. . .11. I. i;
. Luwﬁm ﬁaﬁ. .
.I O,

f
9

hi. I ;
t . 2..
.& a: ..

. .5.

.41..”

ul- Mn.

s
1.! udﬁmm‘.
; .3.€Yuzn..n¢

2 .Iv1

. L .. -
. .. crawl»...
. ......w.L
. a , 1...? «We. a». .
.wumanu...nwucu..i.»mmumk . u
, 17. a a:

£59 . .
5.1, . 1...».

i . a. .. . _ , . .1... ..,.fu:.l 1,71,, .

ﬁfmﬁrwyf .. . . . , raﬁﬂm‘wiipi Jim.)

. . P . r . . . ,. “nit-(v.1 .I..ph.’| .

«raﬂrkuluﬂxﬂﬁw . . _ . . .. . . . . .. . _ . numb..- «nudimb...
,. .., . .. ,u. . . .. . .

.n‘ild‘il
: :45" ... all} .
, iii“: .. -25.-- ,
I 4.11.133:
”My 50
I» e « ﬂ...

.tuﬂ

 

J3R1I. 4 x

I.

 

~. ,‘..,r
.52’?

1'91

 

1.1 uiuldunuk
. .. ). 51H..I.ixifll:.llllrllt

‘ It. .3'0}: :Iii' 1".I. .v I . ‘
l n. J tVJ”.I'L.I.KOt!. , l‘ . ‘n

g . . . .. , . . Funﬂqén
$23!...tkﬁwghﬁ. .w .. . . . ..
.

V .‘L. 4

I... I‘.nu.¢$hlr|!”.nﬂn.lllm
.[vrrl bunnilfourro: H.1‘II‘IK.F;HWN (u...

a . . «m .ﬁﬁnﬂmui 1v. .znT. a!

rampedf. . .

n. \.
l (..

 

IIIII‘
‘l'

I‘l
I!

ANTS ATEU

:I IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

’7 302058 6271

                                

 

 

 

 

 

 

 

 

This is to Certify that the

dissertation entitled

AN INVESTIGATION OF THE DETERMINANTS AND
CONSEQUENCES OF INDIVIDUAL DIFFERENCES IN
MEASURES 0?” SPEED

presented by

Matthew Ross Smith

has been accepted towards fulﬁllment
of the requirements for

Ph. D. degree in Psychology

 

 

‘%«/ .JMW

 

Major professor

Date I2///‘///7

MSU is an Afﬁrmative Action/Equal Opportunity Institution 0-12771

 

 

LIBRARY

Michigan State
University

 

 

A

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11m wCIRC/DateDquGS-p.“

AN INVESTIGATION OF THE DETERMINANTS AND CONSEQUENCES OF
INDIVIDUAL DIFFERENCES IN MEASURES OF SPEED
By

Matthew Ross Smith

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Psychology

1999

ABSTRACT
AN INVESTIGATION OF THE DETERMINANTS AND CONSEQUENCES OF
INDIVIDUAL DIFFERENCES IN MEASURES OF SPEED
By

Matthew Ross Smith

The speed with which individuals react to stimuli might have a large inﬂuence on
a company’s competitiveness and an individual’s success. While current research
suggests a number of variables that can predict individual differences in speed, empirical
veriﬁcation has not been fully conducted. The current study addressed two major
research objectives. First, the question as to what might inﬂuence scores on speeded tests
was investigated by examining both noncognitive and cognitive factors in relationship to
measures of speed. Second, if speed is considered an important individual difference, the
question as to how it relates to measures of job performance was investigated. Particular
emphasis was placed on instances where Speed of response is an important dimension of
job performance and the effects of cognitive ability have been controlled.

The current study used a multiple hurdle, personnel selection context to examine
both research questions. In the study, participants were job applicants who applied for
entry-level positions. As part of the process, participants went through two selection
stages (testing and assessment). In testing, participants completed measures of perceptual
speed, psychomotor speed, cognitive ability, and various noncognitive factors. In
assessment, participants completed a group exercise where they were assessed on

measures of job performance including total number of products produced, proportion of

defects built into products, and ratings on three dimensions (work orientation, problem-
solving, and team skills). Such measures were used to examine the second objective of
the research, namely, the incremental validity of speed.

For the ﬁrst research objective, 853 participants completed both the measures of
speed and the assessment of cognitive ability and noncognitive factors proposed to be
related to measures speed. Mixed support was found for the hypothesized relationships
between measures of speed and factors with the strongest relationships between speed
and measures of conscientiousness, polychronic orientation, comparative anxiety, and
cognitive ability. For the second research objective, 531 participants completed the
group exercise measuring various aspects of job performance. Results found that
measures of speed consistently added ineremental validity to the prediction of job
performance over cognitive ability, particularly with job performance measures
emphasizing effort-based concepts (e.g., number of products produced, quality of

products, and work orientation rating).

Copyright by
Matthew Ross Smith
1 999

ACKNOWLEDGEMENTS

So many people have given me the support and assistance needed to make this
dissertation a reality. I hope I do them justice in the following short paragraphs.

I would like to thank the faculty members at Michigan State for guiding my
development throughout my time at Michigan State and in the years since I have left:
Kevin Ford, Steve Kozlowski, Rick DeShon, Dan Ilgen, and Neal Schmitt. In particular,
I would like to extend a special acknowledgement to Neal who served as my dissertation
chair. Not only did Neal serve as a model of excellence throughout my time at Michigan
State, but he also had the conﬁdence in my ability to ﬁnish the dissertation in the years
Since I left Michigan State to work at Aon Consulting and readily made the available time
to make sure I was fully prepared to ﬁnish it. Hopefully, I can emulate those personal and
professional qualities in the years to come. In addition, I would also like to thank my
fellow graduate students at Michigan State for their camaraderie both academically and
personally and for making graduate school a little easier to bear. Without making this an
awards speech, I’d like to thank Jose, Ron, Jen, Dennis, Whit, Stan, El, and Rob for
providing insights, commentary, and support in my early years at Michigan State and Bill,
David, Catherine, Dan, Ken, and Ann for similar assistance in my later years.

Since I joined Aon Consulting full time in March 1997, there are many people
who have supported my efforts to ﬁnish this dissertation. First, I’d like to thank Tom
Gawel and Dan Galant for providing me the perfect research setting and data and Becky

Gaffey for ensuring that the data was collected despite numerous administrative barriers.

Without their assistance, I might still be looking for a sample to collect the data, a
daunting task for someone trying to ﬁnish their dissertation and work full-time. I’d also
like to thank Renee Kennedy for her help in bringing the manuscript together and not
burying me in detail. Most of all, I’d like to thank both Fred Smith and Kirk Rogg who
have not only served as outstanding professional mentors, but have kept me on track to
complete this work even in light of competing work demands. Without their support, it
would have been very easy to stay ABD.

On a more personal note, I also want to thank God for giving me the necessary
KSAs needed to make this dissertation possible. I want to thank my family, especially
my parents, for teaching me the value of striving to do my best, and providing me with
the support I needed, not only through graduate school, but throughout my life. I’d like to
indirectly thank my brother Todd who was completing his Ph.D. in Physics at the same
time as me for providing some necessary competitive motivation. Finally, I’d like to
thank my wife and family for not only providing me with love and support every step of
the way, but for constantly reminding me what was really important in life. This
dissertation was a four-year affair that saw our lives change dramatically with the addition
of three children, Samantha, Jack, and Mark, as well as increased work and family
responsibilities. In particular, my wife made great sacriﬁces over those years to provide
me the time I needed to complete this dissertation. I believe I said the same thing on my

thesis, but this accomplishment is as much hers as it is mine.

vi

TABLE OF CONTENTS

LIST OF TABLES .............................................................................................................. x
LIST OF FIGURES ........................................................................................................... xii
INTRODUCTION ............................................................................................................... 1
Measuring Speed ..................................................................................................... 5

Speed Versus Level Distinction in Testing ................................................. 6

Empirical Evidence for Speed/Level Difference ......................................... 7

Dimensions of Speed ................................................................................. 11

Measurement Of Speed .............................................................................. 17

Relationships with Measures of Speed .................................................................. 22
Cognitive Inﬂuences: General Cognitive Ability ...................................... 23

Noncognitive Inﬂuences: Personality Traits ............................................. 24
hnpatience/Irritability and Type A Behavior Pattern .................... 25

Conscientiousness ......................................................................... 28

Temporal orientation ..................................................................... 31

Self-monitoring ............................................................................. 32

Situational arousal ......................................................................... 35

Other Inﬂuences ........................................................................................ 37

Age ................................................................................................ 37

Applicant reactions and motivations ............................................. 38

Job Performance .................................................................................................... 41
Research Study: Relationships with Measures of Speed ....................................... 45
General Cognitive Ability and Speed ........................................................ 46

Mechanisms of Noncognitive Inﬂuence .................................................... 49
Impatience/Irritability .................................................................... 5 1

Conscientiousness ......................................................................... 52

Temporal orientation ..................................................................... 54

Self-monitoring ............................................................................. 55

Circadian rhythms ......................................................................... 57

Anxiety .......................................................................................... 59

Other Inﬂuences: Covariates ..................................................................... 60

Research Study: Incremental Validity of Speed on Job Performance ................... 62
METHOD .......................................................................................................................... 65
Participants ............................................................................................................ 65
Measures Investigating Relationships with Speed ................................................ 66
General cognitive ability ........................................................................... 66
Impatience/Irritability ................................................................................ 67
Conscientiousness ..................................................................................... 67

vii

Polychronic vs. monochronic orientation .................................................. 67

Pace ........................................................................................................... 68
Self-monitoring ......................................................................................... 68
Circadian rhythms match ........................................................................... 68
Anxiety ...................................................................................................... 69
Covariate Measures ............................................................................................... 69
Measures of Speed ................................................................................................ 69
Attentive/Perceptual speed ........................................................................ 69
Psychomotor speed .................................................................................... 70
Measures of Job Performance ............................................................................... 7O
Subjective measures .................................................................................. 71
Objective measures ................................................................................... 72
Procedure ............................................................................................................... 72
Testing phase ............................................................................................. 72
Assessment phase ...................................................................................... 73
Data Analysis ........................................................................................................ 78
RESULTS .......................................................................................................................... 81
Relationships with Measures of Speed .................................................................. 81
Bivariate relationships ............................................................................... 81
Covarying out age, test motivation, and perceived face validity ............... 88
Covarying out age, test motivation, perceived face validity,
and cognitive ability .................................................................................. 94
Incremental Validity of Measures of Speed ........................................................ 100
Relationship with performance dimensions related to effort and
discipline ................................................................................................. 107
Relationship with performance dimensions unrelated to effort and
discipline ................................................................................................. 1 10
Correction for criterion unreliability and restriction of range ................. 116
DISCUSSION ................................................................................................................. 1 18
Summary and Implications of Major Findings .................................................... 118
Relationship with measures of speed ...................................................... 118
Incremental validity of measures of speed over cognitive ability ........... 124
Practical and theoretical implications ..................................................... 127
Study Limitations ................................................................................................ 130
Determination of causality ...................................................................... 131
Perceived face validity ............................................................................ 131
Errors in perceptual speed measure ......................................................... 131
Measurement of typical performance ...................................................... 132
Directions for Future Research ............................................................................ 132
APPENDIX A: Power Analyses ........................................................................... 135

viii

APPENDD( B: hnpatience/Irritability Measure — Spence, Pred, & Helmreich

(1987) .......................................................................................... 136
APPENDIX C: Achievement Strivings Measure - Spence, Pred, & Helmreich

(1987) .......................................................................................... 137
APPENDD( D: Monochronic versus Polychronic Orientation — Kaufman,

Lane, & Lindquist (1991); Tuttle (1996) ..................................... 138
APPENDIX E: Pace of Life — Tuttle (1996) ........................................................ 139
APPENDD( F: Self-Monitoring — Snyder (1974) ................................................ 140
APPENDIX G: Circadian Rhythm Scale — Smith, Reilly, & Midkiff (1989) ...... 142
APPENDIX H: Comparative Anxiety Scale - Arvey, Strickland, Drauden,

& Martin (1990) .......................................................................... 145
APPENDD( I: Covariate Measures ..................................................................... 146
BIBLIOGRAPHY ........................................................................................................... 149

ix

Table 1:

Table 2:

Table 3:

Table 4:

Table 5:

Table 6:

Table 7:

Table 8:

Table 9:

Table 10:

Table 1 1:

Table 12:

Table 13:

