Consequently, we hypothesized the following:
Hypothesis 5: (a) Declarative knowledge and (b) procedural
skills will correlate higher with participants’ performance
under maximum than under typical performance conditions.
Finally, when we switch perspectives away from the predictors
toward the criteria, the above assumptions further suggest that
performance under maximum performance conditions should be
limited primarily by participants’ ability, or what they can do, and
less by their motivation to actually do it (Sackett et al., 1988).
Typical performance, the average performance under ongoing
work conditions, places fewer constraints on direction, level, and
persistence of effort and should, rather, assess what people will do
(Sackett et al., 1988). Consequently, maximum performance
should be primarily a function of ability, whereas typical perfor-
mance should be a function of both ability and motivation.
Hypothesis 6: The best predictors of performance under max-
imum performance conditions will be measures of task-
related knowledge and procedural skills, whereas the best
predictors of performance under typical performance condi-
tions will include measures of motivation as well.
Method
In order to test the above hypotheses, the study had to meet three
prime criteria: First, an analysis of the proposed hypotheses re-
quired a setting in which to observe the development of procedural
skill, direction, level, and persistence of effort over both typical
and maximum performance periods. Second, to ensure internal
validity, these performance periods needed to be comparable in
content and specificity of the task and exclude any possible con-
founds to the manipulation (Sackett et al., 1988). Third, the task
should offer a certain degree of external validity. Primarily the first
two considerations caused us to conduct the study in an experi-
mental setting. We maximized ecological validity by using the
realistic task of comparing product prices via the Internet, a task
that is common in many jobs, particularly administrative– clerical
ones.
Sample
One hundred fifty student volunteers at a university psychology
department received research points or a small payment for par-
ticipation. Participants’ average age was 23.3 years, and 68% of
them were female. Data from 4 participants could not be used
because of technical problems, and 8 participants left the experi-
ment prematurely, arguing that the task was too strenuous (3
participants) and/or too boring (6 participants; 1 participant named
both reasons). Hence, results are based on 138 participants. Of
these, 78% reported using the computer on 3 days or more per
week; mean usage was 1.5 hr per weekday. Seventy-seven percent
of the sample reported using the Internet on 3 days or more per
week. Mean usage was 1 hr per day. All participants reported
having used the Internet for 6 months minimum, and 89% of them
reported having used it for 2 years or more.
Procedure
All participants underwent the same procedure. After sitting
down in front of a conventional personal computer, participants
learned the following:
1
We thank an anonymous reviewer for raising this point.
980
KLEHE AND ANDERSON
This study compares the searcher friendliness of different approaches
to present information on the Internet. Imagine you are working for a
hardware retailer. Part of your job is to find out the prices of your
competitors to ensure that your company keeps competitive prices.
Participants were then given the Web addresses of two genuine
competing hardware retailers as well as general information on the
structure of the retailers’ Web sites. We used the above wording to
prime participants to the task to invoke some feelings of being
responsible for comparing prices and also to ensure some degree of
ecological validity to job performance situations in which such
tasks and duties are involved. While participants searched both
Internet sites for 10 min, the experimenter remained in the room to
answer questions and to ensure that participants sufficiently un-
derstood the task. Next, participants filled out online question-
naires regarding task valence and self-efficacy, followed by a
multiple-choice knowledge test. Then, participants searched the
Web sites of both competitors for products for 1.5 hr and then
answered a number of demographic questions. Finally, participants
were debriefed and rewarded. In sum, the total experiment took
about 2.5 hr per participant. Half of the data were collected by
Ute-Christine Klehe, and half were collected by a research assis-
tant ignorant of the study’s purpose.
Task
The items to search for were actual products sold at the time of
the experiment at both hardware retailers. Items to search for were
indicated one at a time in a separate program window, and partic-
ipants’ task was to type in the respective prices found in the two
online catalogues. Different groups of products (e.g., processors,
printers) were evenly distributed across the search list, with no two
subsequent products belonging to the same product group. How-
ever, participants could click forward and backward through the
search list if they wanted to find the next product of the same
product group. The list included 103 products, more than pretest
participants had solved within 90 min.
Justification
The laboratory setting and specific task were chosen for three
reasons: First, this closely defined task, undertaken in a laboratory
setting, allowed for the collection of comparable data on typical
and maximum performance in that it reduced the number of
possible confounds (e.g., varying task demands, settings, leader–
member exchange differences) likely to influence results in most
field settings. Moreover, it allowed the simultaneous assessment of
the procedural motivational and ability-related variables underly-
ing this performance, a requisite if we were to study the above
hypotheses but virtually impossible to achieve in less controlled
field settings.
Second, we deemed it necessary to choose a task that would be
easy to learn yet attentionally demanding in that it required high
motivation in order to sustain the attention to detail needed to
accomplish this task well. Quick learning of the task was important
to ensure that changes in performance were not due to continuous
learning (e.g., Ackerman, Kanfer, & Goff, 1995). Pretest data for
the current study indicated that participants understood the nature
of the task and how to solve it within 5 to 10 min. Furthermore, the
experimenter present during the 10-min practice trial ensured that
participants were able to master the task.
In addition, the task needed to be attentionally demanding and
tiring to ensure that participants were quickly past an initial en-
thusiasm. During the process of socialization into a job and an
organization, individuals oftentimes experience a “honeymoon,”
followed by a realization of the specific job’s disadvantages and a
potential drop in their motivation (e.g., Fichman & Levinthal,
1991; Louis, 1980). Although a laboratory setting does not allow
for the observation of individuals’ performance over weeks and
months, it was important to choose a task that would “wear out”
participants rather quickly, challenging their initial motivation in
the task. Because access to the catalogues’ quick-search options
were blocked, searching demanded back-and-forth-clicking
through diverse catalogue segments, remembering each product’s
specifics, and paying high attention to detail to distinguish between
distinct but similar-sounding products (e.g., Philips Brilliance
170B2T 17-in. TCO 99 monitor vs. Philips Brilliance 170B2Y
17-in. TCO 99 monitor). The fact that 8 participants left the
experiment prematurely, forgoing their rewards (a rather unusual
effect at the university where this study was conducted), indicates
that the current task met this requirement. That said, the assess-
ment of typical performance was likely to grow more valid the
longer the experiment carried on, as any of the abovementioned
effects was more likely to have worn off. Conversely, this implies
that comparisons between earlier assessments of typical perfor-
mance and assessments of maximum performance are likely to
yield more conservative estimates of the impact of typical versus
maximum performance situations on performance than do com-
parisons with later assessments of typical performance.
2
Third, as is true for every laboratory study (as well as every field
study; Sackett & Larson, 1990), the chosen operationalization
needed to guarantee a certain level of external validity. Comparing
prices across competitors is a very common task undertaken for
private and business purposes. The task involved Internet search-
ing, checking, and comparing specific details, all of which are
commonly found in clerical and administrative jobs. Thus, the
search task required high attention to detail and endurance, which
are important in many administrative and technical jobs (e.g.,
Chatman, 1991). Participants’ current level of education (high
school to undergraduate degree) and the frequent reported usage of
the Internet indicate that participants possessed the basic skills
needed to accomplish the task. Also, about 80% of students
worked in addition to their studies, frequently in retail or low-level
administrative jobs (LSVb, 2004). Regarding the stimulus materi-
als, the two catalogues represented the actual online product in-
formation of two major hardware retailers, thus giving further
ecological validity to the study’s findings.
To further enhance the realism of the situation, we constructed
the experiment so that participants’ working conditions were com-
parable to those in many organizations. After receiving instruc-
tions, participants knew that the experimenter would not return
soon. The computer they worked on was a usual work station,
equipped with the standard desktop programs (including games)
and with essentially unlimited access to the Internet. No control
mechanisms were installed to ensure that participants did not use
2
We thank an anonymous reviewer for raising this point.
981
WORKING HARD AND WORKING SMART
the Internet for private purposes or spend their time with other,
task-unrelated activities (e.g., phone calls, reading, Web chatting).
Measures
Typical and maximum performance. Performance was mea-
sured as the average time needed to correctly report a price for
either hardware retailer. Most of the observation period was used
to induce typical performance. During this time, participants were
not made aware that they were being observed and evaluated, nor
were they instructed to do their best. During this time, they were
alone in the room. Such manipulation is in line with Sackett et al.’s
(1988) definition, as (a) participants were not treated as if under
close evaluation, (b) they were not implicitly or explicitly in-
structed to invest their full effort, and (c) the duration was, al-
though not as long as a usual workday, long enough to strain
participants’ persistence.
A 5-min maximum performance situation was induced after
participants had performed the task for about 45 min. At that time,
the experimenter returned to the room and watched participants
work on the task. If asked, the experimenter merely told partici-
pants not to feel disrupted and that he or she was “just checking
how you’re doing.” The experimenter remained in the room,
obviously observing participants’ behavior, for 5 min, then left
participants alone until completion of the experiment. Relative to
the assessment of typical performance, these 5 min are likely to be
a valid assessment of maximum performance, given that (a) the
evaluative nature of the observation was obvious and (b) partici-
pants were implicitly instructed to focus on the task. Finally, (c)
the duration of the evaluation period was rather short.
Self-efficacy. Self-efficacy was assessed in regard to both the
content (computer) and the means (Internet) of the task. Computer
self-efficacy was assessed with a shortened version of the Com-
puter Self-Efficacy Scale (Murphy, Coover, & Owen, 1989) and
included 18 items, such as “I feel confident understanding terms/
words relating to computer hardware.” To meet Bandura’s (1997)
recommendation that measures of self-efficacy be task specific, we
deleted from the scale items that were irrelevant for the content of
the current experiment (e.g., ”I feel confident handling a floppy
disk” and “I feel confident using the printer to make a ‘hard-copy’
of my work”). Participants scored their answers on a 5-point Likert
scale ranging from 1 (totally disagree)to5(totally agree). Scale
internal consistency (Cronbach’s alpha) was .95.