LIST OF TABLES

Types of lSt Order Speed Factors in Cognitive Ability Domains (Carroll,
1993) .......................................................................................................... 14

Summary of Means, Standard Deviations, Reliabilities, and Zero-Order
Intercorrelations Investigating Relationships with Measures of Speed ..... 82

Result of Simple Regression Predicting Perceptual Speed
010 Covariates) ......................................................................................... 85

Result of Simple Regression Predicting Psychomotor Speed
(No Covariates) ......................................................................................... 87

Summary of Partial Correlations Investigating Relationships of Measures
of Speed Controlling for Age, Test Motivation and Perceived Face
Validity ...................................................................................................... 89

Result of Simple Regression Predicting Perceptual Speed Controlling for
Age, Test Motivation, and Perceived Face Validity .................................. 91

Result of Simple Regression Predicting Psychomotor Speed Controlling
for Age, Test Motivation, and Perceived Face Validity ............................ 93

Summary of Partial Correlations Investigating Relationships of Measures
Of Speed Controlling for Age, Test Motivation, Perceived Face Validity,
and Cognitive Ability ................................................................................ 96

Result of Simple Regression Predicting Perceptual Speed Controlling for
Age, Test Motivation, Perceived Face Validity, and Cognitive Ability.... 97

Result of Simple Regression Predicting Psychomotor Speed Controlling
for Age, Test Motivation, Perceived Face Validity, and Cognitive

Ability ........................................................................................................ 99
Hypothesis Summary Investigating Relationships to Measures of

Speed ....................................................................................................... 101
Summary of Means, Standard Deviations, Reliabilities, and Zero-Order
Intercorrelations Investigating the Incremental Validity of Speed .......... 104
Summary of Partial Correlations Investigating the Incremental Validity of
Speed Controlling for Cognitive Ability ................................................. 105

Table 14:

Table 15:

Table 16:

Table 17:

Table 18:

Table 19:

Table 20:

Result of Simple Regression Predicting Proportion of Defects Built
Controlling for Cognitive Ability ............................................................ 108

Result of Simple Regression Predicting Total Assemblies Built
Controlling for Cognitive Ability ............................................................ 109

Result of Simple Regression Predicting Performance Rating of Work
Orientation Controlling for Cognitive Ability ......................................... 111

Result of Simple Regression Predicting Performance Rating of Problem
Solving Controlling for Cognitive Ability .............................................. 113

Result Of Simple Regression Predicting Performance Rating of Team
Skills Controlling for Cognitive Ability .................................................. 114

Signiﬁcant Differences Between Incremental Validity Coefﬁcients for
Effort-Based vs. Non Effort-Based Performance Dimensions ................ 115

Corrected Incremental Validity Over Cognitive Ability Between Measures
of Speed and Job Performance ................................................................ 117

xi

LIST OF FIGURES

Figure l. Hypothesized Interaction Between Self-Monitoring and Personality
Characteristics on Measures of Speed ....................................................... 58

Figure 2. Relationships with Measures of Speed ...................................................... 61

Figure 3. Interaction Between Polychronic Orientation and Self-Monitoring on 95

xii

INTRODUCTION

The speed in which individuals complete tasks has been of interest to researchers
since the beginnings of psychology. Oﬁen, speed was equated with measures of
intelligence for such historic researchers as Galton and Spearman who equated constructs
such as general sensory discrimination with general intelligence (Vernon, 1987). Other
researchers such as Thorndike and colleagues (1926) distinguished the difference
between level of performance and speed of performance as logically distinct components
of intellectual measurement. However, speed has more recently been viewed more as a
potential byproduct and hindrance in the measurement of other abilities, typically
cognitive abilities (Hartigan & Wigdor, 1989; Peterson, 1993; Rindler, 1979). While
speed/time issues have been considered by some researchers (i.e., McGrath & Rothchrod,
1983), an investigation of its determinants and impacts has not been adequately
described. There are three reasons why speed would be an important topic for researchers
today: (1) the importance of speed in today’s jobs, (2) the unexplored relationships
between speed and noncognitive factors, and (3) the unexamined validity relationship
between speed and measures of job performance.

First, speed of response has emerged as a critical feature in an organization’s
competitive strategy whereby a customer is offered a competitive value in terms of
reduced time for products and services (Carnevale, 1991; Muchinsky, 1997; Schuler &
Jackson, 1987). The US. economy is currently converting ﬁ'om what has been

traditionally a mass production economy to one of a service nature (Heneman &

Heneman, 1994). With such changes, new competitive standards that characterize the
service economy include expectations of high productivity levels, elevated awareness of
customer service, and an increased pressure on providing goods and services in a timely
manner (Camevale, 1991). An important skill applicable to jobs with a service focus
includes the ability to allocate time and resources to complete tasks in which speed can
play a vital role (SCANS, 1992). In addition to economic changes, the downsizing of
employee workforces and the increased workload of the survivors of such interventions
can also increase the role of speed in the performance of organizational duties,
particularly those duties that are more repetitive and routine. When employees need to
acquire the tasks of displaced workers, time allocation and the speed in which tasks are
completed can be important not only to organizational performance but to survival in the
organization as well. The role speed can play in the effective performance of a job can
also vary as a function of the level of speed inherent in the performance criterion of the
occupation. Speed can play a role in such speciﬁc tasks as searching radar to locate
ships, identifying errors in documents and products, preparing food in a restaurant at rush
hour, and scanning stock market prices to make decisions. The speed in which an
individual performs such tasks can be the critical point in determining whether or not
someone is successful on the job. Other organizational implications of speed include the
decrease of stress through the overload reduction and adaptation to organizational
cultures that expect people to adhere to certain time and scheduling demands and

pressures.

The questions concerning what factors are related to speed of response falls along
the familiar lines of performance being determined by ability and motivation. Abilities
related to speed oﬁen fall under the category of cognitive abilities (i.e., Carroll, 1993).
While a relationship between reaction time and intelligence has been demonstrated (i.e.,
Jensen, 1993), the equivalence of the two constructs is less than complete. Many
researchers agree that such a relationship between the two constructs exists, however,
they also agree that noncognitive factors such as motivation, persistence, carefulness,
anxiety, personality, temperament, interest, etc., play a role in the measurement of speed
and should be taken into account when modeling test/task performance (Carlson &
Widaman, 1987; Carroll, 1993; Jensen, 1980). In addition, factor analyses of speeded
tests oﬁen demonstrate an unique speed factor in addition to that represented by cognitive
ability (Carroll, 1993; Hunter, 1986). However, what the relationship is between
noncognitive factors and measures of speed has never been fully examined or empirically
demonstrated beyond the mere conversation level. When the difﬁculty of the tasks are
low, the inﬂuence of cognitive abilities may be reduced whereas relationships with
noncognitive factors may be increased. Empirical veriﬁcation of the existence of
relationships between noncognitive factors and speed could be an essential missing step.

While some research in personnel selection contexts has demonstrated that
speeded tests of cognitive ability add little to the prediction of job performance (i.e.,
Hunter, 1986), such research has failed to explore the nature of the criterion in terms of
speeded performance (Kendall, 1964) or has used cognitive ability tests that were

speeded, thus inﬂating part-whole correlations between power tests of cognitive ability

and measures of speed (Carroll, 1993). While not denying the importance of cognitive
ability in the prediction of a variety of constructs (Brand, 1987), a further examination of
criterion issues in job performance could add to the understanding of the inﬂuence of
speeded tests. With a rising increase in the signiﬁcance of speed in today’s workforce,
this could lead to an expansion of the criterion space of job performance to include
speeded components as an important factor. With such a rise in time pressure and
service, the relationship of speed to measures of job performance could increase over and
above the relationship of cognitive ability with job performance.

Overall, the objective of this research is twofold. First, the question as to the
relationship between various individual difference measures and speed will be
investigated by examining a variety of noncognitive and cognitive factors. Second, if we
consider speed to be an important individual difference, how does it relate to measures of
job performance, particularly in instances where speed of response is an important
dimension of overall performance and the effects of cognitive ability have been
controlled.

The subsequent literature review takes the following approach to outline the major
issues involved in the accomplishment of these objectives. The ﬁrst step is to investigate
the measurement of speed and the potential measurement concerns involved with speeded
tests that must be considered when thinking about the conceptual issues. Topics involved
in the consideration of speeded test performance include the conceptual and empirical
difference between speed and level constructs in testing, the examination of measures of

speed as an unidimensional versus multidimensional construct, the interplay between the

way speed is measured and the conceptual issues affected, and speed versus accuracy
concerns when time pressures are central in the testing scenario. Once the elaboration of
the conceptual domain of speed is completed, the next step examines the predictor space
in terms of what would predict performance on measures of speed. Using information
based on what we know about performance on speeded tests and on the nature of both
cognitive and noncognitive constructs, literature will be reviewed on variables that could
be related to performance on these measures.

Once that has been achieved, the focus then switches to the second objective of
the research, namely the incremental validity of measures of speed in the prediction of
job performance. The nature and multidimensionality of job performance, using job
performance research as a basis, is examined with particular emphasis being placed on
performance dimensions that should be related to individual differences in measures of
speed. The type of performance setting that will be used in this research, the assessment
center, will also be highlighted to examine various construct dimensions and operational
issues. Based on the literature outlined in the previous sections, the ﬁnal section of the
introduction outlines the hypotheses of the current research study beginning with those
associated with the ﬁrst objective, relationships to measures of speed, and ending with the

second objective, incremental validity issues.

Measuring Speed
In most practical applications of testing, particularly the testing of cognitive-

related abilities, time limits are placed in the testing scenario. Time limits are placed in

tests to ensure that testing is administered in a reasonable, efﬁcient amount of time or to
assess speed of performance if rate—of-work is deemed an important factor to be assessed.
In early group testing research, evidence was gathered to support the assumption that
speed (time limits) and power (no time limits) tests served as interchangeable measures of
the same construct, such that the speed with which people solve problems is indicative of
the level of difﬁculty of the problems they can solve (Mon'ision, 1960; Rindler, 1979).
These conclusions regarded speed as an inconsequential inﬂuence on the assessment of
cognitive abilities and disregarded the effects of speed on test scores. However, such
statements have been subsequently challenged to view speed and power as separate
components of test scores.
Speed Versus Level Distinction in Testing

Speed can be broadly deﬁned as the time or rate at which tasks of a speciﬁc kind
and difﬁculty are performed. This is conceptually different from notions of level which
deal with the point of difﬁculty at which an individual can perform with a certain amount
of accuracy (Carroll, 1993). In traditional testing, level is matched to the construct of
interest that the test proposes to measure (e. g., intelligence). The correctness of the
response is taken as an indication of an individual’s level of ability with respect to the
task or range of tasks that deﬁne an ability. Anastasi (1988) describes a pure speed test as
one in which individual differences in test performance rely entirely on the speed of
performance. The test is made up of items that have low difﬁculty and has a time limit
that is short enough so no one is able to ﬁnish the test. In such a test, the construct of

interest is speed itself. A pure power test has no time limit or a time limit that is long

enough to allow everyone to complete all the items, but the item difﬁculty is steeply
graded and includes some items which are too difﬁcult for anyone to solve. It most
testing situations, it is a rare case when we either have a pure speed or a pure power test
due to practical issues such as making sure testing is administered in a reasonable,
efﬁcient amount of time. In most cases, test scores contain both speed and level variance.
While conceptually different, the next question concerning speed versus level concepts
focuses on whether there are any empirical differences between the two.
Empirical Evidence for Speed/Level Difference

Not only are speed and level different conceptually, but empirical studies have
found that most tests have speed and level components that are often uncorrelated
(Carroll, 1993). Typical testing scenarios introduce speed variance along with the
variance of the construct the test is designed to measure into total test variance, such that
speed of work inﬂuences test scores as much as accuracy/level. Early group testing
research that disregarded the effects of speed on test scores was originally re-examined by
Davidson and Carroll (1945). They highlighted the interchangeable misconception of
speeded and power tests by indicating that the correlation between time-limit scores and
power scores are part-whole correlations in that the time-limit scores measure the same
construct as the power score and thus contain substantial amounts of common variance.
This would artiﬁcially raise the level of the correlation (Carroll, 1993). Such problems
call for the use of separate tests to measure both speed and level components. As
mentioned above, speed tests in their purest sense contain items with low difﬁculty and

are measured by the number of items completed in the time interval or the time it takes to

complete the whole task. Level or power tests in their purest sense have no time-limits;
person’s scores on the test are the number of items that are correct (Anastasi, 1988).
While practically difﬁcult to have pure measures, separate tests should be used.
Correlational demonstrations of the distinctive contribution of speed and level to
time-limit scores was ﬁrst found in work by Baxter (1941), Davidson and Carroll (1945),
and later by others (see Rindler, 1979 for review) in which the relationship between speed
(rate of response) and level (response accuracy) scores was found to be negligible. The
empirical differences between speed and level factors on test-taking has also been
examined by factor-analytic methods. Probably the ﬁrst investigator to ﬁnd a general
speed factor in a series of “mental ability” tests was DuBois (1932) who found a general
factor in a series of speeded tests that had low correlations with number-correct on
arithmetic reasoning and vocabulary tests. In an early study examining the relationship
between speed and level effects, Tate (1948) had high school students individually tested
with selected items, ranging widely in difﬁculty on four tests: arithmetic reasoning,
number series completion, sentence completion, and spatial relations. Performance on
each item was timed in seconds and converted to logarithmic units along with level
scores. For each of the four tests, correlations between speed and level scores were
negligible while correlations among the level scores and among the speed scores were
more substantial. Myers (1952) in a study of the factorial composition of different speed
portions of a nonverbal reasoning test found two orthogonal factors interpreted as the
ability to answer problems correctly and the tendency to answer the problems quickly. In

a reanalysis of Lord’s (1956) study of speed factors in a series of tests, Carroll (1993)

found a seven-factor principal factor solution which extracted two orthogonal second-
order factors that could be deﬁned as speed (ﬁrst-order factors of numerical, perceptual,
and verbal speed) and level (ﬁrst-order factors of visualization, quantitative and
arithmetic reasoning, and lexical knowledge) factors. In a reanalysis of Mangan’s (1959)
study on speed and power variables, Carroll (1993) also extracted two orthogonal second-
order factors that could be deﬁned as broad speediness at performing easy tasks and level
in performing difﬁcult tasks. Based on Thissen’s (1980) work on estimating a latent trait
model for speed and level parameters on items from three tests (the Progressive Matrices,
a verbal analogies test, and a spatial ability test), Carroll (1980) displayed loadings in a
varimax-rotated principal factor-analysis matrix on two uncorrelated factors (speed and
level) that accounted for 77% of the variance between a six parameter correlation matrix
(an ability and speed parameter for each test). Horn (1978; 1981) found that speed of
thinking (both in terms of speed in producing correct and incorrect answers) and power of
thinking were not highly correlated on intellectual problems of nontrivial difﬁculty. Even
though some people were quicker than others in solving problems, this quickness of
thinking was independent of the qualities of thinking represented by ﬂuid and crystallized
intelligence. However, the quickness of thinking factor on problems of nontrivial
difﬁculty was separate from a scanning speediness factor. Kyllonen (1985) analyzed a
series of more or less traditional psychometric tests that were administered to Air Force
personnel by computer so that both speed (latency) and accuracy (percent correct) scores
could be obtained. They reported that at the second-order, a general speed and general

level factor were obtained that were virtually uncorrelated (r = -.05).

While the above results concerning “rate-of-test taking” measures of speed point
to the distinction between speed and level constructs, the relationship between speed and
level constructs is more complex, such that certain cognitive dimensions of abilities
appear to show differential associations between speed and power. In other research
involving scores of more elemental measures of general processing speed, relationships
have been noted between speed and level. A variety of studies have found relationships
between reaction time in simple tasks with little intellectual content and nonspeeded
complex tests of reasoning ability (see Vernon, 1987 for a review). In general,
individuals with higher scores on intelligence tests tend to apprehend, scan, retrieve, and
respond to stimuli more quickly than individuals who score lower on intelligence tests.
The theory behind the relationship highlights the neurological basis vof intelligence based
on how fast information is transmitted, processed, and held in working memory (Jensen,
1993). Common paradigms used include (a) same/different judgments on visual displays
(Hunt, 1978), (b) choice reaction time where individuals hit the button of a target whose
light is activated (Jensen, 1987), (c) inspection time where individuals view two lines and
must judge which of the two is shorter (Deary & Stough, 1996; Nettelbeck, 1987), and
(d) neurological measures such as visual evoked potentials measuring nerve conduction
velocity or the speed with which electrical impulses are transmitted by the nervous
system (Reed & Jensen, 1992). These speed measures appear to have small, but
signiﬁcant correlations with level abilities, such that the size of the correlations appear to
be related to the complexity of information processing associated with the speeded task.

For instance, complex reaction time measures will have higher correlations with

10

intelligence than simple reaction time measures. However, such research does indicate
that even with simple reaction time tasks, cognitive ability, a level construct, does have a
signiﬁcant relationship with measures of general processing speed

In summary, results of studies of the relationship between speed and level are
more complex than the original proposition that speed and power scores were
interchangeable. First, speed and level are logically and empirically distinct aspects of
task performance with speed dealing with the rate at which tasks are performed and level
relating to the level of task difﬁculty at which someone can perform. Equating test scores
on speeded versus power versions of the test can be inappropriate based on the test-
maker’s purpose and the time-limit enforced. Second, while speed and level are distinct
components, there does appear to be a small relationship between measures of speed and
cognitive measures of level, particularly when the measure of speed increases in cognitive
complexity. Given that speed is a distinct, yet related component of tasks outside of level
constructs, the next question is whether the speed component is a general, unidimensional
construct across tasks or a more multidimensional construct.
Dimensions of Speed

While speed can be seen as logically distinct from measures of level, the
dimensions of speed assessed, whether multiple or unidimensional, need to be clariﬁed.
A large amount of research examining speed constructs in the domain of testing,
particularly cognitive ability testing, has employed factor analytic methods to investigate
the factor structure and content of speed variance in test scores. Some studies argue that

different measures of speed could be well represented by a general factor, in a sense, a

ll

speed “g,” accounting for a majority of the variance in measures of speed (Kyllonen,
1985; Levine, Preddy, & Thorndike, 1987; Lord, 1956; Miller & Vernon, 1992). Other
studies have argued that speed is more multidimensional having factors such as decision
time and movement time (Kranzler, 1990), cognitive speed and motor speed (Verster,
1983), reaction time, intellectual speed, speed of perception, and motor speed (Rimoldi,
1951), and attentive speediness and quickness in deciding on answers (Horn, 1985;
1988). However, such methods employed in both historical and more current studies
have methodological shortcomings insofar as the establishment of a clear understanding
of the dimensions of speed scores is concerned. The largest concern is the sample size to
number of tests ratio, particularly for many of the early studies. In most studies, the
typical procedure is to administer a large group of tests (cognitive ability and speed
measures) to a sample of 75-100 university students. Such procedures fail to ensure that
the groupings of factor structures are not simply effects of sampling error (Gorsuch,
1988; Nunnally & Bernstein, 1994). In addition, many of the measures of speed have
been vastly different from reaction time to number of items ﬁnished in a time period.
Therefore, a review of speed studies might give the clearest picture of the dimensionality
of speed.

The most recent and comprehensive review of measures of speed was conducted
in an examination of various measures of cognitive abilities by Carroll (1993). Carroll
found that most of the variance in test scores could be partitioned into (a) level factors
where probability for success decreases with task difﬁculties and (b) speed factors where

probability of success increases when time permitted to perform the task increases. Such

12

a classiﬁcation is based on a conceptual analysis of the range of tasks that measure the
factor to determine if variation is due to difﬁculty or to rate of performance.

In his review, Carroll (1993) examined the speed and level factors in various level
domains of cognitive abilities and found a variety of speed factors (see Table 1). To
examine whether these ﬁrst-order factors represented different dimensions of speed or
whether they represented a more general factor of “broad speediness,” Carroll expanded
upon this work by examining studies of second-order factors that could represent speed
variance. In all, Carroll (1993) found 57 token higher-order speed factors from 52 data
sets that were classiﬁed as broad speediness. In making further classiﬁcations of the
factors listed, Carroll examined the similarity of the ﬁrst-order factors that loaded on the
second-order factor and postulated three higher-order factors of speed. The ﬁrst category
that represented most of the factors was an attentive speediness factor containing such
ﬁrst-order factors as perceptual speed, numerical facility, and speed of test-taking. This
factor represents a quickness in identifying elements and distinguishing between
elements, particularly when there is a pressure to maintain attention (Horn, 198 8).

The second speed category tended to represent various kinds of reaction time
tasks such as the Hick paradigm. Reaction times in these tasks are typically measured on
very simple perceptual and cognitive tasks where individuals react to a stimulus by
pressing the button that corresponds to the one in which a light is activated. This factor
might be separate from the ﬁrst category for two reasons. First, this factor could
represent a more general or fundamental form of processing speed where cognition plays

little to no role or involves a simple decision that is more neuropsychological in nature.

13

Table l

Typps of lst-Order Smed Factors in Cognitive Ability Domains (Carroll, 1993)

 

 

 

DOMAIN SPEED FACTORS
Language Reading Speed
Rate of Work in Performing Verbal Tasks
Reasoning Speed of Reasoning Factor (similar to language’s rate of work)

 

Learning and Memory

Difficult to Separate Speed from Level

 

Visual Perception

Spatial Relations (rate of performing rotation tasks)

Closure Speed (rate at apprehending what is represented by a
distorted visual presentation)

Closure Flexibility (rate at ﬁnding a form embedded in another)

Perceptual Speed (rate of search & comparison of visual forms)

 

 

 

Auditory Reception none
Idea and Language Ideational Fluency (rate at producing responses that ﬁt criteria)
Production Naming Facility (speed at naming pictures and objects)

 

Associational Fluency (rate of producing responses that are drawn
from restricted classes associated with given stimuli)
Expressional Fluency (rate of producing meaningful discourse)
Word Fluency (rate of producing different verbal responses)
Sensitivity to Problems (rate of producing different ideas that are
pertinent to real or imagined problems)

Figural Fluency (rate of producing different ﬁgural designs)
Figural Flexibility (rate of producing different hypotheses about

ﬁgural and spatial problems)

 

l4

 

Second, the similarity of the tasks (i.e., reaction time) loading on these factors could
indicate that the method of measurement might be leading to higher relationships among
these tests.

A ﬁnal category of second-order speed factors could be interpreted as general
psychomotor speed that is primarily concerned with ﬁnger, hand, and arm movements
and is relatively independent of cognition (Carroll, 1993). Examples of tasks in this
domain include arm movement time (Carlson, Jensen, & Widaman, 1983; Kranzler,
1990), and psychomotor tasks such as packing blocks, tapping, and card sorting (Clausen,
1966; Verster, 1983).

Research emphasis in this study will only be applied to the ﬁrst and third
dimensions of speed: (1) attentive/perceptual and (2) psychomotor speed, respectively.
Focus on reaction time, while an important concept to the study of speed, will not be
examined for several reasons. First, the dimension, as mentioned above, could be a by-
product of the type of tasks that are employed in these measurement designs. Many of
these reaction time tasks that load on this factor require a quick response to detect
whether two lines are the same length, to detect which light is lit, etc.. Measurement is
based on how fast individuals can detect differences or identify stimuli. While
representing a more biological measure of speed or the efﬁciency of information
movement, such tasks might not transfer as readily to the type of job performance that
one would encounter in a work environment where a successful response to a task

requires more time to conduct a series of movements/decisions than a simple reaction to a

15

stimulus. Second, an rurique aspect of measures of speed is that there are more
opportunities to investigate relationships with more noncognitive factors such as the
importance of maintained effort or persistence in task completion. Individuals need to
perform the task at a certain rate in an extended time period to achieve a meaningful
outcome. Reaction time tasks instead rely on short bursts of speed in response to stimuli,
rather than engaging in speed over a time period. In speeded tests, noncognitive factors
can be explored in relationship to measures of speed, and how an individual can maintain
such speed over time. ' Such speeded tasks that require maintenance of a certain rate
might have more relevance to work environments than a quick response to an initial
stimulus. Third, issues of face validity and test-taker reactions in hiring situations have
become an important factor to consider in the candidate review process (Rynes, 1993).
Having job candidates complete reaction-time tasks where they react to lights or make
judgments on the similarity of lines could open the organization to queries concerning the
job-relevance of the test as well as fairness and legality. While not denying the reaction-
time paradigm as a valid measure of speed, for this research, speed will only be assessed
by examining only attentive and psychomotor speed tasks.