For the same reasons (Bandura, 1997), Internet self-efficacy was
assessed with 15 items selected from two published scales on the
basis of their relevance to the search task. Six items were taken
from Eastin and LaRose (2000), and nine were taken from Joo,
Bong, and Choi (2000). Included were items such as “I feel
confident using the Internet to gather data” and “I feel confident
linking to desired screens by clicking.” Excluded were task-
unrelated items such as “I feel confident turning to an on-line
discussion group when help is needed.” Participants scored their
answers on a 5-point Likert-scale ranging from 1 (totally disagree)
to5(totally agree). Scale internal consistency was .92.
Task valence. Freitas et al. (2002) assessed task valence
through three items asking participants about their level of interest,
fun, and anticipated success in the experiment. As we considered
anticipated success to be a measure of self-efficacy rather than task
valence, we dropped this aspect from the scale and assessed task
valence through five statements ( “I am looking forward to search-
ing for further products,” “I am interested in the products,” “Learn-
ing about the competitors’ prices is fun,” “Learning about the
competitors’ prices is interesting,” and “I would like to know
more about the products”) developed and pretested for this
experiment. Participants scored their answers on a 5-point Lik-
ert scale ranging from 1 (totally disagree)to5(totally agree).
Cronbach’s alpha was .83.
Motivation/working hard. In accordance with Humphrey, Hol-
lenbeck, Ilgen, and Moon (2004), the current study included indi-
cators of how hard and how smartly individuals worked on the
task. While participants worked on the task, all their actions were
recorded and recoded into indicators of direction, level, and per-
sistence of effort.
Direction of effort—that is, whether individuals engage in the
task (Campbell, 1990)—was assessed by the percentage of time
participants spent working on the task itself. Whenever partici-
pants selected a program or Web page not related to the task or
whenever they stopped moving on any screen for more than 30 s
(the maximum time needed in pretests to scan the relevant content
of a page; this can happen whenever individuals do something
completely unrelated to the experiment— e.g., make phone
calls— or when they allow themselves to get distracted and start
reading extended product descriptions unnecessary for the task at
hand), the time was deducted from their score. Internal consistency
of the direction measure ranged from .70 to .74 during the three
assessment periods.
Level of effort was assessed through the number of task-related
clicks per minute. This measure resembles Humphrey et al.’s
(2004) overall measure of working hard, which included the num-
ber of task-related actions taken within the given time frame of the
experiment. Internal consistency ranged from .67 to .78 during the
three assessment periods.
Persistence is the degree to which level of effort is sustained
over time (Campbell, 1990). To assess the development of level of
effort over time, we sampled each participant’s level data at the
beginning, at the end, and after three equally long intervals during
each evaluation period. For the two typical performance periods,
such a measure entailed samples from the 4th, 14th, 24th, 34th, and
44th minutes, whereas it represented all of Minutes 1 to 5 of the
maximum performance period. On the basis of the within-subject
correlation between time of assessment (Measurement Points 1 to
5 within each measurement period) and this participant’s level of
effort at that time, we used each participant’s linear regression
weight of level of effort over time as a measure of his or her
persistence. An index around or above zero indicates that individ-
uals maintained or even increased their level of effort over the
evaluation time and thus represents high persistence. A negative
index, in contrast, indicates a decrease in effort over time and,
hence, lower persistence.
Declarative knowledge. To test participants’ declarative
knowledge regarding the content and means of the task, we de-
veloped and pretested a 13-item multiple-choice test for this study,
in collaboration with a certified computer engineer who was
knowledgeable about computer retailing. Items focused on specif-
ics of the search task, such as the following:
Which of the following is NOT true about subject trees? (a) Subject
trees contain links to every existing Web page (b) Another name for
982
KLEHE AND ANDERSON
one is a directory (c) Subject trees are organized hierarchically (d)
Different sites support different subject trees.
Other items asked for the meaning of particular features fre-
quently mentioned in the search task, such as the following
What is TCO regarding computer hardware? (a) A Windows compat-
ibility standard (b) A pollution norm issued by the Swedish Confed-
eration of Professional Employees (c) A measure of the Total Cost of
Ownership (d) Total Clear Objective.
The declarative knowledge assessed in the first type of question
tests participants’ basic understanding of the task they are facing.
For example, a successful search requires participants to see that
the relevant information is organized (c) hierarchically yet (d)
differently across the two online catalogues and that one (a) cannot
retrieve information about one retailer by following the subject
tree of the other. The second type of item is relevant for partici-
pants as it represents a mental map that facilitates the understand-
ing, chunking, and recognition of relevant information. For exam-
ple, a participant who knows that TCO stands for (b) a pollution
norm issued by the Swedish Confederation of Professional Em-
ployees will likely also know that this norm was issued at some
time (1995) and adapted a few years later (1999). Consequently,
for this participant, search tasks including information on TCO
(e.g., Belinea 101556 15-in. [38.1 cm] TCO 99 monitor) will be
simplified in that he or she knows to look for one of two possible
values instead of having to memorize the respective information as
an incoherent string of letters. The test’s overall internal consis-
tency was .69.
Procedural skills/working smart. In accordance with Hum-
phrey et al. (2004), we used a measure of how efficiently individ-
uals completed the task as a measure of working smart. Although
the objective of the task was explained to participants, they did not
receive any instructions as to how best to approach the task. In line
with Humphrey et al.’s measure of inefficient actions taken to
accomplish the task, we assessed procedural skills by evaluating
the efficiency of the search strategies participants used when
performing the task. The complexity of the task grows out of two
sources: (a) the products’ specifications and (b) the different
approaches the two hardware retailers use for presenting their
information on the Internet. Thus, a good search strategy is one
that attempts to reduce complexity by altering each search’s spec-
ificities as little as possible from the last search conducted, as any
change in either product group searched for or hardware retailer
searched at will cause additional complexity to the task and will
cost participants cognitive resources and time.
We evaluated the search strategy used during every new search.
Searching for the same product (e.g., Toshiba SD-W2002RAM
DVD drive) with the same retailer as a participant had done in his
or her last search (e.g., Retailer A) was not possible, as participants
should already have found and noted the respective price. The two
strategies closest to this, and thus the smartest, were either (1) to
search for the same product at the other retailer (e.g., search for
Toshiba SD-W2002RAM drive at Retailer B), as this allowed
participants to keep focused on the same particular product, or (2)
to stay with the same retailer (Retailer A) as in the last search but
search for a different product from the same product group (an-
other DVD drive). Although this strategy demanded that partici-
pants skip about a handful of products in the search list before
arriving at the next product from the same product group, it was
still an advisable strategy, as it demanded only minimal alterations
to participants’ last search and no or few changes in the catalogue
page that they were searching (e.g., they could stay on the Web
page listing DVD drives and their prices instead of moving upward
to the more general section of the catalogue, only to move down to
another particular product section— e.g., printers—thereafter and
reorient themselves within that page).
Less advisable was a strategy that either (3) changed the product
group while sticking with the same retailer or (4) changed the
retailer while staying in the same product group, although not with
the same product. In the former case, mostly when searching for
the next item on the search list (e.g., Epson AcuLaser EPL-6100N
printer), participants would have to adapt their search for the new
product group’s specifications, focus on different product details,
click their way up and down through the respective retailer’s
catalogue to a different page, and reorient themselves within that
page. In the latter case, participants could search within the same
product group as they had focused on during their last search (e.g.,
DVD drives) but would not carry major benefits from this, as they
would have to switch to the alternative retailer (e.g., Retailer B),
which used a widely different Web design and organization of
material.
Finally, (5) participants could search for a product that was not
part of the same product group, and they could do so at the
alternative retailer. As this strategy demanded the most changes
between the latest and the current search, it was the least advisable.
Additionally, it was wise to use (6) more than one window of the
Web browser for searching to facilitate and speed the search. This
is because the usage of two or more browser windows saves
individuals time otherwise needed to reconnect with the respective
retailer and enter the online catalogue.
Procedural skills, or working smart, were calculated from the
above six strategic components, which had been proposed and
validated through the pilot test. Because these strategies are largely
interdependent, they were combined into one measure. As a strat-
egy grew more advisable the more the product or product group
and the retailer remained the same from one search to the next, we
weighted different strategies according to their usefulness by add-
ing half a point for every aspect of the search that remained the
same: working smart 2 (Strategy 1 or 2) 1.5 (Strategy
3or4) 1 (Strategy 5 or 6). Internal consistency for working
smart was .69 to .75.
Results
Experimenter Effects
No differences emerged between the data collected by Ute-
Christine Klehe and the data collected by the research assistant,
who was uninformed about the experiment’s purpose, F(137)
1.87. This indicates that data were unaffected by the person col-
lecting them.
Hypothesis 1: Working Hard
Hypothesis 1 stated that motivation in the form of (a) direction,
(b) level, and (c) persistence of effort would be higher under the
maximum performance condition than under typical performance
983
WORKING HARD AND WORKING SMART
conditions. Means, standard deviations, correlations, and internal
consistencies of all study variables are reported in Table 1.
Direction. During the maximum performance period, partici-
pants focused 82% of the time on the task, compared with 70% and
59% during the first and second typical performance periods. A
repeated-measurement analysis of variance (ANOVA) revealed a
significant difference, F(2, 128) 57.85,
2
.48, p .001.
Bonferroni-corrected pairwise comparisons revealed that partici-
pants’ direction of effort on the task was significantly higher
during the maximum performance period than during both the first
(d 6.65) and the second (d 10.90) typical performance periods
(both ps .001). In addition, participants focused more on the task
during the first, compared with the second, typical performance
assessment (d 5.80, p .001).