In summary, examining speeded performance on tasks seems to be represented by
three dimensions of speed: attentive/perceptual speed, reaction time, and psychomotor
speed. While all three are potentially meaningful, attentive and psychomotor speed
dimensions seem to have the most relevance to work in which the maintenance of a
speeded response over time is an important aspect of work performance. These two

dimensions will be examined in the current research study. Before proceeding to the

16

constructs with relationships to speeded performance, some time will be spent to examine
various measurement concerns that can have a direct impact on the conceptual
investigation of speeded test performance.

Measurement of Speed

As mentioned in the previous sections, a speed test in its purest sense is one where
individuals of varying levels of ability are given low difﬁculty items to solve in a certain
time period that is short enough so no one ﬁnishes the test. In an alternative way, speed '
can also be assessed by the amount of time it takes for an individual to complete a test
comprised of items at low levels of difﬁculty. Other more non-traditional measures of
speed are highlighted in research investigating the speed of information processing and
general cognitive ability. These measures of speed include simple reaction time or other
neurological measures such as nerve conduction velocity. While useful to investigating
the relationship between general cognitive ability and speed, these measures will not be
used in the investigation of speed in the current research study for various reasons
highlighted in the previous section examining the dimensions of speeded performance.

In terms of traditional paper-and-pencil tests, some issues need to be considered
prior to making an assessment of speed. Items on such tests can fall into four categories:
(a) items marked correct, (b) items with incorrect answers, (c) items considered and left
unmarked, and (d) unmarked items which the examinee had insufﬁcient time to consider.
Items that are marked wrong and items omitted after full consideration may be combined
into one category because test-takers considered the item but were unable to derive a

correct answer (Rindler, 1979). To assess speed, a separate test of speed rather than a

17

time-limit test of another construct of interest is more useful. A single test administration
of a different construct of interest is likely to yield an under-estimate of the speed factor
containing systematic error deriving ﬁ'om three sources: (1) failure to properly account
for items skipped over that individuals do not have sufﬁcient time to consider, (2) failure
to account for the number of difﬁcult items subjects do not read, but rather ﬁll in at
random as time approaches its limit, and (3) failure to account for effects of cognizance
of time limits on overall rates of response and consequent accuracy (Rindler, 1979).

Most traditional methods of assessing test speededness are insufﬁcient for the
present discussion because these methods view speed as a hindrance to measuring the
trait of interest rather than an interesting phenomenon in and of itself. Methods such as
tau (Cronbach & Warrington, 1951) and determinant analysis (Lord, 1971) compare the
performance of the same subjects in timed and untimed administrations of parallel tests to
estimate the extent to which the true standard scores or factor structures in the group
would be changed if time limits were extended (Morrison, 1960). However, some of the
methods used to assess speed as a byproduct can still be applicable to measuring speed.
As mentioned above, giving individuals a test with a larger number of low difﬁculty
items than can be ﬁnished in a certain time period would be an appropriate measurement
of speed with the number of items completed representing speed.

An issue that arises with the measurement of speed occurs when all the items are
not easily answered by examinees. Such difﬁculty could arise due to item content as well
as time pressure instituted by the test examiner. In cases where error is caused by

carelessness or time pressure rather than actual ability to successfully answer the item

18

correctly, traditional measures of item difﬁculty such as p—values may be misleading as to
the difﬁculty of an item by incorrectly indicating that an incorrect item is due to a lack of
ability. If given a power version of the test, an individual might be able to easily solve all
items and item difﬁculty would be minimal. When test items produce errors for
individuals, speed versus accuracy issues need to be addressed. In this case, error
becomes an issue due to the rise in the assessed difﬁculty (i.e., p-values) of the items in
the test such that individuals trade accuracy for speed, thus generating more errors.

The question arises as to how to handle incorrect versus correct responses in
speeded tests. When the rate of error is virtually non-existent, the issue of correct versus
incorrect responses is not pertinent. However, when errors occur such as when there is an
increase in time pressure or cognitive difﬁculty, the question of how to measure speed in
light of incorrect responses becomes applicable. One group of researchers claim that the
speed is more or less constant across incorrect and correct responses. Tate (1948) found a
high correlation of speed for correct and incorrect items in a free-response Arithmetic
test. Horn (1978, 1981) also found that speediness in obtaining correct answers is
correlated positively with speediness in obtaining incorrect answers. In a major review of
the domain of cognitive speed, Carroll (1993) found that speed factors can be obtained in
tests measuring level constructs and that the linear independence of speed and level
factors in such tests has been established.

The other side of the debate, led by David Lohman, has argued for the assessment
of both speed of response and accuracy of response in relationship to speed-accuracy

tradeoffs. In such tradeoffs, a participant, while under time pressure, would sacriﬁce the

19

chance of getting the item correct (i.e., “trade-oft”) to answer more items in the time
period. Such latency measures for incorrect responses might have complex determinants
beyond ability (Audley, 1973; Lohman, 1979). For instance, if an individual was given
unlimited time to ﬁnish a test, he/she would most likely be able to answer every item
correctly. However, since there is a time constraint, the individual might take less time
answering and double-checking items, thereby leading to more errors that are unrelated to
the ability to answer the item correctly. Such an increase in errors can be small while the
response time can be reduced substantially (Lohman, 1994). Other factors such as fatigue
(within individuals), response style, and carelessness (between subjects) can affect such
trade-offs (Lohman, 1994). In examining speed versus accuracy trade-offs (SATO), it
has been found that normal SATO curves are generally monotonically decreasing and
negatively accelerating (Lohman, 1986; Wickelgren , 1977). Lohman (1994) argues that
this nonlinear relationship makes the investigation of speed in test performance without
assessing accuracy inappropriate.

In summary, unless test performance is error-free, the assessment of speed should
take differences in accuracy or inaccuracy into account. In measures of speed where
errors occur, the above issues raised in speed—accuracy trade-offs should be addressed.
For the current research, speed-accuracy issues will be addressed in several ways. The
tests used to assess speeded performance will be error-free as much as possible to avoid
any issues with applicants making trade-offs to complete more items. Also, research
investigating speed-accuracy trade-offs is typically conducted in studies exploring

individuals in test-taking situations and how they respond to time pressures while

20

_ completing items on a test battery. While vitally important to the assessment of test
reliability and validity, it is more micro-oriented than the current study investigating more
stable general differences in speed among individuals. As such, the current research will
attempt to control only those factors that most inhibit an accurate aSsessment of speed
differences among individuals. First, while the author recognizes that within subject
variance such as conﬁdence and fatigue can affect test scores, the battery assessing speed
in the current study will consist of measures requiring no more than 10 minutes of the
applicant’s time to reduce long-term within-subj ect effects from having a large
relationship on measures of speed. In addition, the current study is more interested in
investigating between-subject differences in speed rather than those at the within-subj ect
level and as such will focus most of the measurement attention on between-subj ect
factors. Second, between-subject factors, particularly ability to correctly solve problems
and carelessness, can lead to a misinterpretation of measures of speed. As such, care will
be taken to assess cognitive ability and conscientiousness while investigating
relationships with speed. By incorporating and controlling for variance in speed due to
those measures, it is the author’s intention to get a more accurate measure of speed as
well as determining, at least in a linear model, the relationships of speed with other
measures. In addition, other constructs such as test-taking motivation, age, and perceived
face validity of the test which may impact an assessment of speed will be assessed and
controlled for in the current research study. By accounting for more of the determinants
of test responses, I hope to increase the predictive validity of speeded tests (Ackerrnan &

Humphreys, 1990). While both errors and latency will be considered in this research,

21

focus will be maintained on the speed component of measures and its relationships with

other factors.

Relationships with Measures of Speed

Traditional measures of speed have two key components. First, performance is
based on how fast an individual can complete items or tasks of relatively low difﬁculty.
Second, individuals have to persist in their task performance to get a higher score. For
instance, individuals must maintain their pace over ten minutes to complete as many
items as possible or they are timed on how long it takes to complete a series of simple
tasks. Based on these components, measures of speed can be related to a wide range of
factors from cognitive abilities to noncognitive factors. While cognitive ability typically
is identiﬁed as an inﬂuence of speeded test scores, many researchers acknowledge that
speed in test-taking or other situations can be affected by more than an individual’s level
of cognitive ability (e.g., Carlson & Widaman, 1987; Carroll, 1993; Jensen, 1980; Marr &
Stemberg, 1987). Even early research postulated, and on some occasions minimally
displayed, the effect of noncognitive measures on speeded performance (i.e., Himmilweit,
1946; Thurstone, 1937; Wesman, 1960). While generating early interest and postulating
the effects of various constructs, empirical research examining these effects has been
minimal. In this research, the examination of variables related to speed will focus ﬁrst on
cognitive variables, and then on important noncognitive variables to empirically verify

whether or not such relationships exist. While not speciﬁcally going into the hypotheses

22

of the current research proposal, the next section will outline and deﬁne the conceptual
' nature of those constructs that may inﬂuence scores on speeded tests.
Cogpitive Inﬂuences: General Cognitive Ability

One of the most widely studied constructs in psychology is general cognitive
ability, or intelligence, and its effects on life outcomes. Individual differences in
intelligence are related to such things as the ability to understand complex ideas, to learn
from experience, to adapt to the environment, and to engage in reasoning and thought
(Neisser et a1, 1996). In a study examining the Raven’s Progressive Matrices Test
(Raven, 1965), a well known intelligence test, Carpenter and colleagues (1990) found that
most individuals use an incremental strategy for encoding and inducing the regularities in
each problem, but higher scoring individuals were better able to induce abstract relations
and dynamically manage a large set of problem-solving goals. Dating back to the early
part of the century, Spearman (1904) proposed a two factor theory of abilities consisting
of a general cognitive ability (g) and speciﬁc abilities (s). Spearman noticed that scores
on tests measuring mental processes were all positively correlated such that total scores
on different tests typically rank order persons in similar ways. Using factor analysis,
Spearman found that a large proportion of variance among measures of complex
cognitive abilities could be attributed to a general cognitive ability factor. Such ﬁndings
are still typical in current testing situations (Jensen, 1992). This general ability factor has
been given labels such as intelligence, general aptitude, learning potential, attentional
resources, and cognitive abilities. For purposes of this proposal, the term general

cognitive ability will be used. This ability can be broadly deﬁned as the mental process

23

whereby individuals learn from experience and adapt to their environment (Gregory,
1996). The strength of the correlations of g with such outcomes as job performance, skill
acquisition, aptitudes, personal income, health and ﬁtness, etc., has been well-
documented (Ackerman, 1987, 1989; Brand, 1987; Hunter, 1986; Ree 8c Earles, 1992).
In addition to general cognitive ability, a variety of noncognitive variables can also be
related to speed.
Noncogpitive Inﬂuences: Personalig Traits

Questions have oﬁen been raised regarding the validity of noncognitive correlates
of performance (e. g., Hunter & Hunter, 1984). However, personality traits have recently
displayed incremental validity as a predictor in conjunction with ability tests in the
prediction of performance (Baehr & Orban, 1989; Day & Silverrnan, 1989; McHenry,
Hough, Toquam, Hanson, & Ashworth, 1990). Such results have generated new interest
in using more work-related constructs of personality to predict important work-related
outcomes in organizational settings (e.g., Guion, 1991). One possible explanation for this
incremental validity is that personality characteristics might be related to aspects of the
job performance dimension that are not associated with cognitive ability. Based in a
large part by Project A, Campbell (1990) proposed an eight-factor taxonomy of job
performance across jobs: (a) job-speciﬁc task proﬁciency, (b) nonjob-speciﬁc task
proﬁciency, (c) written and oral communication tasks, ((1) demonstrating effort, (e)
maintaining personal discipline, (1) facilitating peer and team performance, (g)
supervision, and (h) management/administration. In examining the criterion taxonomy,

job performance dimensions of demonstrating effort and maintaining personal discipline

24

seem to be most predictable by personality variables (Schneider & Hough, 1995). In
Project A, personal discipline was better predicted with personality variables than
cognitive variables (R = 0.32 versus R = 0.16) while effort was predicted equally well by
both personality and cognitive variables (McHenry, Hough, Toquam, Hanson, &
Ashworth, 1990). Contextual performance including non-role speciﬁc behavior such as
persisting with extra effort when necessary, and following organizational rules and
procedures even when it is personally inconvenient (Bonnan & Motowidlo, 1993) also
seem to lend themselves to prediction by personality variables. Other research by Hough
and colleagues (Hough, 1992; Hough et al, 1990) found personality being more predictive
of commendable behavior and law abiding behavior (contextual performance) than it is of
more typical measures of task and job performance. In particular, measures of potency,
achievement, dependability, adj ustrnent, and agreeableness tended to have the highest
relationship with such criteria. Mumford and colleagues (1993) found that personality
variables (low evaluation apprehension, high self-discipline, and high creative
achievement) distinguished those individuals who scored highly on both ill-deﬁned and
well-deﬁned tasks. Such emphasis on effort and discipline are similar to the sort of
characteristics needed to score high on speeded tasks. Maintaining effort as well as
performing rapidly are essential to high performance on such tasks. Noncognitive factors
that should inﬂuence such criterion dimensions as effort, discipline, and rapid

performance are examined in relationship to speeded test performance.

Impatience/Irritabilitv and Type A Behavior Pattern. The manner in which

 

individuals view time could have a profound effect on their behavior and the speed of that

25

behavior. Originating in the medical research, one of the more popular personality
constructs in psychology is the Type A Behavior Pattern (TABP). Individuals who are
considered Type A tend to be more concerned about time and are characterized as
walking, eating, and talking rapidly, being impatient, striving for achievement and
competitiveness, feeling time pressure, etc.. Recent researchers have found that TABP is
not an unidimensional construct, but reﬂects the constellation of behaviors constituting
the construct (Edwards, Baglioni, & Cooper, 1990; Spence, Helmreich, & Pred, 1987).
Most research on Type A behavior patterns has used self-report scales, especially the
Jenkins Activity Survey (Jenkins, Zyzanski, & Rosenman, 1979) and the Framingham
scales (Haynes, Levine, Scotch, Feinleib, & Kannel, 1978). Edwards and colleagues
(Edwards et al, 1990) examined three popular measures of TABP (the Framingham
questionnaire, the Bortner questionnaire, and the Jenkins Activity Survey) and found that _
all three measures seem to tap different underlying constructs, contain ample
measurement error, and fail to recognize the multidimensional nature of the construct.
Such ﬁndings led Edwards et al (1990) to conclude that global measures of TABP be
abandoned in favor of measures that recognize the multidimensionality of the construct,
particularly those dimensions that are of interest to the researcher’s question. In
examining the Jenkins Activity Survey, Spence and colleagues (Spence et a1, 1987) found
that the measure revealed two independent factors: Achievement Strivings (AS) and
Impatience-Irritability (II). In addition, they found that each factor had differential
effects on performance and health with AS correlating with grade point average and II

correlating with physical complaints.

26

A promising construct related to the TABP is the notion of time urgency which
has been deﬁned as individuals with an accelerated pace or individuals who consider time
as a scarce resource and plan its use thoughtfully (Bumam, Peenebaker, & Glass, 1975;
Landy, Rastegary, Thayer, & Colvin, 1991). Research has begun to investigate the
multidimensional nature of time urgency. Edwards and colleagues (1990) identiﬁed
seven dimensions of time urgency: (a) general speed, (b) doing many things at once, (c)
eating fast, ((1) putting words in the mouths of others, (e) impatience, (f) punctuality, and
(g) time pressure. Research by Landy and colleagues (Conte, Landy, & Mathieu, 1995;
Landy et a1, 1991) have further reﬁned the measurement of time urgency using
behaviorally anchored rating scales (BARS). Landy and colleagues (1991) found that
time urgency is a multidimensional construct that includes (a) time awareness, (b) eating
behavior, (c) scheduling, (d) nervous energy, (e) list making, (f) speech patterns, and (g)
deadline control using a multitrait-multirater approach with a sample of ofﬁce and
professional workers. In addition, measures of time urgency also displayed low
intercorrelations with the State-Trait Anxiety Scale (Spielberger, Jacobs, Crane, Russell,
Westberry, Barker, Johnson, Knoght, & Marks, 1979) and the Job Descriptive Index
(Smith, Kendall, & Hulin, 1969) suggesting that the time urgency dimensions were not
surrogates for constructs such as job satisfaction or anxiety. Conte et a1 (1995), using
conﬁrmatory factor analysis, found good ﬁt indices for a ﬁve dimensional model of time
urgency: (a) time awareness, (b) scheduling, (c) list making, (d) eating behavior, and (e)
deadline control. In their study, time awareness, scheduling, list making, and deadline

control were signiﬁcantly associated with Spence et al.’s (1987) AS. Speech patterns,

27

eating behavior, and nervous energy were signiﬁcantly associated with 11. Such
relationships would perhaps demonstrate a higher order two-factor solution of AS and II
as stipulated by Spence and colleagues (1987). While the II factor is well-deﬁned in the
above paragraphs, the notion of achievement strivings is outlined below in reference to
notions of conscientiousness.

Corgientioujsness. Recent personality literature has focused on the emergence of
a ﬁve-factor model of all personality types (Digman, 1990; Goldberg, 1993). While
debates emerge concerning whether or not the ﬁve factor model encompasses all
personality traits or whether we should focus on personality types instead of traits (sec
Schneider & Hough, 1995, for a review), one of the most robust personality traits linked
with important work-related outcomes is conscientiousness. Conscientiousness is
associated with such descriptive terms as responsible, dependable, planful, organized,
persistent, and achievement-oriented (Barrick, Mount, & Strauss, 1993).

In a series of studies, Barrick and Mount have demonstrated the consistency of the
relationship of conscientiousness to work outcomes. In a meta-analysis on the
relationship of the Big Five and performance criteria across occupational groups. Barrick
and Mount (1991) found that conscientiousness is a valid criteria of job proﬁciency,
training proﬁciency, and personnel data (true r = 0.22). In other studies, they examined
the relationship between the Big Five and job performance in a sample of 146 ﬁrst-line
supervisors and mid-level managers and found that conscientiousness correlated
signiﬁcantly with supervisor ratings of job performance (r = 0.25, corrected r = 0.35).

Finally, in a third study, Barrick, Mount, & Strauss (1993) tested a‘ structural model

28

relating conscientiousness to job performance in a sample of 91 sales representatives in
an appliance manufacturing organization and found that autonomous goal setting and
goal commitment mediated the relationships between conscientiousness and sales volume
and supervisory ratings. In a different meta-analysis examining personality and job
performance outcomes, Tett and colleagues (1991) found that conscientiousness had an
estimated true r of 0.18, slightly less than the Barrick and Mount’s (1991) meta-analysis.

In a separate line of research, Hough and colleagues (Hough 1989, 1992; Hough,
Eaton, Dunnette, Karnp, & McCloy, 1990), using a nine-factor model of personality
related “conscientiousness-type” personality measures of achievement and dependability
to a variety of job and life criteria measures. Measures of achievement were signiﬁcantly
related to criterion measures of effort and leadership, personal discipline, and physical
ﬁtness and military bearing (Hough et al, 1990) and with job performance constructs of
job proﬁciency, training success, educational success, commendable behavior, and law
abiding behavior (Hough, 1992). Measures of dependability were signiﬁcantly related to
criterion measures of effort and leadership, personal discipline, and physical ﬁtness and
military bearing with mixed results for technical proﬁciency and general soldiering
(Hough et a1, 1990). In addition, dependability was correlated with job proﬁciency,
training success, educational success, commendable behavior, and law abiding behavior
(Hough, 1992). In similar work investigating models of supervisory job performance
ratings, Borrnan and colleagues (1991) found that achievement orientation and

dependability had small direct effects on ratings, but larger indirect effects in the

29

prediction of supervisory ratings through its effects on awards and disciplinary actions,
respectively.

Whereas some researchers investigate conscientiousness as a single variable (i.e.,
Barrick & Mount, 1991), other researchers consider such conscientiousness sub-
dimensions as achievement and dependability as important factors to be considered
instead of a general measure of conscientiousness (Borman et a1, 1991; Hough, 1992).
Achievement is similar to the AS factor in measuring the Type A behavior pattern and is
deﬁned as the tendency to strive for competence in one’s work such that a person works
hard, sets high standards, tries to do a good job, persists in the completion of a task, etc.
(Hough, 1992). Dependability refers to such personal characteristics as being disciplined,
well~organized, planful, honest, trustworthy, etc. (Hough, 1992). Acknowledging that all
of these constructs are highly related, this research will isolate achievement and compare
it with a general measure of conscientiousness to investigate potential differential
relationships with speed. For example, while both constructs entail the notion of
persistence, a person high in achievement may strive to work faster realizing that success
is based on time. Likewise, achievement and conscientiousness may have differential
relationship with the construct of impatience/irritability. Achievement may have a higher
relationship with impatience irritability, thereby decreasing the amount of incremental
validity added to the prediction of speed.

In addition to the personality characteristics associated with TABP, other

personality-related characteristics could inﬂuence speeded test scores. In particular, an

30

individual's temporal orientation toward the completion of tasks, and their level of self-
monitoring in relationship to situational inﬂuences could inﬂuence speeded performance.

Temptmal orientption. In addition to the above personality traits, individuals’
attitudes, values, and beliefs about time can differ (Cottle, 1976) without incorporating
the impatience/irritability aspects associated with type A behavior pattern. Such
differences can inﬂuence the rate at which individuals perform tasks in a typical day and
subsequently the rate at which a speeded task is performed. Two temporal orientations
that can affect an individual’s performance in speeded tasks are polychronic versus
monochronic orientation and pace.

Work by Bluedorn and associates based on Hall’s (1983) original work has
examined monochronic versus polychronic orientation as a continuous variable that
inﬂuences the way individuals approach life’s work (Bluedorn & Denhardt, 1988;
Bluedorn, Kaufman, & Lane, 1992). People who have a monochronic orientation
approach tasks one at a time whereas individuals with a polychronic orientation are
simultaneously involved in two or more activities. For instance, if a person had three
activities to complete (1, 2, 3) and they had to complete 1 before they started 2, and had
to complete 2 before they started 3, they would be displaying a monochronic orientation.
However, if the individual does 1 for a while and then starts 2, and then goes back to 1,
etc., such that they always make progress on each task, they are displaying a polychronic
orientation. Characteristics of individuals with a monochronic orientation include being

task-oriented, emphasizing promptness, and sticking to plans whereas characteristics of

31

individuals with a polychronic orientation involve changing plans and emphasizing
relationships over tasks (Hall, 1983).

The concept of pace or tempo has been deﬁned in many ways including notions of
rate (Lauer, 1981), speed (Amato, 1983), and velocity (Kelly, 1988; Warner, 1988).
Operationally, measures of pace of life and pace of work have included walking speed
(Amato, 1983; Bomstein, 1979), speed of post ofﬁce transactions (Levine & Bartlett,
1984), speed of currency exchange (Amato, 1983), and speed of lunch consumption
(Lucia, 1985). Pace of life differences have been noted across individuals, type of
environment (urban versus rural), and nationalities (Levine, 198 8). The concept of pace
or “personal tempo” has been proposed to be a dimension of speed in timed performance
tasks (i.e., Marr & Stemberg, 1987). While research has supported the existence of
reliable individual differences in the rate of performing tasks, there is no clear empirical

evidence that such differences are associated with personal tempo (Carroll, 1993). The
hypothesis that such “in-grained” tendencies in the performance of activities could
inﬂuence the outcomes in measures of speed has never been adequately investigated in an
empirical fashion.

Self-monitoring. In a recent review of personality characteristics and work
outcomes, a possible explanation provided for lower-than-expected validities for
personality characteristics was the presence of moderators between the relationship of
personality characteristics and work outcomes (Schneider & Hough, 1995). One possible
moderator that could display interactive effects is self-monitoring. Self-monitoring

represents the extent to which a person in a social situation pays attention to social cues

32

and actively attempts to construct a pattern of behavior that is appropriate to that
particular context (Snyder, 1979). Sources of information such as social norms,
situational and interpersonal speciﬁcations of appropriateness on how to behave, social
attitudes, and personal dispositions can be used as guides to the construction of
behavioral patterns (Snyder & Ickes, 1985). However, individuals differ in the extent to
which they rely on social cues to regulate their behavior. Individuals high in self-
monitoring tend to monitor and regulate their behavior based on the behavioral
requirements of situations, whereas individuals low in self-monitoring permit their
behavior to be guided by their dispositional characteristics, thus ignoring the demands of
the situation. Relationships between dispositions and behavior of individuals high in
self-monitoring could be lower than for low self-monitoring individuals because the high
self-monitoring individuals would pay more attention to situational and interpersonal
cues of social appropriateness. The exception for high self-monitors would be in cases
where it is socially desirable to act on one’s true attitudes and dispositions (Snyder &
Kendzierski, 1982). In contrast, individuals who are low in self-monitoring would have a
more consistent and larger correspondence between dispositions and behavior such that
their dispositions guide their behavior rather than being guided by information in the
social situation (Snyder & Ickes, 1985). As such, it would be expected that measures of
typical and maximum performance (Sackett et al, 1988) would be similar for individuals
at a lower level of self-monitoring.