Level. During the maximum performance period, participants,
on average, made 10.63 task-related clicks per minute, compared
with 9.54 clicks during the first and 7.87 clicks during the second
assessment of typical performance. A repeated-measurement
ANOVA revealed a significant difference, F(2, 122) 15.54,
2
.20, p .001. Bonferroni-corrected pairwise comparisons re-
vealed that participants’ level of effort on the task during the
maximum performance period lay significantly above their level of
effort during the first (d 2.95, p .01) and the second typical
performance periods (d 5.58, p .001). In addition, participants
invested more effort in the task during the first, compared with the
second, typical performance assessment (d 3.68, p .001).
Persistence. A 3 (performance condition) 5 (persistence)
repeated-measurement ANOVA revealed significant effects for
both persistence, F(4, 126) 11.60,
2
.27, p .001; and
performance condition, F(2, 128) 13.47,
2
.17, p .001; as
well as for the interaction between these two factors, F(8, 122)
7.64,
2
.33, p .001 (see Figure 1
). Trend analyses (Stevens,
2002) revealed that the decline in effort during the first typical
performance period followed a linear, F(1, 129) 6.51,
2
.05, p .05, or a quadratic trend, F(1, 129) 4.35,
2
.03,
p .05. The second typical performance period showed a
comparable, although stronger, decline of effort in the form of
a linear, F(1, 129) 65.88,
2
.34, p .01, or quadratic
trend, F(1, 129) 12.99,
2
.09, p .01. No such trends
were found during the maximum performance period, Fs(1, 129)
0.07– 0.54,
2
.00, ns.
Working smart. Participants scored, on average, 1.29 and 1.37
on the measure of working smart during the two typical perfor-
mance periods, respectively. During the maximum performance
period, they scored, on average, 1.20 on this measure. A repeated-
measurement ANOVA showed a significant difference, F(2, 123)
6.40,
2
.10, p .01. Bonferroni-corrected pairwise com
-
parisons revealed that participants worked smarter during the
second typical performance period than during the maximum
performance period (d 3.49, p .01), with no significant
differences emerging between the other comparisons.
Taken together, the results confirm Hypothesis 1: Motivation, in
the form of direction, level, and persistence of effort, was higher
during the maximum performance period than during either of the
two typical performance periods.
Table 1
Means, Standard Deviations, Internal Consistencies, and Correlations Between Variables
Dependent
variable MSD123456789 10111213141516171819
Performance
1. Typical 1 0.65 0.33 .75
2. Maximum 0.86 0.51 .65 .69
3. Typical 2 0.70 0.37 .73 .49 .67
Motivation
Direction
4. Typical 1 0.71 0.24 .59 .36 .48 .74
5. Maximum 0.82 0.24 .42 .41 .33 .66 .72
6. Typical 2 0.59 0.21 .36 .27 .55 .49 .41 .70
Level
7. Typical 1 9.54 4.15 .41 .18 .36 .46 .25 .15 .74
8. Maximum 10.63 4.63 .36 .21 .26 .41 .33 .10 .57 .78
9. Typical 2 7.87 4.03 .33 .29 .44 .19 .11 .55 .24 .20 .67
Persistence
10. Typical 1 0.08 0.32 .05 .15 .05 .07 .03 .10 .05 .04 .04
11. Maximum 0.01 0.50 .02 .06 .01 .12 .12 .03 .12 .07 .05 .03
12. Typical 2 0.33 0.43 .09 .09 .11 .00 .06 .07 .05 .03 .09 .11 .16
Self-efficacy
13. Computer 3.73 0.76 .45 .31 .44 .32 .22 .33 .30 .30 .34 .07 .01 .11 .95
14. Internet 3.85 0.58 .37 .18 .33 .31 .20 .27 .31 .30 .27 .01 .02 .08 .81 .92
15. Task valence 2.54 0.82 .36 .23 .39 .28 .24 .42 .14 .20 .24 .13 .12 .04 .45 .41 .83
Ability
Working smart
16. Typical 1 1.27 0.51 .45 .31 .29 .47 .31 .28 .25 .12 .25 .22 .10 .02 .30 .29 .26 .70
17. Maximum 1.22 0.60 .45 .60 .32 .27 .29 .24 .07 .02 .33 .33 .12 .00 .19 .08 .08 .53 .69
18. Typical 2 1.37 0.35 .38 .37 .40 .21 .14 .17 .09 .07 .34 .13 .04 .03 .13 .12 .14 .39 .55 .75
19. Knowledge 0.51 0.20 .33 .50 .25 .21 .27 .14 .19 .20 .07 .26 .01 .05 .22 .34 .18 .21 .33 .28 .69
Note. N 138. Internal consistencies are in italics on the diagonal. Correlations above .17 are significant at p .05, correlations above .23 are significant
at p .01, correlations above .27 are significant at p .001.
984
KLEHE AND ANDERSON
Hypothesis 2: Typical and Maximum Performance
Hypothesis 2 proposed that participants’ performance would be
higher under maximum than under typical performance conditions.
On average, participants entered 0.65 and 0.70 correct prices per
minute during the two typical performance periods and 0.87 cor-
rect prices per minute during the maximum performance period. A
repeated-measurement ANOVA revealed a significant difference,
F(2, 136) 22.09,
2
.25, p .001. Bonferroni-corrected
pairwise comparisons revealed that this effect was not due to
differences between the two assessments of typical performance
(d 2.18, p .10) but that participants’ performance during the
maximum performance period significantly surpassed their perfor-
mance during both the first (d 6.45, p .001) and the second
typical performance period (d 4.13, p .001), thus confirming
Hypothesis 2.
Hypothesis 3: Changes in Motivation Cause Varying
Performance Under Typical Versus Maximum
Performance Conditions
Hypothesis 3, which proposed that varying performance under
typical versus maximum performance conditions was primarily
due to changes in performers’ motivation, was tested via stepwise
regressions on typical performance. We started with participants’
level of maximum performance as a predictor in Step 1; added
participants’ direction, level, and persistence of effort as well as
their procedural skills during the maximum performance condition
in Step 2; and added the same variables during the respective
typical performance period in Step 3. As can be seen in Table 2,
Hypothesis 3 was fully supported for the prediction of perfor-
mance during the first typical performance assessment
(R
2
adjusted
.64, p .01). Only direction (␤⫽.27, p .01), level
(␤⫽.15, p .01), and persistence of effort (␤⫽.14, p .05),
yet not the degree to which people used smart procedures (␤⫽.10,
p .05), added incremental validity to the explanations of per-
formance (R
2
.12, p .01). For the second assessment of
typical performance, however, Hypothesis 3 was not supported
(R
2
adjusted
.53, p .01): In addition to direction of effort (␤⫽
.39, p .01), working smart turned out to be a significant predic-
tor as well (␤⫽.22, p .01), indicating that the distinction
between performance under typical and maximum performance
conditions may not always rely exclusively on a change in per-
formers’ motivation (R
2
.22, p .01).
Hypothesis 4: Correlations With Motivational Variables
To test whether motivation correlated higher with performance
under typical than under maximum performance conditions, we
ran two kinds of analyses: When both the predictors and the
criteria were assessed at both moments involved in the comparison
(direction, level, persistence, and working smart), we compared
correlations via Steiger’s (1980) z. Steiger’s procedure, however,
does not allow for comparisons in which a once-assessed predictor
(self-efficacy, task valence, knowledge) is correlated with two
assessments of the criterion. Consequently, we compared such
correlations with Williams’s (1959) t (t
w
). Analyzing data with
alternative procedures (Hotelling, 1940; Meng, Rosenthal, & Ru-
bin, 1992; Olkin, 1967) led to the same conclusions.
Direction. During the two typical performance periods, direc-
tion of effort correlated .59 and .55 (both ps .01) with perfor-
mance, with no significant difference between correlations (z
0.68; n 138). During maximum performance, the correlation
dropped to .41 ( p .01), presenting a significant decrease in
relationship (z ⫽⫺2.53; z ⫽⫺1.73; n 138, p .05).
Level. During the two typical performance periods, level of
effort correlated .41 and .44 (both ps .01) with performance, and
the correlations were not significantly different (z 0.25; n
138). During the maximum performance period, this correlation
dropped to .21 ( p .01), presenting a significant decrease in
relationship (z ⫽⫺2.24; z ⫽⫺2.17; n 138, p .05).
Persistence. During the two typical performance periods, per-
sistence correlated .05 and .11 with performance. During the
maximum performance period, this correlation was .06. No sig-
nificant differences emerged between any of the correlations (zs
0.43 to 0.90; n 138).
Self-efficacy. Internet self-efficacy correlated .37 and .33 (both
ps .001) with typical performance, with no significant difference
emerging between correlations, t
w
(135) 0.68. During maximum
performance, the correlation dropped to .18 ( p .05), presenting
a significant decrease in relationship, t
w
(135) ⫽⫺2.81; t
w
(135)
1.81, p .05.
Computer self-efficacy correlated .45 and .44 (both ps .001)
with typical performance, with no significant difference between
0
2
4
6
8
10
12
measurement
point 1
measurement
point 2
measurement
point 3
measurement
point 4
measurement
point 5
Time
Level of Effort -
task-related clicks per minute
typical 1
maximum
typical 2
Figure 1. Persistence during the two typical performance periods and the maximum performance period.
985
WORKING HARD AND WORKING SMART
correlations, t
w
(135) 0.02. During the maximum performance
period, this correlation dropped to .31 ( p .001), presenting a
significant decrease in relationship, t
w
(135) ⫽⫺2.04; t
w
(135)
1.69, p .05.
It is interesting that self-efficacy showed a similar pattern of
relationships with direction of effort under typical (rs .31 and
.27, respectively, both ps .01) and maximum (r .20, p .05)
performance conditions, even though this difference in correlations
did not reach significance, t
w
(135) ⫽⫺1.63; t
w
(135) ⫽⫺0.78.
Also, computer self-efficacy showed a similar pattern of relation-
ships with direction of effort under typical (rs .32 and .33,
respectively, both ps .01) and maximum (r .22, p .05)
performance conditions, even though, again, this difference in
correlations did not reach significance, t
w
(135) ⫽⫺1.47;
t
w
(135) ⫽⫺1.24. Internet and computer self-efficacy further cor
-
related significantly with level of effort—although not, however,
with persistence of effort— during the three observation periods,
with no clear differences between typical and maximum perfor-
mance periods.
Task valence. Task valence correlated .36 and .39 (both ps
.01) with performance in the typical performance conditions, with
no significant difference between correlations, t
w
(135) 0.49.
During the maximum performance period, the correlation dropped
to .23 ( p .05), a significant decrease, t
w
(135) ⫽⫺2.03;
t
w
(135) ⫽⫺2.06, p .05. A graphic representation of this
finding, illustrative of the combined effects of intrinsic (task va-
lence) and extrinsic (maximum performance conditions) motiva-
tors on performance, is presented in Figure 2. In sum, motivational
variables correlated higher with performance during the typical
than during the maximum performance periods, largely supporting
Hypothesis 4.
Hypothesis 5: Correlations With Knowledge and Skills
Hypothesis 5 proposed that ability should correlate higher with
participants’ performance under maximum than under typical per-
formance conditions.
Knowledge. During the typical performance periods, results
on the knowledge test correlated .33 and .25 (both ps .01) with
performance, with no significant difference between correlations,
0.6
0.4
Performance (scores standardized)
0.2
0
low (-1SD) high (+1SD)
-0.2
-0.4
typical 1
-0.6
maximum
typical 2
-0.8
Task valence (scores standardized)
Figure 2. The interaction between typical versus maximum performance
condition and task valence.
Table 2
Stepwise Regression of Typical Performance on Maximum Performance and Participants’
Changes in Working Hard (Direction, Level, and Effort) and Smart (Hypothesis 3)
Step and variable
Performance
Typical 1 Typical 2
Adjusted
R
2
R
2
Adjusted
R
2
R
2
Step 1
Max. performance .70
**
.49 .49
**
.56
**
.31 .32
**
Step 2
Max. performance .60
**
.53 .06
**
.50
**
.32 .04
Max. direction .01 .03
Max. level .25
**
.19
*
Max. persistence .01 .01
Max. working smart .10 .01
Step 3
Max. performance .54
**
.64 .12
**
.34
**
.53 .22
**
Max. direction .13 .03
Max. level .11 .11
Max. persistence .03 .01
Max. working smart .08 .16
Typ. direction .27
**
.39
**
Typ. level .15
**
.11
Typ. persistence .14
*
.12
Typ. working smart .10 .22
**
Note. N 138. The data had no multicollinearity and no outliers. Max. maximum; Typ. typical.
*
p .05.
**
p .01.
986
KLEHE AND ANDERSON
t
w
(135) 1.31. During the maximum performance period, this
correlation was .50 ( p .01), representing a significant increase in
relationship, t
w
(135) 2.59; t
w
(135) 3.12, p .05. Further
-
more, individuals who scored better on the knowledge test also
showed more direction and level of effort during the first typical
performance period (rs .21 and .19, respectively, both ps .05)
and the maximum performance period (rs .27 and .20, respec-
tively, both p .05), although not, however, during the second
performance period (rs .14 and .07, respectively). Here, differ-
ences between correlations were not significant, t
w
(135) ⫽⫺0.13
to t
w
(135) 1.44.
Working smart. During the typical performance periods,
working smart correlated .45 and .40 (both ps .01) with perfor-
mance, with no significant difference between correlations (z
0.61; n 138). During the maximum performance period, this
correlation was .60 ( p .01), representing a significant increase in
relationship (z 2.03; z 2.52; n 138). Thus, Hypothesis 5 was
supported for both task-related declarative and procedural knowl-
edge.
Hypothesis 6: Predicting Typical and Maximum
Performance
Hypothesis 6 proposed that the best predictors for performance
under maximum performance conditions would be measures of
task-related knowledge and procedural skills, whereas the best
predictors for performance under typical performance conditions
would include measures of motivation as well. This hypothesis
was analyzed with stepwise multiple regressions predicting per-
formance with the help of all predictors included in the study. As
can be seen in Table 3, the best predictor of typical performance
during both performance periods was direction (␤⫽.44 and ␤⫽
.42, respectively, both ps .01). The second most relevant pre-
dictor was the procedural skills used during the task (␤⫽.21 and
␤⫽.25, respectively, both ps .01). In addition, the motivational
variable computer self-efficacy predicted performance both times
(␤⫽.20, p .05, and ␤⫽.24, p .01; total R
2
adjusted
.44, p
.01). In the second assessment of typical performance, persistence
(␤⫽.18, p .01) and task valence (␤⫽.15, p .05) emerged
as two additional significant predictors (total R
2
adjusted
.53,
p .01).
Under the maximum performance condition (Table 4), direction
of effort remained an important predictor for performance (␤⫽
.29, p .01), yet it fell back in relative importance compared with
procedural skills (␤⫽.40, p .01). In addition, declarative
knowledge regarding the content and means of the task incremen-
tally predicted performance (␤⫽.29, p .01; total R
2
adjusted
.53, p .01). Thus, the findings disconfirm Hypothesis 6, that
maximum performance would only be accounted for by declara-
tive knowledge and procedural skills. However, compared with the
assessments of typical performance, the role of ability increased,
whereas the role of motivation decreased, in accounting for per-
formance.
Discussion
The contribution of this study is fourfold. In the next four
sections, we elaborate on the four areas in which the study con-
tributed.
Validity of the Sackett et al. (1988) Model
The typical versus maximum performance distinction has re-
ceived considerable theoretical (e.g., Ackerman & Humphreys,
1990; Arvey & Murphy, 1998; Boudreau, 1991; Dewberry, 2001;
Guion, 1991, 1998) and emerging empirical attention (DuBois et
al., 1993; Klehe & Latham, 2006; Lim & Ployhart, 2004; Ployhart
et al., 2001; Smith-Jentsch et al., 2001), and researchers have
nearly uniformly accepted Sackett et al.’s (1988) assumptions on
faith alone (for exceptions, see Barnes & Morgeson, 2007, and
Mangos, Steele-Johnson, LaHuis, & White, 2007). This study
presents an initial detailed test of the mechanisms proposed by
Sackett et al. (1988) and DuBois et al. (1993) to underlie the
distinction between typical and maximum performance: It evalu-
ated the development of both motivation (direction, level, persis-
tence) and procedural skills (working smart) under both typical
and maximum performance conditions. Confirming Sackett et al.’s
(1988) assumptions, participants worked harder under the maxi-
mum performance condition (Hypothesis 1), which led their max-
imum performance to surpass their typical performance (Hypoth-
esis 2). Participants’ increased motivation during the maximum
performance condition, compared with the two typical perfor-
mance conditions, explained some of this difference (Hypothesis
3). Furthermore, measures of motivation (e.g., direction and level
of effort), but also participants’ perceived self-efficacy regarding
the contents and the means of the task as well as their intrinsic
enjoyment of the task in the form of task valence, correlated higher
Table 3
Stepwise Regression of the First and Second Assessments of
Typical Performance on the Measures of Ability and Motivation
Included in the Study
Model and
variable
Performance
Typical 1 Typical 2
Adjusted
R
2
R
2
Adjusted
R
2
R
2
Model 1
Direction .61
**
.37 .37
**
.60
**
.36 .36
**
Model 2
Direction .50
**
.56
**
Working smart .25
**
.41 .05
**
.28
**
.43 .07
**
Model 3
Direction .44
**
.47
**
Working smart .21
**
.26
**
Computer SE .20
*
.44 .03
*
.26
**
.48 .06
**
Model 4
Direction .47
**
Working smart .26
**
Computer SE .28
**
Persistence .19
**
.51 .03
**
Model 5
Direction .42
**
Working smart .25
**
Computer SE .24
**
Persistence .18
**
Task valence .15
*
.53 .02
*
Note. N 138. The data had no multicollinearity and no outliers. SE
self-efficacy.
*
p .05.
**
p .01.
987
WORKING HARD AND WORKING SMART
with performance under typical than under maximum performance
conditions (Hypothesis 4), whereas both declarative knowledge
and procedural skills correlated higher with performance under the
maximum than under the typical performance conditions (Hypoth-
esis 5). Finally, measures of motivation, such as direction of effort,
computer self-efficacy, and persistence, played an important role
in predicting typical performance, whereas measures of ability,
namely procedural skills and knowledge regarding the means and
the content of the task, played an increased role under maximum
performance conditions (Hypothesis 6). All this lends support for
Sackett et al.’s (1988) original concept of typical versus maximum
performance.
These findings have important practical implications. First, they
confirm that the typical–maximum performance distinction is real
and should be included and studied in models of job performance
(e.g., Campbell, 1990; Viswesvaran & Ones, 2000). Second, they
have implications for when one is predicting performance through
measures of ability or motivation, as in, for example, employee
selection procedures (see also Smith-Jentsch, 2007). The useful-
ness of most predictors included varied greatly across typical
versus maximum performance conditions (Campbell, 1990;
DuBois et al., 1993; Guion, 1998). As Guion (1991) noted, re-
searchers run extensive validation studies, and organizations make
huge financial investments in the selection of new employees,
without knowing which of these two aspects of performance they
are predicting or even trying to predict (Klehe & Anderson, 2005).
Campbell (1990) argued that basing selection decisions on predic-
tors of maximum performance could be one cause for the weak
relationship often found between results of personnel selection
procedures and typical performance on the job. Such a mismatch
could also bear considerable financial consequences for the orga-
nization in question: Boudreau (1991) noted that results from
utility analyses regarding a selection procedure’s prediction of
typical job performance are likely to be biased if the dollar value
of performance is based on maximum performance criteria, and
vice versa.