Empirically, self-monitoring is typically described as a personal moderating

variable between dispositions and behaviors. Based on original research by Snyder

33

(1974), a 25-item scale was developed that demonstrated both convergent and divergent
validity as well as internal consistency. Items endorsed by high self-monitors included
such things as “I may deceive people by being friendly when I really dislike them” or “I
would probably make a good actor,” while items endorsed by low self-monitors included
such things as “I can only argue for ideas which I already believe” and “I ﬁnd it hard to
imitate the behavior of other people” (Snyder, 1974). In later research based on the
original scales, the Self-Monitoring Scale was reduced to 18 items, removing items from
the original scale that had low item-total correlations (Gangestad & Snyder, 1985).

There have been some questions regarding what the self-monitoring scale actually
measures. One major concern is that being a true-false scale, self-monitoring has lower
inter-item correlations than desirable, but its length (18 or 25 items) raises the alpha level
to acceptable standards for unidimensionality (Briggs & Cheek, 1988). Such claims of
multidimensionality argue that the subscales should be treated separately (Briggs &
Cheek, 1980; Lennox & Wolfe, 1984). In a systematic examination of the Self-
Monitoring scale, Briggs and colleagues have identiﬁed two factors that seem to emerge
from use of the scale: (1) social surgency dealing with exhibitionism, social potency, and
extraversion, and (2) other-directedness dealing with shyness and lack of self-esteem
(Briggs, Cheek, & Buss, 1980; Briggs & Cheek, 1986, 1988). In response to the
criticisms, Snyder and Gangestad (1986) argue that the self-monitoring scale has
demonstrated important relationships with many relevant criteria and even though
interpretable factors result from rotation, most of the items also load on the ﬁrst unrotated

factor which reﬂects a single latent variable. While a resolution to the debate has never

34

transpired, the utility of the scale as a predictor of interpersonal behavior has been
supported by most researchers.

In addition to general cognitive ability and personality characteristics, situation-
related factors can also potentially inﬂuence scores on speeded tests. Time of day
inﬂuences and anxiety toward testing situations could affect an individual's level of
awareness as well as the amount of effort being placed to the task at hand. Potential
situational inﬂuences will be deﬁned in the following sections.

Situational arousal. Another category of noncognitive factors that could inﬂuence
speeded test performance involves measures of arousal induced through situational
demands. Arousal can be characterized by an individual’s level of alertness, vigor, and
activation ranging from extreme excitement on one end to extreme drowsiness at the
other end (Humphreys & Revelle, 1984). Indices for arousal can range ﬁom self-report
measures to physiological measures such as heart rate, breathing rate, and skin
conductance. In addition, whereas the relationship between arousal and performance has
been curvilinear, a monotonic linear relationship between arousal and performance can
occur when difﬁculty of the task is low (Humphreys & Revelle, 1984; McGrath, 1976)
While personality traits such as impulsivity/irritability represent a more general “trait-
like” notion of arousal, arousal induced through situational circumstances could add
additional variance to the prediction of speeded test scores. For purposes of this research,
two factors, circadian rhythms and comparative anxiety are examined.

It has been well-established that there are signiﬁcant time-of-day effects in many

types of human performance (Blake, 1967; Colquhoun, 1971; Kerkhof, 1985; Revelle,

35

1989). These effects have been attributed to variations in circadian arousal levels
affecting the capacity and efﬁciency of working memory as well as more general notions
of fatigue or alertness (F lokard, Wever, & Wildgruber, 1983). Some studies suggest that
there are inter-individual differences in the time of day at which one reaches his/her peak
or acrophase such that there are two clusters of individuals who reach their peak either
early or later in the day (Home & Ostberg, 1977). According to a review of the circadian
rhythm literature, “morningness-eveningness” tends to be the most powerful
interindividual difference variable related to biological, attitudinal, and behavioral
differences on peak arousal (Kerkhof, 1985; Smith, Reilly, & Midkiff, 1989). Several
measures have been developed to assess morningness and eveningness (i.e., Folkard,
Monk, & Lobban, 1979; Home & Ostberg, 1976; Torsvall & Akerstedt, 1980). In a
recent review, Smith and colleagues (1989) re-evaluated the psychometric characteristics
of various scales and found inter-item deﬁciencies in all of them. In response, they
developed a composite scale that combined the best characteristics of the scales and
found high internal consistency as well as relationships to external criteria associated with
morningness and eveningness.

In a personnel selection context, a common concern between concurrent and
predictive strategies is that current employees may not be equal to applicants in terms of
motivational issues (Arvey, Strickland, Drauden, & Martin, 1990). Motivational
differences between samples used in concurrent and predictive designs may act to distort
and inﬂuence the estimates of validity or introduce random error (Barrett, Phillips, &

Alexander, 1981; Cranny & Smith, 1982). Another major difference is that individuals

36

may differ in terms of their anxiety toward tests and these differences might inﬂuence
their test scores. Tests takers may experience profound emotions and feelings as a result
of the situational pressure of taking tests which affect their future (Hashemian, 1978;
Nevo & Svez, 1985). Such anxiety would interfere with their ability to concentrate on
the task at hand (Kanfer & Ackerrnan, 1989), potentially inﬂuencing their ability to
complete speeded tasks where time pressure is evident.

Therefore, inﬂuences on speeded test performance can come from a variety of
sources: levels of general cognitive ability/efﬁciency, personality dispositions, and
situational factors. Other characteristics can inﬂuence speeded test scores as well as these
dimensions. However, these inﬂuences are not a focus in the present research due to
restriction of range problems, namely individuals being studied will be similar on these
characteristics. These characteristics include the age of the examinee (most participants
are young) and various applicant reactions such as their test-taking motivation and

interest in the job, and the perceived face validity of the tests.

Other Inﬂuences

Agg. One of the more apparent antecedents of an individual’s speed of behavior
is an individual’s age. When older and younger adults perform similar cognitive tasks,
older individuals tend to perform more slowly than the younger adults. Age declines
have been found in a wide variety of experiments (Birren & Fisher, 1995; Salthouse,
1985). Some research has attributed such slowing to a general factor of speed of the

central nervous system (Birren, 1965; Birren, Woods, & Williams, 1980). In preliminary

37

research, Brinley (1965) compared a group of older (mean age = 71.4, SD = 6.1; n = 51)
and younger (mean age = 24, SD = 5.2; n = 60) adults on 21 tasks. He performed a
regression of mean time scores for the elderly on the mean time scores for the young and
found that age accounted for 98% of the variance in adult response time (old =
l.68(young) - .27; r = .99), supporting a generalized slowing across tasks. In other meta-
analytic work that integrated data from studies using a wide range of information
processing tasks, research has also found evidence supporting a general slowing
hypothesis (Cerella, 1985; Cerella, Poon, & Williams, 1980; Hale, Myerson, & Wagstaff,
1987). Such a generalized slowing throughout the central nervous system is purported to
be manifested in any task that requires higher levels of information processing.

An area in which age-related decrements have not been consistently demonstrated
has been when the task differs in terms of the speed and knowledge required for its
completion. If performance is mainly a function of speed, then age-related effects should
be large, but if knowledge is an important aspect of the task, then age effects can be
expected to be smaller (Salthouse, 1993; Schaie & Willis, 1993). Such a distinction is
similar to that of Horn-Cattell ﬂuid-crystallized intelligence distinction. Such outcomes
were found in areas such as lexical vs. nonlexical tasks (Lima, Hale, & Myerson, 1991),
and arithmetic tasks (Geary, French, & Wiley, 1993).

Applicant regtions pnd motivation. Another factor that can inﬂuence test scores
is an applicant’s reaction to the test (Arvey et al, 1990; Rynes, 1993; Smither, Reilly,
Millsap, Pearhnan, & Stoffey, 1993). Whereas applicants may have differential reactions

to tests in terms of anxiety, most candidates should have high levels of motivation to do

38

well on the task at hand such that restriction of range in predictive validity settings could
become a major issue (Arvey et al, 1990). In a recent study, Chan and colleagues (Chan,
Schmitt, DeShon, Clause, & Delbridge, 1997) found that test-taking motivation assessed
after the completion of a ﬁrst test affected subsequent performance on a parallel test after
race and ﬁrst test performance were controlled. Face validity assessed after the
administration of ﬁrst test also affected performance on a subsequent equivalent test, but
only indirectly through test-taking motivation. However, the study examined test-taking
motivation in a student sample, so restriction of range in terms of test-taking motivation
was less of a concern than with an applicant sample where motivation to do well should
be uniformly high. Such differences in motivation have been found in similar samples
examining incumbents (Mean = 4.97, SD = 1.71) versus applicants (Mean = 6.08, SD =
.71) on test-taking motivation (Arvey et al, 1990). A problem with test-taking motivation
research is the difﬁculty in establishing causality (Arvey et al, 1990; Chan et a1, 1996). It
is still unclear as to whether an individual’s level of motivation inﬂuences test scores or
whether one’s performance inﬂuences their attribution of motivation.

A similar motivational inﬂuence on test scores in an applicant setting is the
candidate’s actual interest in obtaining a position with an organization. Some candidates
may be merely applying for a position because their parents, spouse, etc., are pressuring
them to apply even if they have little interest in obtaining employment. Other candidates
may be applying because they need to maintain government beneﬁts and have to apply to
so many positions each week/month. If the person is not really interested in obtaining

employment with the organization, his/her level of effort may be reduced compared to

39

someone who really wants the position. While this scenario is most likely a
characterization of only a minority of actual job candidates, such an inﬂuence should be
controlled in examining other factors.

In addition to motivation and interest, another potential inﬂuence on test scores is
the candidate’s perception of the face validity of the test. Face validity is deﬁned as the
extent to which examinees perceive the content of the selection procedure to be related to
the content of the job (Smither et al, 1993). Individuals who perceive that the tests are
unrelated to the job may react negatively and reduce the amount of effort expended, thus
inﬂuencing test scores. In carefully created selection tests, the inﬂuence of negative face
validity reactions should be minimized.

Overall, the variability on the above factors should be minimal with most
candidates for entry level positions being of similar age with similar perceptions and
motivations. However, these factors may inﬂuence test scores for a minority of actual job
candidates. Therefore, age, motivation, and perceived face validity are measured and
controlled in the research study.

After examining the relationships of factors to measures of speed, the research
will switch focus to examine the second objective of the research investigating the
inﬂuence of individual differences in measures of speed on broader measures of j ob
performance. To adequately examine the relationship, attention will ﬁrst be placed on the
nature of performance using job performance as a basis. In examining the performance
domain, consideration will be placed on those dimensions of performance that should be

most highly related to individual differences in measures of speed. Finally, assessment

40

centers, a method by which performance data will be collected in this study, will be
highlighted in relationship to typical administration procedures and types of performance

dimensions assessed.

Job Performance

Until now, the focus of the present research has been on investigating
relationships with measures of speed. However, a second interest of the research is also
to examine how measures of speed relate to broader measures of job performance, in
addition to the relationship between cognitive ability and performance. To examine the
criterion domain taking a construct-oriented approach, care must be taken to specify the
multidimensional nature of the domain as well as specify the dimensions that should be
most related to notions of speeded test performance compared to those dimensions most
related to cognitive ability variables (Schneider & Hough, 1995).

Researchers have proposed various taxonomies of job performance. Campbell
(1990), based on the work in Project A, suggested that the highest-level job performance
dimensions include: (a) job-speciﬁc task proﬁciency, (b) non-j ob-speciﬁc task
proﬁciency, (c) written and oral communication tasks, ((1) demonstrating effort, (e)
maintaining personal discipline, (0 supervision, and (g) management/administration.
Borman and Brush (1993) denoted 18 managerial factors that could be summarized in
four superfactors: (a) interpersonal dealings and communication, (b) leadership and
supervision, (c) technical activities and the “mechanics of management,” and (d) useful

personal behaviors and skills. In a different twist, Borman and Motowidlo (1993)

41

distinguish task performance from contextual performance. Contextual performance
serves more of an environmental support role including such activities as volunteering,
persisting with extra enthusiasm to complete a task, helping and cooperating, and
supporting organizational objectives (Borman & Motowidlo, 1993).