Distinction Between Typical and Maximum Performance
The findings of this study demonstrate that the distinction be-
tween typical and maximum performance is slightly more complex
than originally proposed by Sackett et al. (1988). This is for two
reasons. First, this study shows that although Sackett et al.’s (1988)
general propositions hold true, motivation may still explain a
considerable proportion of variance in performance under maxi-
mum performance conditions. Although correlations of direction
and level of effort, self-efficacy, and task valence with perfor-
mance shrank during the maximum performance period, they still
remained significant, supporting findings reported by Kirk and
Brown (2003).
Campbell (1990) argued that performance requires a minimum
of both procedural skills and motivation to occur. The current
results confirm this argument. Although the relative weight of
measures of motivation and ability when we predicted typical and
maximum performance pointed in the direction proposed by Sack-
ett et al. (1988), direction of effort and procedural skills remained
the two most important predictors of performance, no matter
whether performance was assessed under typical or under maxi-
mum performance conditions. One may argue, of course, that
typical versus maximum performance conditions reflect a contin-
uum and that our manipulation of a maximum performance situ-
ation was not the strongest one conceivable. Yet the concurrence
of our findings with those of Kirk and Brown (2003) suggests that
maximum performance as a pure measure of ability may be diffi-
cult to establish and that Sackett et al.’s (1988) hope to gain,
through a comparison of typical and maximum performance, an
estimate of the role of ability during typical performance may be
overly optimistic.
In addition, our results suggest that participants may not always
work at their smartest under maximum performance conditions,
which, in turn, could explain the significant difference in working
smart between the maximum and the second typical performance
period as well as the fact that working smart accounted for incre-
mental validity in the prediction of the second typical performance
assessment. Maybe the slightly detrimental effect found for the
maximum performance condition on working smart was due less
to the maximum performance condition per se than to participants
being distracted from the task by the presence of the experimenter
(Sanders, 1981; see also Zajonc, 1965). Equally likely, however, is
that the evaluative nature of the maximum performance condition
raised evaluation anxieties among at least some participants, im-
pairing their ability to work smart, an effect similar to findings
reported in educational psychology (Wine, 1971) and social psy-
chology (Sanna, 1992) and now partially supported for maximum
performance situations (Klehe, Anderson, & Hoefnagels, 2007).
We thus call for future research to test whether and how typical
versus maximum performance conditions influence not only mo-
tivation but also the procedural skills needed for accomplishing the
task.
Criterion Measurement Problems and Issues
This study raises some doubt as to the validity of supervisory
evaluations for estimating typical job performance. Although the
manipulation used in the current study for inducing maximum
performance was purposely not extreme, it was not much different
from actions undertaken by supervisors to learn about their em-
ployees’ performance. Thorsteinson and Balzer (1999) suggested
that, unlike coworkers, who may be capable of gathering daily
information about individuals’ typical motivation and perfor-
Table 4
Stepwise Regression of the Assessment of Maximum
Performance on the Measures of Ability and Motivation
Included in the Study
Model and
variable
Maximum performance
Adjusted R
2
R
2
Model 1
Working smart .60
**
.35 .36
**
Model 2
Working smart .48
**
Direction .36
**
.46 .11
**
Model 3
Working smart .40
**
Direction .29
**
Knowledge .29
**
.53 .07
**
Note. N 138. The data had no multicollinearity and no outliers.
**
p .01.
988
KLEHE AND ANDERSON
mance, supervisors may only be allowed to observe and rate
performers’ maximum performance. During this laboratory exper-
iment, the mere presence of the experimenter motivated partici-
pants to work significantly harder than they did during the rest of
the experiment. It is quite likely that this effect is considerably
stronger in real-life settings, when people work on tasks they have
performed not just for 2 hr but for years, so that even the last
novelty effect has worn off, and when employees know that the
impressions formed about them by a supervisor, unlike those
formed by an experimenter in a psychology department, can have
actual consequences for their careers. Hence, it is not surprising
that Sackett et al. (1988) found supervisory ratings of performance
to be significantly more related to the maximum than to the typical
speed and accuracy with which supermarket cashiers processed
grocery items. Yet further research on the assessment of typical
versus maximum performance by different rating sources is clearly
needed.
Research Settings: Laboratory and Field Studies
This experiment shows that the distinction between typical and
maximum performance can be studied in the laboratory. Being
able to study the distinction between typical and maximum per-
formance in the laboratory is important in the face of the scarcity
of empirical research currently available on this important distinc-
tion. A number of research questions may well be open for
laboratory research, and the introduction of a laboratory design to
the study of typical and maximum performance may facilitate
future research that otherwise might not be conducted in the
absence of truly parallel situations of typical and maximum per-
formance in field settings. Examples of such research include the
effect of typical and maximum performance situations on the use
of impression management tactics, the combination of typical
versus maximum performance with other theories on motivation
(e.g., expectancy theory; Van Eerde & Thierry, 1996; Vroom,
1964), and the search for moderators to reactions to typical versus
maximum performance situations (e.g., Klehe & Anderson, 2007).
However, it is important that findings from such experimental
studies be extended to and—as far as logistic considerations al-
low—replicated in naturalistic field study settings. Future research
can usefully combine both approaches to address questions of
causality and generalizability.
Study Limitations and Strengths
Of course, the use of a laboratory setting is not only new for the
study of typical versus maximum performance but is also associ-
ated with a number of limitations, primarily in regard to the study’s
task, sample, and duration. Concerning the duration of the task,
one might conceivably argue that 90 min is too short a period to
reflect typical performance. Yet, on a continuum between “pure”
typical and maximum performance situations, the typical perfor-
mance periods were less evaluative, less instructive to focus on the
task, and longer, thus satisfying all three of Sackett et al.’s (1988)
requirements. The fact that 8 participants quit the experiment
prematurely (forgoing their reward) and the finding that persis-
tence added incremental validity to the explanation of the second
assessment of typical performance also suggest that 1.5 hr was too
long for participants to maintain their maximum motivation and,
thus, their maximum performance. Plainly, it was long enough for
at least some participants to lose sufficient interest in the task to
quit.
As noted above, however, in some cases the results in the first
and the second typical performance period did not converge per-
fectly (e.g., direction, level, and persistence of effort were, on
average, higher during the first than during the second period). In
these cases, it appears more prudent to rely primarily on the second
assessment as an indicator of typical performance, as any experi-
mental demand, learning, or honeymoon effects that were possibly
still active in the first half of the experiment would have worn off.
It is interesting to note that the comparison between the maximum
and the second typical performance assessment was also the com-
parison that showed stronger differential effects in response to the
hypotheses.
At the same time, one may wonder whether 5 min is long
enough for individuals to present their maximum performance.
Although the time was surely short enough for participants to
remain focused on the task, such a short duration may turn prob-
lematic in tasks that are more complex and time consuming per
trial (e.g., solving complex puzzles) than in the current case.
Certainly, future studies are called for that use longer periods of
measurement of maximum performance. On a related note, this
study, just like nearly all other studies on typical versus maximum
performance published to date (e.g., Kirk & Brown, 2003; Lim &
Ployhart, 2004; Ployhart et al., 2001; Sackett et al., 1988; Smith-
Jentsch et al., 2001), justified the manipulation of typical versus
maximum performance conditions on theoretical grounds but did
not include a manipulation check, a decision primarily due to
logistic considerations (a manipulation check after the maximum
performance period would have made data collection during the
following typical performance period impossible), even though a
manipulation check would obviously have been desirable.
The current experiment used a repetitive but rather strenuous
task, judging from the relatively low task valence indicated by
participants as well as from comments participants made in their
final remarks or when leaving the experiment prematurely. Most
participants were undergraduate psychology students, a sample
unlikely to choose a task like this out of intrinsic motivation.
Although this is likely to limit the generalizability of the study’s
results, it poses no automatic threat to the study’s ecological
validity. First, students are frequently confronted with similar
high-attention yet relatively mindless tasks that call for high dili-
gence during their part-time jobs or, in altered form, even during
their studies (e.g., entering data, scanning literature). Second,
participants still showed meaningful individual differences not
only in task valence but also in exerted effort throughout the
performance periods. The causes of these differences (e.g., inter-
personal differences in interests or personalities; Kanfer & Heg-
gestad, 1999) are beyond the scope of this article. However, they
are likely a reflection of meaningful differences in participants’
intrinsic motivation during the task. The manipulation of maxi-
mum performance, in contrast, represents a classic approach for
ensuring extrinsic motivation. Although this manipulation
achieved an average increase in participants’ direction, level, and
persistence of effort, the influence of interindividual differences in
motivation, particularly intrinsic motivation in the form of partic-
ipants’ task valence, on performance decreased during this mea-
surement period.
989
WORKING HARD AND WORKING SMART
Directions for Future Research
Given the above limitations, results from this study are likely to
generalize primarily to job roles and tasks that are detail oriented,
repetitive, and lower in cognitive demand and for which perform-
ers hold relatively little intrinsic interest. Different findings might
have emerged in a sample more intrinsically interested in the task,
such as hardware engineers or retailers. For example, the differ-
ence in motivation and performance during typical and maximum
performance conditions might have been smaller, maybe even
reversed. In addition, research on self-determination theory (Deci,
Koestner, & Ryan, 1999; Ryan & Deci, 2000) suggests not only
that— given sufficient intrinsic motivation—performers may work
as hard under typical performance conditions as they do under
maximum performance conditions but that their intrinsic motiva-
tion, and hence their subsequent typical performance, may suffer
from the introduction of maximum performance periods. Self-
determination theory suggests that evaluations and rewards foster
motivation to the degree that they provide information on the
performer’s level of performance but that they reduce motivation
to the degree that they are perceived as controlling and hence
hindering performers’ striving for autonomy. The observation of
typical and maximum performance periods over time and the
motivating and/or demotivating effects of either on the other at
different levels of initial intrinsic motivation are likely to be an
important avenue for future research (see also Klehe & Anderson,
2007).
Furthermore, research on social facilitation and inhibition, the
increase or decrease of individuals’ performance in the presence
or, more precisely, under the evaluation of others (Cottrell, 1968,
1972; Henchy & Glass, 1968), suggests that Sackett et al.’s (1988)
assumptions may only hold true as long as individuals actually
have the ability to perform the task and know that they have it
(Sanna, 1992). Klehe, Anderson, and Hoefnagels (2007) linked the
typical versus maximum performance literature to that of social
facilitation and inhibition and found that maximum performance
conditions can actually hinder performance on a task that perform-
ers perceive as difficult or that they have not yet sufficiently
learned at the time of assessment.
If this link with the mechanisms of social facilitation and inhi-
bition, and thus also with social loafing (Sanna, 1992), also holds
true in future studies, it could enrich both sets of literatures—for
example, by introducing known moderators from the social loafing
literature (Karau & Williams, 1993) as possible moderators to the
different performance levels under typical versus maximum per-
formance conditions or, conversely, by testing the impact of ability
versus motivation on performance within a classic social facilita-
tion design.
Conclusion
Given the aims of this study, we purposefully chose a laboratory
rather than a field setting for a number of reasons, foremost among
them the ability to observe the development of motivation and
performance across typical and maximum performance. Conse-
quently, this study has succeeded in supporting most of Sackett et
al.’s (1988) main assumptions regarding the development of mo-
tivation during typical and maximum performance periods and its
consecutive impact on performance. Yet this study also invites
future research, given (a) that results on the prediction of perfor-
mance with measures of ability and motivation were slightly more
complex than expected, (b) that the results may not generalize to
tasks and samples of high task valence and intrinsic motivation,
and (c) that this study proves the usefulness of the laboratory for
the study of typical versus maximum performance.
References
Ackerman, P. L., & Humphreys, L. G. (1990). Individual differences in
industrial and organizational psychology. In M. D. Dunnette & L. M.
Hough (Eds.), Handbook of industrial and organizational psychology
(Vol. 1, 2nd ed., pp. 223–282). Palo Alto, CA: Consulting Psychologists
Press.
Ackerman, P. L., Kanfer, R., & Goff, M. (1995). Cognitive and noncog-
nitive determinants and consequences of complex skill acquisition. Jour-
nal of Experimental Psychology: Applied, 1, 270–304.
Arvey, R. D., & Murphy, K. R. (1998). Performance evaluation in work
settings. Annual Review of Psychology, 49, 141–168.
Bandura, A. (1997). Self-efficacy: The exercise of control. New York:
Freeman.
Barnes, C. M., & Morgeson, F. P. (2007). Typical performance, maximal
performance, and performance variability: Expanding our understanding
of how organizations value performance. Human Performance, 20,
259 –274.
Borman, W. C. (1991). Job behavior, performance, and effectiveness. In
M. D. Dunnette & L. M. Hough (Eds.), Handbook of industrial and
organizational psychology (Vol. 2, 2nd ed., pp. 271–326). Palo Alto,
CA: Consulting Psychologists Press.
Boudreau, J. W. (1991). Utility analysis for decisions in human resource
management. In M. D. Dunnette & L. M. Hough (Eds.), Handbook of
industrial and organizational psychology (Vol. 2, 2nd ed., pp. 621–745).
Palo Alto, CA: Consulting Psychologists Press.
Campbell, J. P. (1990). Modeling the performance prediction problem in
industrial and organizational psychology. In M. D. Dunnette & L. M.
Hough (Eds.), Handbook of industrial and organizational psychology
(Vol. 1, 2nd ed., pp. 687–732). Palo Alto, CA: Consulting Psychologists
Press.
Chatman, J. A. (1991). Matching people and organizations: Selection and
socialization in public accounting firms. Administrative Science Quar-
terly, 36, 459 484.
Cottrell, N. B. (1968). Performance in the presence of other human beings:
Mere presence audience and affiliation effects. In E. C. Simmel, R. A.
Hoppe, & G. A. Milton (Eds.), Social facilitation and imitative behavior
(pp. 245–250). Boston: Allyn & Bacon.
Cottrell, N. B. (1972). Social facilitation. In C. G. McClintock (Ed.),
Experimental social psychology (pp. 185–236). New York: Holt, Rine-
hart & Winston.
Cronbach, L. J. (1960). Essentials of psychological testing (2nd ed.).
Oxford, England: Harper.
Deci, E. L., Koestner, R., & Ryan, R. M. (1999). A meta-analytic review
of experiments examining the effects of extrinsic rewards on intrinsic
motivation. Psychological Bulletin, 125, 627– 668.
Dewberry, C. (2001). Performance disparities between Whites and ethnic
minorities: Real differences or assessment bias? Journal of Occupa-
tional and Organizational Psychology, 74, 659 673.
DuBois, C. L. Z., Sackett, P. R., Zedeck, S., & Fogli, L. (1993). Further
exploration of typical and maximum performance criteria: Definitional
issues, prediction, and White–Black differences. Journal of Applied
Psychology, 78, 205–211.
Eastin, M. S., & LaRose, R. (2000). Internet self-efficacy and the psychol-
ogy of the digital divide. Journal of Computer-Mediated Communica-
tion, 6. Retrieved May 11, 2007, from http://jcmc.indiana.edu/vol6/
issue1/eastin.html
990
KLEHE AND ANDERSON
Fichman, M., & Levinthal, D. A. (1991). Honeymoons and the liability of
adolescence: A new perspective on duration dependence in social and
organizational relationships. Academy of Management Review, 16, 442–
468.
ForsterLee, R. (2007). Personality, interest, and motivational determinants
of maximal and typical performances on a verbal knowledge task.
Human Performance, 20, 287–292.
Freitas, A. L., Liberman, N., Salovey, P., & Higgins, E. T. (2002). When
to begin? Regulatory focus and initiating goal pursuit. Personality and
Social Psychology Bulletin, 28, 121–130.
Guion, R. M. (1991). Personnel assessment, selection, and placement. In
M. D. Dunnette & L. M. Hough (Eds.), Handbook of industrial and
organizational psychology (Vol. 2, 2nd ed., pp. 327–397). Palo Alto,
CA: Consulting Psychologists Press.
Guion, R. M. (1998). Assessment, measurement, and prediction for per-
sonnel decisions. Mahwah, NJ: Erlbaum.
Henchy, T., & Glass, D. C. (1968). Evaluation apprehension and the social
facilitation of dominant and subordinate responses. Journal of Person-
ality and Social Psychology, 10, 446 454.
Hotelling, H. (1940). The selection of variates for use in prediction, with
some comments on the general problem of nuisance parameters. Annals
of Mathematical Statistics, 11, 271–283.
Humphrey, S. E., Hollenbeck, J. R., Ilgen, D. R., & Moon, H. (2004). The
changing shape of large scale programs of research: MSU-DDD as an
illustrative example. In S. G. Schiflett, L. R. Elliott, E. Salas, & M. D.
Coovert (Eds.), Scaled worlds: Development, validation and applica-
tions (pp. 200 –219). Aldershot, England: Ashgate.
Joo, Y. J., Bong, M., & Choi, H. J. (2000). Self-efficacy for self-regulated
learning, academic self-efficacy and Internet self-efficacy in Web-based
instruction. Educational Technology Research and Development, 48,
5–17.
Kanfer, R., & Heggestad, E. D. (1999). Individual differences in motiva-
tion: Traits and self-regulatory skills. In P. L. Ackerman, P. C. Kyllonen,
& R. D. Roberts (Eds.), Learning and individual differences: Process,
trait, and content determinants (pp. 293–313). Washington, DC: Amer-
ican Psychological Association.
Karau, S. J., & Williams, K. D. (1993). Social loafing: A meta-analytic
review and theoretical integration. Journal of Personality and Social
Psychology, 65, 681–706.
Kirk, A. K., & Brown, D. F. (2003). Latent constructs of proximal and
distal motivation predicting performance under maximum test condi-
tions. Journal of Applied Psychology, 88, 40 49.
Klehe, U.-C., & Anderson, N. (2005). The prediction of typical and
maximum performance. In A. Evers, N. Anderson, & O. Smit-Voskuijl
(Eds.), Handbook of personnel selection (pp. 331–353). Oxford, En-
gland: Blackwell.
Klehe, U.-C., & Anderson, N. (2007). The moderating influence of per-
sonality and social loafing on motivation in typical versus maximal
performance situations. International Journal of Selection and Assess-
ment, 15, 250 –262.
Klehe, U.-C., Anderson, N. R., & Hoefnagels, E. A. (2007). Social facil-
itation and inhibition during maximum versus typical performance sit-
uations. Human Performance, 20, 223–239.
Klehe, U.-C., Anderson, N. R. & Viswesvaran, C. (2007). More than peaks
and valleys: Introduction to the special issue on typical and maximum
performance. Human Performance, 20, 173–177.
Klehe, U.-C., & Latham, G. P. (2006). What would you do—really or
ideally? Constructs underlying the Behavior Description Interview and
the Situational Interview in predicting typical versus maximum perfor-
mance. Human Performance, 19, 357–382.
Lim, B. C., & Ployhart, R. E. (2004). Transformational leadership: Rela-
tions to the five-factor model and team performance in typical and
maximum contexts. Journal of Applied Psychology, 89, 610 621.
Locke, E. A., Mento, A. J., & Katcher, B. L. (1978). The interaction of
ability and motivation in performance: An exploration of the meaning of
moderators. Personnel Psychology, 31, 269 –280.