Specifying the criterion space provides assistance in highlighting the type of
criterion dimensions that are most readily predicted by personality and noncognitive
variables (Schneider & Hough, 1995). Likewise, certain dimensions from the taxonomies
listed above should have higher relationships with measures of speed than others. In
terms of Campbell’s taxonomy, criteria that emphasize persisting on tasks over time, such
as the demonstration of effort and the maintenance of personal discipline, should be more
related to measures of speed as exhibited in typical speeded tests (Borman et al, 1991;
Hough et al, 1990; Hough, 1992; McHenry et al, 1990). In addition, non-job-speciﬁc
task proﬁciency should also be related to measures of speed if speed of performance is
necessary for success across multiple tasks that are not speciﬁc to one’s job.

A measure of job performance can be obtained in many ways such as objective
production measures, subjective ratings, etc.. One way in which such multidimensional
measures of performance could be collected is through assessment center activities.
Assessment centers are one of the more popular methods for selecting, developing,
evaluating, or promoting individuals in work organizations (Gaugler, Rosenthal,
Thornton, & Bentson, 1987). Typically, such devices require individuals to perform
some group task where they are subsequently rated by assessors. Some of the deﬁning

characteristics of such centers include a high level of face validity (Schmidt, Greenthal,

42

Hunter, Bemer, & Seaton, 1977; Macan, Avedon, Paese, & Smith,l994), the use of
multiple assessors to observe multiple assessees’ behavior in exercises, and the rating of
assessees on multiple dimensions including such skills as problem-solving, planning,
organizing, delegation, or written communication (Schneider, 1992; Task Force on
Assessment Center Standards, 1989). The typical procedure of an exercise is twofold. I
First, assessors observe assessees’ behavior in either a live, videotaped, or written format.
Second, assessors rate assessees on dimensions with the assistance of standardized
scoring guidelines. Exercises are typically constructed based upon job analysis to mirror
tasks that are performed on the job or to mirror critical aspects of the task domain
(Schneider, 1992). In fact, such exercises are often considered “work samples” of
activities that are displayed on the job that compare favorably with supervisor ratings and
productivity measures of job performance (see Smith, 1991 for a review). In these group
exercises, assessees perform in small groups, typically made up of other assessees, to
solve organizational problems, generate plans, and manufacture mock products
(Schneider, 1992). Using behaviorally anchored rating scales, assessors make ratings for
each target person. Factor analysis studies of assessment center ﬁnal ratings have
commonly demonstrated a three factor solution for the constructs measured in assessment
center activities: problem solving, interpersonal skills, and initiative/work orientation
(Schmitt, 1977). Such dimensions should demonstrate differential relationships to
measures of speed. Given the prior research highlighting the types of performance
dimensions measures of speed should be most related to, it seems reasonable to assume

that dimensions involving effort and discipline such as work orientation should have

43

larger relationships with measures of speed than dimensions such as interpersonal and
problem-solving skills.

Given a work sample/assessment center context to measure job performance,
notions of the continuum of typical versus maximum performance (Sackett et al, 198 8)
need to be taken into account. In simple terms, maximum performance is typically
associated with one’s best effort on the job, whereas typical performance is more
associated with behaviors that would most likely be exhibited on the job over extended
periods of time. In terms of variables related to speed, general cognitive ability and
reaction time measures would be most illustrative of a maximum characteristic given that
most individuals in testing situations give their best efforts to try to solve problems or
complete tasks in the time allotted, particularly in situations involving short time periods.
Noncognitive variables such as temporal orientation and achievement would be more
illustrative of a typical characteristic because you would expect an individual to display
those traits more consistently over time. When dealing with measures of job
performance, a construct can be measured across the continuum between typical and
maximum performance. Maximum performance has three main characteristics (Sacket et
al, 1988): (a) there must be explicit awareness that one is being evaluated, (b) there must
be an awareness of and acceptance of instructions to maximize effort, and (c)
performance must be measured over a short enough time duration that the perforrner’s
attention remains focused on the accepted goal of maximum performance. In assessment
center/work sample situations, the precise point on the continuum from typical to

maximum is unresolved, primarily due to issues of time. It can be argued that task length

and task complexity can reduce candidate awareness of the evaluative aspect of the
situation and produce a sample of typical job performance. However, it can also be
argued that the situation heightens the candidate’s level of effort and attention to be
considered maximum performance. One objective of the current research is to investigate
the incremental validity of measures of speed. Therefore, speed, which falls more on
maximum side of the continuum, will be evaluated both against more maximum elements
of job performance, where conceptually it should have a larger relationship, and more
typical elements.

In summary, job performance dimensions involving a demonstration of effort, a
maintenance of personal discipline, and nonjob-speciﬁc task proﬁciency across multiple
tasks not speciﬁc to one’s job should be most related to measures of speed. Primarily, this
relationship should provide incremental validity in the prediction of job performance over
that of general cognitive ability. Also, the nature of typical versus maximum
performance in relationship to speed as an incremental predictor over general cognitive

ability will be examined.

Research Study: Relationships with Measures of Speed
The following research study investigates the relationships of factors with
measures of speed as well as its usefulness in predicting job performance. Using the
literature cited above, the proposal outlines the hypotheses of the research in the
following fashion. First, relationships with measures of speed are examined by

highlighting general cognitive ability and noncognitive factors as important factors that

45

can inﬂuence an individual’s level of arousal and effort and subsequently their
performance on measures of speed. Second, the incremental validity of measures of
speed in the prediction of job performance, above and beyond the inﬂuence of cognitive
ability, is examined.

General Cogpjtive Abilig and Speed

Many recent studies have demonstrated that the speed with which simple
perceptual and cognitive tasks are performed is correlated with psychometric intelligence
(i.e., Ceci, 1990; Vernon, 1987). Research has demonstrated that individuals with higher
scores on general cognitive ability tests tend to apprehend, scan, retrieve, and respond to
stimuli faster than those with low scores on general cognitive ability tests (Neisser et al,
1996). The more complex the stimuli and the judgments required, the larger the
relationship of speed with psychometric intelligence (Jensen, 1993).

The type of speeded task used to investigate this relationship has been varied from
the typical paper-and-pencil test paradigm to more complex biological measures.
Original work required participants to make same/different judgments regarding visual
displays and found that the speed of response was correlated with psychometric verbal
ability (Hunt, 1978; Jackson & McClelland, 1979). Other cognitive paradigms used to
study speed include choice reaction time, inspection time, and neurological measures. In
choice reaction time, the individual participant must move his/her ﬁnger from a “home”
button to one of eight other buttons arranged in a semicircle around it. One of the lights
is illuminated that indicates it as a target and the individual moves his/her ﬁnger and

presses the button corresponding to the target (Jensen, 1987). Typically, reaction time

46

correlates as high as -.4 to -.5 with intelligence test scores with more complex decision
paradigms (e.g., odd-man-out, higher number of lights, etc.) exhibiting higher
correlations (Jensen, 1993, Neisser et al, 1996). In this case, the negative correlation
indicates the higher the intelligence, the less time it takes to react to a target stimulus.
Another paradigm for measuring processing speed is inspection time. In the
standard version of the paradigm, two vertical lines are displayed very brieﬂy on each
trial, followed by a pattern mask. The subject is required to judge which of the two lines
is shorter. For any given individual, inspection time is deﬁned as the minimum exposure
duration for which the lines must be displayed in order for the individual to meet a given
level of accuracy, for example 71% or 85% (Deary & Stough, 1996; Nettelbeck, 1987).
As opposed to choice reaction time, it is stimulus duration required by the participant to
reach a given level of accuracy that is assessed, not the speed of a participant’s response
in making the discrimination, so inspection time is often referred to as the speed of intake
of information (Deary & Stough, 1996). Like choice reaction time, inspection time is
consistently correlated with psychometric intelligence. In a recent meta-analysis,
Kranzler and Jensen (1989) found an overall correlation of -.30 between IQ and
inspection time that rose to -.55 (longer times are associated with lower IQ scores) with
measurement corrections and omissions of studies with serious methodological problems.
Such measures of inspection time appear to be stronger with measures of performance or
ﬂuid intelligence (Deary & Stough, 1996) and have been replicated in post-meta-analysis

studies (Bates & Eysenck, 1993, Deary, 1993).

47

A ﬁnal category of speed measures are ones that examine more direct indices of
neural processing including averaged evoked potentials (AEP), nerve conduction velocity
(NCV), and cereme glucose metabolism (CGM) which are usually thought to reﬂect and
neural efﬁciency. Such research is relatively new in application. Vernon (1991) found
that AEP, measuring length and amplitude of wavelengths, as well as other EEG
measures was correlated with intelligence test scores for both adults and children. CGM
measured by positron emission tomography (PET) scans, provide an index of the activity
of the brain either at rest or when subjects are involved with some task (Vernon, 1991).
Some preliminary research has found that cortical absolute CGM rates were correlated
between -.50 and -.54 with better performance on mental ability tests associated with less
cortical activity (Parks et al, 1988). Higher scoring subjects had to expend less energy
than lower scoring subjects. NCV is essentially a measure of the speed with which
electrical impulses are transmitted by the nervous system. In one study, NCVs correlated
r = .26 (corrected for restriction of range r = .37) with scores on an unspeeded test of
intelligence (Reed & Jensen, 1992). Overall, studies of biological measures of neural
efﬁciency were related to psychometric intelligence.

Such research demonstrates a relationship between general cognitive ability and
information processing speed in that individuals would be more efﬁcient in processing
information or have higher levels of arousal to attend to stimuli. While a relationship
should be evident in this study as well, the size of the relationship could be reduced
because speeded tests require the maintenance of speed over a certain time period instead

of the maximum information processing speed paradigms (e.g., reaction time) that are

48

assessed in the above research. Testing conditions in the current research should
introduce a range of noncognitive inﬂuences that are not present in the above information
processing speed research.

Hypothesis 1: Individuals with higher general cognitive ability scores should obtain
higher scores on speeded tests.

Cognitive ability should be related to speeded performance through an increased
efﬁciency or to apprehend, scan, retrieve, and respond to stimuli. As stated earlier,
noncognitive factors should also have an inﬂuence on speeded test scores. Before
examining the speciﬁc noncognitive constructs affecting speed, the proposal will
highlight how noncognitive variables inﬂuence speed of response on tests by examining
two speciﬁc mechanisms: the amount of effort devoted to on-task activities and the
amount of arousal or alertness to stimuli.

Mechanisms of Noncogpitive Inﬂuence

While the role of cognitive abilities in predicting learning and job performance is
well documented (Ackerman, 1987; Hunter & Hunter, 1984), the role of noncognitive
inﬂuences are relatively unexplored in relationship to measures of speed. In speeded
tests, performance is determined by the amount of work that is completed in a set period
of time or the amount of time required to complete a given task. In such scenarios, the
difﬁculty of the items and/or tasks are low, thereby reducing the level of ability needed to
correctly solve the problems. Such tasks should decrease the inﬂuence of cognitive
ability on test scores by reducing the problem difﬁculty and relying on speed of

performance.

49

Noncognitive factors can inﬂuence speeded test performance in a variety of ways.
Using an information-processing framework, noncognitive factors can inﬂuence measures
of speed through a resource allocation process (N aylor, Pritchard, & Ilgen, 1980; Kanfer
& Ackerrnan, 1989). Individuals differ in the way they allocate their time and attentional
effort to the achievement of a particular goal and the extent to which that attentional
effort is maintained over time. Certain tasks require more attentional resources to be
directed toward the task versus attentional resources directed to off-task activities.
Another distinction is the inﬂuence of both distal and proximal processes in task
performance (Kanfer & Ackerrnan, 1989). Distal motivational processes concern the
choice to engage some part of one’s resources to the attainment of a goal whereas
proximal motivational processes determine the distribution of effort across on-task, off-
task, and self-regulatory activities during task activity (Kanfer, 1992). In measures of
speed, the difﬁculty of the task in terms of the ability to solve problems is low, but the
nature of the testing situation forces individuals to remain on-task over time to complete
as many problems as they can or complete the activity as quickly as possible.

In another model concerning the inﬂuence of information-processing and
personality, Revelle and colleagues highlight the organization of personality and
performance in a cognitive information-processing framework (Humphreys & Revelle,
1984; Revelle, 1989; Zinbarg & Revelle, 1989). Their model proposed a theory linking
individual differences in Introversion-Extroversion, Achievement Motivation, and
Anxiety to information processing components of complex task performance (Humphreys

& Revelle, 1984). In their model, personality, in combination with situational moderators

50

(e. g., time of day, fatigue, etc.), affect an individual’s arousal and effort which in turn
affect three components of performance: sustained information transfer, long-term
memory, and short-term memory. In their model, on-task effort and arousal both increase
the resources allocated for sustained information tasks or tasks whereby individuals are
required to process a stimulus, associate an arbitrary response to the stimulus, and
execute the response over an appreciable length of time (Humphreys & Revelle, 1984).
On such tasks and similar tasks in prior research, the relationship between personality
traits and performance should be enhanced because persistence is an important
component for success (Helmreich, Sawin, & Carsrud, 1986; Schneider & Hough, 1995;
Weiss & Adler, 1984).