Louis, M. R. (1980). Surprise and sense making: What newcomers expe-
rience in entering unfamiliar organizational settings. Administrative Sci-
ence Quarterly, 25, 226–251.
LSVb [Landelijke Studenten Vakbond]. (2004). Studenten en studenten-
leven [Students and student life]. Utrecht, The Netherlands: Onderzoeks-
bureau van de Landelijke Studenten Vakbond.
Maier, N. R. F. (1955). Psychology in industry (2nd ed.). Oxford, England:
Houghton Mifflin.
Mangos, P. M., Steele-Johnson, D., LaHuis, D., & White, E. D., III. (2007).
A multiple-task measurement framework for assessing maximum-
typical performance. Human Performance, 20, 241–258.
Marcus, B., Goffin, R. D., Johnston, N. G., & Rothstein, M. G. (2007).
Personality and cognitive ability as predictors of typical and maximum
managerial performance. Human performance, 20, 275–285.
McCloy, R. A., Campbell, J. P., & Cudeck, R. (1994). A confirmatory test
of a model of performance determinants. Journal of Applied Psychology,
79, 493–505.
Meng, X.-L., Rosenthal, R., & Rubin, D. B. (1992). Comparing correlated
correlation coefficients. Psychological Bulletin, 111, 172–175.
Murphy, C. A., Coover, D., & Owen, S. V. (1989). Development and
validation of the Computer Self-Efficacy Scale. Educational and Psy-
chological Measurement, 49, 893–899.
Olkin, I. (1967). Correlations revisited. In J. C. Stanley (Ed.), Improving
experimental design and statistical analysis. Chicago: Rand McNally.
Ones, D. S., & Viswesvaran, C. (2007). A research note on the incremental
validity of job knowledge and integrity tests for predicting maximal
performance. Human Performance, 20, 293–303.
Ployhart, R. E., Lim, B. C., & Chan, K. Y. (2001). Exploring relations
between typical and maximum performance ratings and the five factor
model of personality. Personnel Psychology, 54, 809 843.
Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the
facilitation of intrinsic motivation, social development, and well-being.
American Psychologist, 55, 68–78.
Sackett, P. R. (2007). Revisiting the origins of the typical-maximum
performance distinction. Human Performance, 20, 179 –185.
Sackett, P. R., & Larson, J. R., Jr. (1990). Research strategies and tactics
in industrial and organizational psychology. In M. D. Dunnette & L. M.
Hough (Eds.), Handbook of industrial and organizational psychology
(Vol. 1, 2nd ed., pp. 419 489). Palo Alto, CA: Consulting Psychologists
Press.
Sackett, P. R., Zedeck, S., & Fogli, L. (1988). Relations between measures
of typical and maximum job performance. Journal of Applied Psychol-
ogy, 73, 482– 486.
Sanders, G. S. (1981). Driven by distraction: An integrative review of
social facilitation theory and research. Journal of Experimental Social
Psychology, 17, 227–251.
Sanna, L. J. (1992). Self-efficacy theory: Implications for social facilitation
and social loafing. Journal of Personality and Social Psychology, 62,
774 –786.
Smith-Jentsch, K. A. (2007). The impact of making targeted dimensions
transparent on relations with typical performance predictors. Human
Performance, 20, 187–203.
Smith-Jentsch, K. A., Salas, E., & Brannick, M. T. (2001). To transfer or
not to transfer? Investigating the combined effects of trainee character-
istics, team leader support, and team climate. Journal of Applied Psy-
chology, 86, 279 –292.
Steiger, J. H. (1980). Tests for comparing elements of a correlation matrix.
Psychological Bulletin, 87, 245–251.
Stevens, J. P. (2002). Applied multivariate statistics for the social sciences
(4th ed.). Mahwah, NJ: Erlbaum.
Thorsteinson, T. J., & Balzer, W. K. (1999). Effects of coworker informa-
991
WORKING HARD AND WORKING SMART
tion on perceptions and ratings of performance. Journal of Organiza-
tional Behavior, 20, 1157–1173.
Van Eerde, W., & Thierry, H. (1996). Vroom’s expectancy models and
work-related criteria: A meta-analysis. Journal of Applied Psychology,
81, 575–586.
Viswesvaran, C., & Ones, D. S. (2000). Perspectives on models of job
performance. International Journal of Selection and Assessment, 8,
216 –226.
Vroom, V. H. (1964). Work and motivation. Oxford, England: Wiley.
Williams, E. J. (1959). The comparison of regression variables. Journal of
the Royal Statistical Society, 21, 396 –399.
Wine, J. (1971). Test anxiety and direction of attention. Psychological
Bulletin, 76, 92–104.
Witt, L. A., & Spitzmu¨ller, C. (2007). Person-situation predictors of
maximum and typical performance. Human Performance, 20, 305–
315.
Zajonc, R. B. (1965, July 16). Social facilitation. Science, 149, 269 –274.
Received February 17, 2005
Revision received September 16, 2006
Accepted October 23, 2006
Members of Underrepresented Groups:
Reviewers for Journal Manuscripts Wanted
If you are interested in reviewing manuscripts for APA journals, the APA Publications and
Communications Board would like to invite your participation. Manuscript reviewers are vital to the
publications process. As a reviewer, you would gain valuable experience in publishing. The P&C
Board is particularly interested in encouraging members of underrepresented groups to participate
more in this process.
If you are interested in reviewing manuscripts, please write to the address below. Please note the
following important points:
To be selected as a reviewer, you must have published articles in peer-reviewed journals. The
experience of publishing provides a reviewer with the basis for preparing a thorough, objective
review.
To be selected, it is critical to be a regular reader of the five to six empirical journals that are most
central to the area or journal for which you would like to review. Current knowledge of recently
published research provides a reviewer with the knowledge base to evaluate a new submission
within the context of existing research.
To select the appropriate reviewers for each manuscript, the editor needs detailed information.
Please include with your letter your vita. In the letter, please identify which APA journal(s) you
are interested in, and describe your area of expertise. Be as specific as possible. For example,
“social psychology” is not sufficient—you would need to specify “social cognition” or “attitude
change” as well.
Reviewing a manuscript takes time (1– 4 hours per manuscript reviewed). If you are selected to
review a manuscript, be prepared to invest the necessary time to evaluate the manuscript
thoroughly.
Write to Journals Office, American Psychological Association, 750 First Street, NE, Washington,
DC 20002-4242.
992
KLEHE AND ANDERSON
... Consideration is only to those conditions or hazards that are inherent like the work and not the extreme situations, that is, where the risk of a specific situation or accident occurring is unlikely. The frequency of exposure to undesirable working conditions or hazards should also be considered (Kaber et al., 2016) (Klehe & Anderson, 2007). ...
Article
Full-text available
The purpose of this research is to investigate the relationship that exists between hygiene factors and job satisfaction (JS) among the Royal Malaysian Navy (RMN) personnel in the workplace. The focus shall be directed toward the diligent personnel on board at Lumut Naval Base. The present study adopts a quantitative analysis methodology, employing a straightforward descriptive approach and correlation analysis through SPSS 23, wherein a sample of 220 employees was randomly selected for analysis. The empirical evidence suggests that there exist notable associations between job satisfaction (JS) and various factors, namely company policy, salary and benefits, and interpersonal relationships. The variables of job security and working conditions do not exhibit a statistically significant correlation with job satisfaction. The findings of this study suggest that there exists a noteworthy correlation between variables towards Job Satisfaction (JS). Therefore, there is no significant correlation between Job Security and Working Conditions in JS. The study's impact extends beyond the organization, encompassing the broader field, as it adds to the body of literature about the research elements. The utilization of models and theoretical frameworks in prospective investigations presents a distinct point for comprehending and interpreting phenomena that represents a noteworthy innovation within the context study.
... Factors from Job Stress and coworkers still often occur in influencing nurse performance. This condition can occur due to the differentiation of individual work with performance targets which are a function of declarative knowledge, procedural knowledge and skills, and work motivation (Klehe & Anderson, 2007). This factor represents an understanding of the requirements of the task being performed. ...
Article
Full-text available
The aim of this research is to analyze and prove the influence of Job Stress and Coworkers on the performance of nurses at Adent Hospital Medan. This research will also prove the moderating effect of compensation on the relationship between Job Stress and co-workers on performance. The population in this study were 150 permanent nurses who worked at the Medan Adventist Hospital. The sampling technique used was Simple Random Sampling, obtaining a sample of 109 respondents. Data analysis techniques use multiple regression with t-test (partial test), F test (simultaneous test) and coefficient of determination test (R2) as well as moderating analysis using residual analysis. The results of this study prove that Job Stress and co-worker variables simultaneously influence nurse performance. Partial testing proves that the Job Stress variable influences nurse performance, as well as the individual Co Workers variable influences nurse performance. Moderation testing using the residual test proves that compensation is able to moderate the relationship between Job Stress and coworkers on nurse performance. The implications of the results of this research are able to prove that the importance of Job Stress management and co-workers in improving nurse performance is strengthened by compensation.
... Moreover, how we define important life outcomes in terms of maximum performance versus typical performance (see Sackett et al. 1988;Klehe and Anderson 2007) can also influence the conclusions drawn from research on intelligence and personality. For instance, life outcomes that require maximum performance (e.g., "working smart" needs peak performance but only on occasion) may depend more heavily on intelligence, whereas outcomes that rely on typical performance (e.g., "working hard" needs average performance but maintained for a long time) may more heavily depend on personality factors such as persistence, self-esteem or optimism. ...
Article
Full-text available
Understanding how personality (e.g., DeNeve and Cooper 1998; Steel et al. 2019) and intellectual abilities (e.g., Gottfredson 1997, 2004b; Brown et al. 2021; Kulikowski 2021) contribute to shaping aspects of individuals’ lives is essential for in the advancement of many scientific disciplines such as psychology, management and medicine. However, although personality and intelligence and their impacts on life outcomes have been a subject of extensive research and interest among scholars across different disciplines, personality and intellectual abilities are most often studied in separation. When delving into the research on important life outcomes such as educational outcomes, career success, interpersonal relationships, mental health and overall well-being, it is vital to recognize that it is the joint influence of personality and intelligence that determines life outcomes (see, e.g., Deary et al. 2010; Damian et al. 2015; Cheng and Furnham 2012). Solely focusing on either intelligence or personality in isolation creates an artificial situation (Judge et al. 1999; Roberts et al. 2007), and only by considering them in tandem can we avoid oversimplifications and gain a deeper understanding of the processes that influence human lives (for more on the importance of ability–personality integration, see also Colom et al. (2019)). Therefore, this Special Issue serves a dual purpose. Firstly, we want to give a platform to papers that investigate the relationships between personality, intelligence and a wide range of life outcomes. Secondly, we aim to stimulate scholarly discourse by illuminating the often neglected and underexplored topic of the simultaneous effects of intelligence and personality in shaping individuals’ life trajectories. With this editorial, our objective is to highlight the main challenges that should be addressed to facilitate the research on the effects of intelligence and personality on life outcomes and offer a comprehensive overview of potential avenues for future exploration. We hope to inspire further research on personality and intelligence effects on important life outcomes.
... Typische Leistung wird als diejenige bezeichnet, die jemand dauerhaft im Alltag zeigt ("will do", Dalal et al., 2014). Letztere hängt natürlich auch von den Fähigkeiten ab, wird aber wesentlich von der aktuellen Motivation der Person bestimmt (Klehe & Anderson, 2007), also der Ausdauer und der Anstrengung. Für drei im Modell von Koopmans et al. (2011) (Lohaus, 2009). ...
... The alternative is to administer DT tests without explicitly directing examinees to "be creative" or generate original ideas because it can be useful to understand what individuals do spontaneously (Acar et al., 2020). Thus, when non-explicit (standard) instructions are given, the results may be more indicative of DT as it occurs in the natural environment (natural performance), while explicit instructions may be more indicative of examinees' potential and what is possible (maximal performance) (Klehe & Anderson, 2007). Such potential is of particular interest when DT tests are used to identify students for gifted programs. ...
Article
Divergent thinking (DT) tests are sometimes used to select students for gifted programs. Studies on these tests, mostly conducted on non-gifted students, suggest that performance is influenced by the type of instruction given (standard vs. hybrid "be fluent AND original") and time-on-task. The current study aimed to examine the effect of instructions and time-on-task on divergent thinking performance in gifted and non-gifted students in a 2 [gifted versus non-gifted] × 2 [standard versus hybrid instructions] design. The results showed that gifted students outperformed non-gifted students in fluency, while no significant difference was found between the two groups in originality. Creativity instructions improved both originality and fluency scores in verbal but not figural tests. As for time-on-task, gifted students took more time when completing DT tests as well as when they were given explicit instructions to "be creative." Implications for gifted identification are discussed.
... High capacity of organizational adaptation is required to implement the agile working regime, and the development of ICT tools and services is a critical driving force behind this process. Though most studies in the literature related to the development of remote work types (see, for example, Rapp et al., 2006;Klehe & Anderson, 2007;Oliva, 2017) were mainly concerned with productivity, well-being and life-work balance issues, a number of contributions particularly highlighted the value of ICTs in the innovation of organizational processes (Albano et al., 2019;Butera 2020) and their role in the redefinition of places, times and working methods towards a progressive flexibility of activities. Furthermore, Loré & Frey (2020) pointed out that the realization of agile working requires the acquisition of digital skills also at the individual level beyond the organizational one, in line with what other authors theorized in the years preceding the pandemic (see Chiaro et al., 2015). ...
Chapter
Full-text available
This chapter wants to shed light on the value of the autonomy in the organization of the individual work, with respect to the production of new scientific knowledge in non-university research institutions. The empirical base is the experience of smart working implemented in the Public Research Organizations (PROs) during the COVID-19 pandemic emergency, and the individual adaptation/reaction of the research personnel in two Italian PROs: the National Research Council (CNR) and the National Institute for Astrophysics (INAF). During the health-emergency period, scholars were required to work for most of the time at home, and several restrictions on free movement of people were imposed. The aim of this paper is to understand how this special condition influenced the activities of researchers and technologists and the production of original research work, thus impacting positively or negatively their creativity –namely, their attitudes of going beyond the exploitation of existing capabilities and routines.
Article
Full-text available
The study aims to determine the influence of academic performance to employability of the students in the field of education.The experimental research design will be structured for this study in which numeric score rating gathered the perception of the respondents observed in a structured pattern. This research design will deal at the problem of the study scientifically, to establish a clear cause and effect of Education Major students’ academic performance to their employability.The students’ perception on their employability in the field of education revealed that many disagreed their course is very important in their future work, and expressed that they are not more likely to work on a private school Majority of the students disagreed that they are more likely to work on a public school, and shared that their field of study is in-demand in terms of work opportunities in the education sector. Further, more students disagreed on the other indicators which are; they are confident that their skills and knowledge will match their work and confident that they will work in teaching or education sector, teachers are more likely to teach students than doing admin works, student skills are honed enough to move in an education work environment and they chose their field of study based on work demands of the education sector, teachers are more likely to do admin works than teaching students, students are more ready to teach and experience work environment, and lastly students already planned where they are going to work.
Article
Full-text available
The Education Program has a significant impact on students' academic performance and employability in the education sector. The study aimed to understand how students' perceptions of their academic record and their influence on their employability were analyzed. The study focused on English major students in China who expressed dissatisfaction with their academic performance and the idea that their course was relevant to their future employment. Many students disagreed with the idea that they are more likely to find employment in a private school and believed that their field of study is in high demand. They also disagreed with other indicators such as confidence in their skills and knowledge, the role of teachers in teaching students, and the readiness of students to work in an education work environment. The study concluded that a student's ability to advance in the education sector is significantly influenced by their academic record, which is a result of the correlation between the two.
Article
Purpose In this study, the authors explore the association between employee education level and the efficiency of corporate investment using data from a sample of Chinese listed firms during the period from 2011 to 2018. By examining the impact of education on investment efficiency, the authors' study provides valuable insights that contribute to a deeper understanding of the underlying economic mechanisms related to education. Design/methodology/approach The authors conduct multivariate regression analyses to examine the relationship between investment efficiency (following Richardson, 2006) and the level of employee education, along with a series of control variables. To ensure the reliability of the authors' findings, the authors subject the their results to a comprehensive set of robustness tests, such as a staggered difference-in-difference (DiD) regression approach, an instrumental variable (IV) method and the use of alternative employee education level and investment efficiency measurements. Findings The findings offer compelling evidence that higher levels of education have a positive impact on firms' investment efficiency, and this effect remains robust across various model specifications and endogeneity considerations. Moreover, the influence of education is more pronounced in firms that prioritize employee training, maintain effective internal communication and offer attractive financial rewards. Furthermore, the results suggest that the relationship between education and investment efficiency is influenced by the firms' business nature and competitive environment. Factors such as business complexity, labor intensity and business location also play a role in shaping the impact of education on investment outcomes. Originality/value The study emphasizes the crucial role of education in influencing investment decisions and performance within firms. By delving into this previously unexplored area, the authors' research contributes to the existing literature, establishing that the level of employee education is a significant determinant of corporate investment efficiency. This valuable insight has substantial implications for firms aiming to enhance their investment decision-making processes and overall performance. Understanding the positive impact of education on investment efficiency can empower organizations to leverage their human capital effectively and achieve better investment outcomes, ultimately contributing to long-term success and competitiveness in the market.
Article
Two experiments indicate that the areas of social facilitation and social loafing are complementary and can be conceptualized in terms of self-efficacy theory. In Experiment 1, efficacy expectancies were manipulated by false performance feedback, and outcome expectancies were manipulated by 3 group conditions (alone, coaction, and collective). In Experiment 2, efficacy expectancies developed spontaneously while participants performed easy or difficult tasks, and outcome expectancies were manipulated by 3 evaluative conditions (alone, evaluation, and no evaluation). Consistent with social facilitation research, high-efficacy evaluated participants performed better than alone participants, whereas low-efficacy evaluated participants performed worse than alone participants. Consistent with social loafing research, high-efficacy nonevaluated participants performed worse than evaluated participants, whereas low-efficacy nonevaluated participants performed better than evaluated participants.
Article
Understanding and predicting behavior in organizations requires a consideration of person and situation factors, and how these factors interact. This paper develops and longitudinally tests a model of person-organization fit (POF). POF is defined, and antecedents (selection and socialization) and consequences (commitment, performance, and tenure) are examined.
Article
A solution is suggested for an old unresolved social psychological problem.
Article
Hotelling's test of significance for difference in efficiency of predictors is reformulated in terms of regression analysis. A test proposed by Healy is shown to differ from Hotelling's test in general.
Article
The motivation constructs of Need for Cognition, Liking in Thinking, and Typical Intellectual Engagement, NEO Five-Factor Inventory factors, and RIASEC interests along with age were used to predict verbal tasks completed under timed (maximal performance) or untimed (typical performance) conditions. The results from a sample of 341 participants (245 female and 116 male), ranging in age from 18 to 71 years (M = 29.33, SD = 12.08), indicated that the men and women diverge in the relationships between their noncognitive domains and the verbal tasks. That is, women's verbal abilities are influenced by their age, and personality, interests, and motivations, whereas only personality and age contributed to verbal abilities in men. The results also displayed that the context in which a cognitive task is performed induced either maximal or typical efforts with different relationships with the predictor variables. The implications of the findings in relation to workplace training and development are discussed.