Overall, noncognitive factors seem to inﬂuence performance on measures of
speed through their impact on an individual’s level of arousal and the amount of effort
they put into the task at hand (Humphreys & Revelle, 1984). Arousal is the state of an
individual that indicates how alert or active they are ranging from extreme drowsiness on
one hand to extreme excitement on the other hand. Individuals high on arousal should be
more alert to stimuli and respond more rapidly. Effort is a more general means of trying
hard or being involved in a task. Individuals who try hard for various reasons (e.g.,
incentives, importance of task, etc.) should maintain focus on a task until its completion.
Considering the effects of effort and arousal, such processes will be more explicitly
examined when we discuss each of the individual noncognitive factors.

I_r_npptience/Init§bilitv (II). Individuals who are high on levels of 11 could be

characterized as having more nervous energy, doing things rapidly, feeling time pressure,

51

etc. (Conte et al, 1995; Edwards et al, 1990; Landy et al, 1991). II has been shown to be
arousal-related such that individuals who are high on 11 display higher levels of alertness
or arousal (Humphreys & Revelle, 1984). Such high levels of activity could characterize
behaviors that are considered speeded. Particularly, in a speeded test scenario, an
applicant’s score depends on how quickly he or she can complete the task as well as how
long he or she can maintain such speed. Individuals whose activity level is high based on
levels of II should maintain a fast tempo throughout the testing condition and should
strive to ﬁnish the task as quickly as possible.
Hypothesis 2: Individuals with higher levels of Impatience/Irritability should obtain
higher scores on measures of speed

Conscientiousness. The positive relationship of conscientiousness and work
performance has been demonstrated in numerous situations (Barrick & Mount, 1991;
Barrick et al, 1993; Hough et al, 1990; Hough, 1992). In Project A, a large scale military
selection project, Hough and colleagues found that measures of conscientiousness,
achievement and dependability dimensions, were strongly related to performance
measures that had effort and discipline components (Hough et al, 1990; Hough, 1992).
Other work has demonstrated that individual differences in dispositions such as
achievement motivation may inﬂuence variability in job performance where persistence is
an important component of success (Helmreich et al, 1986; Weiss & Adler, 1984). These
performance dimensions appear to be most related to successful performance in measures

of speed where an applicant must maintain speed over a period of time, thus requiring

52

effort as well as discipline over time. It is primarily through an increase in effort that
conscientiousness will be related to measures of speed.

Hypothesis 3: Individuals with higher levels of Conscientiousness should obtain higher
scores on measures of speed.

While conscientiousness, through effort and discipline over time, should have a
relationship with speed, a potential conﬂict could be an individual’s propensity to stay on
task and to make sure assignments get completed with 100% accuracy. One could argue
that such carefulness could slow down a highly conscientious individual in speeded tasks.
However, a competing goal of an individual who is high in conscientiousness would be to
complete as many items as possible or ﬁnish in the smallest amount of time to be
successﬁil. The question asks which is more important: being careful in test completion
or ﬁnishing as many items as possible to be successful.

Previous researchers have criticized measures of conscientiousness as consisting
of two constructs, achievement and dependability, instead of a general construct of
conscientiousness with each component having different conceptual deﬁnitions and
relationships with external variables (Borman et al, 1991; Hough, 1992). While both
achievement and dependability should be highly related conceptually and empirically,
they may display differential relationships when examining the speed-accuracy tradeoff
in speeded test performance. In particular, the achievement component may display a
relationship with how individuals treat errors in favor of completing more items. While
both notions of achievement and dependability, as well as conscientiousness, represent

notions of persistence, effort, and discipline, the speeded nature of the test situation may

53

assist an individual who is high in achievement to recognize that the key to success in the
speeded test scenario is how fast he/she can complete the task. This could reduce the
amount of focus placed on the accuracy of the response in favor of completing more
items to succeed. It is this component of achievement that may lead effort to be directed
at favoring speed over the carefulness of checking errors

Hypothesis 4: Achievement should have a positive relationship with the number of errors
committed on measures of speed.

Temporal orientation. Two constructs that are used to deﬁne an individual’s
temporal orientation are (1) monochronic vs. polychronic orientation, and (2) pace.
Individuals with a monochronic orientation approach tasks one at a time whereas
individuals with a polychronic orientation are simultaneously involved in two or more
activities (Bluedorn et al, 1992). Individuals who are monochronic tend to be more task-
oriented, prompt, and deadline/plan conscious whereas individuals with polychronic
orientations are more relationship than task conscious. In speeded test scenarios, success
might be more in line with individuals who favor the monochronic orientation. These
individuals should have a predisposition to focus effort on the completion of the task at
hand, which in the employment testing scenario is the measure of speed. Individuals with
a polychronic orientation may be more concerned with other tasks they might need to
complete, and thus divert their attention and effort to other tasks, thereby reducing their
level on the measure of speed. Even though the testing scenario may force them to focus,

their ability to do so through prior habits of diverging the effort may inhibit their success.

54

Hypothesis 5 : Individuals with a polychronic orientation should perform worse on
measures of speed than individuals with a monochronic orientation.

The concept of pace is similar to notions of impatience/irritability in terms of an
individual’s predisposition to perform tasks in a rapid fashion. However, pace could be
construed as a tempo variable without the irritability aspect of 11. While missing the
irritability aspect, the element of arousal or activity level should still be high in
individuals who maintain a quicker tempo in the performance of activities. Research has
noted that pace of life differences have been documented across individuals (Levine,
1988) and has been proposed to be a dimension of speed in timed performance tasks
(Marr & Stemberg, 1987). This connection to measures of speed has never been fully
investigated empirically. In the speeded test scenario, individuals who have a quicker
pace in the performance of their everyday activities should have higher arousal levels and
be more alert to stimuli presented, make quicker decisions, and make quicker movements
in responding.

Hypothesis 6: Individuals with a quicker life pace should obtain higher scores on
measures of speed.

Self-monitoring. Self-monitoring is related to the extent to which a person in a
social situation pays attention to social cues and actively attempts to construct a pattern of
behavior that is appropriate to that particular context (Snyder, 1979). In the context of
employment testing, particularly measures of speed, the situation cues applicants that to
have an opportunity to obtain the job, they must exert maximum effort and work rapidly

to be successful. Individuals who are high in self-monitoring should regulate their

55

behavior to exert such effort while individuals who are low in self-monitoring should
perform behaviors, be they high or low effort, that are similar to those in other situations
outside the employment context. In that sense, the disposition-behavior relationship
would be more consistent rather than situationally inﬂuenced when an individual is
characterized as being low in self-monitoring.

In the speeded test scenario, the previous hypotheses propose disposition-behavior
relationships between conscientiousness and polychronic orientation on measures of
speed. In these situations, self-monitoring could serve as a moderating variable between
disposition-behavior relationships in that individuals who are high in self-monitoring, but
low on conscientiousness and possessing a polychronic orientation, may exhibit
behaviors more characteristic with being high in conscientiousness and having a
monochronic orientation. This would be the case particularly in situations whereby the
mechanism of inﬂuence is the direction of effort. People who are high in self-monitoring
would recognize the need to exert more effort and work at a faster pace to meet
performance demands. Such a moderator should serve to reduce the relationship between
the dispositions and speeded test performance for individuals who have high levels of
self-monitoring. Other research investigating test-taking motivation also found that the
criterion-related validity of personality tests was higher in samples with less test-taking
motivation than samples with high test-taking motivation (Schmit & Ryan, 1992) which
would similarly argue that the situation regulates behavior to a certain extent. A similar
way of explaining the disposition-behavior relationship would view candidates as using a

self-presentation strategy to maximize the possibility of employment (Sackett, Burris, &

56

Ryan, 1989). In their case, the characteristics and demands of the situation are vital to the
prediction of behavior. With self-monitoring, such self-presentation strategies should be
i more unconscious in that individuals adapt to situational cues in a more automatic
fashion. However, individuals with a low level of self-monitoring would behave
consistently in these scenarios as they would in other situations, thus maintaining the
validity between dispositions and behaviors.
Hypothesis 7-8: Self-monitoring should moderate the relationship between
conscientiousness and monochronic orientation with performance on measures of speed
such that individuals who are high on self-monitoring would have a lower relationship
between dispositions and measures of speed than individuals with lower levels of self-
monitoring. The nature of the hypothesized interaction is displayed in Figure I.
Circadian Rhﬂhms. Time-of-day has been demonstrated to have signiﬁcant
effects in many types of human performance (Blake, 1967; Colquhoun, 1971; Kerkhof,
1985; Revelle, 1989). These effects have been attributed to variation in circadian arousal
levels affecting an individual’s level of alertness or fatigue. If an individual is tested at
the time when their arousal level is peaked, they should be more active, aroused, and
alert. Such high levels of arousal should positively affect one’s performance on speeded
tests. The higher levels of arousal should allow one to be more alert, respond more
quickly to stimuli, etc.. Individuals could be characterized as having peak times either in
the morning or in the aftemoon/evening (Bodenhausen, 1990). If the testing time

(morning or afternoon/evening) corresponds to the time when an individual’s level of

57

 

 

Low Conscientiousness 0r Polychronic
Orientation
Slow

High Conscientiousness or
Monochronic Orientation

 

 

 

SPEED

Fast

 

Low High
SELF-MONITORIN G

 

 

 

Figpre l

Hypothesized Interaction Between Self-Monitoring and Personality Characteristics on
Measures of Speed

58

circadian rhythms is at a peak, the individual should have higher levels of alertness
leading to higher performance on the speeded test.

Hypothesis 9: Individuals whose testing time (morning or afternoon/evening) matches the
time when their circadian rhythm level is highest should perform better on measures of
speed than individuals whose testing time does not match their peak level based on
circadian rhythms.

A_n_x_im. The speeded test scenario has a deﬁnite time pressure component
involved. In a way, individuals are “racing” against the clock to either complete as many
tasks as possible in a limited time period or they are attempting to complete the task in as
little time as possible. Such pressure could increase an individual’s anxiety and thereby
reduce the amount of on-task effort that is exerted on the completion of the measure of
speed. Instead, individuals could be concerned about how well they are doing, how well
others are doing, how much time is left, etc.. The more time pressure or the more
difﬁcult the task, the higher the levels of reported comparative anxiety and external
attribution of poor scores (Arvey et al, 1990). Research by Arvey and colleagues (1990)
found that comparative anxiety due to the testing situation was signiﬁcantly related to
scores on speeded tests of data comparisons (r = -.213, p < .05). A major problem with
the relationship between anxiety and speeded test performance is the difﬁculty in
determining causality (Chan et al, 1996). It is difﬁcult to determine if the anxiety is
causing poor performance or if poor performance is leading is external attributions of

anxiety. Therefore statements of causality cannot be made with complete certainty.

59

However, in either case, a negative relationship between anxiety and performance on
speeded tests is e'Xpected.

Hypothesis 10: Individuals who are high as opposed to low in reported anxiety should be
associated with poorer performance on measures of speed.

Other Inﬂuences: Covan’aﬁs

In addition to the above inﬂuences, age, test-taking motivation, and perceived face
validity could have an inﬂuence on measures of speed. However, restriction of range in
an applicant setting, most typically for entry level positions, could nullify the impact of
each of the variables. Each of the variables will be assessed in the research but will serve
as control variables.

Overall, the model for ﬁrst research objective exploring the relationships of
important factors with speed is highlighted in hypotheses 1-10 and displayed in Figure 2.
Hypothesis 1 represents the effect of cognitive ability on measures of speed and
hypotheses 2-6 represent the relationships of noncognitive personality characteristics
(impatience/irritability, conscientiousness, monochronic/polychronic orientation, and
pace) with speed. Of these six, only hypothesis 4 explores the relationship between a
noncognitive measure (achievement) and errors committed on speeded tests. Hypotheses
7 and 8 highlight the interactive effects of self-monitoring with measures of speed.
Hypotheses 9 and 10 examine the relationships between two person-situational inﬂuences
(circadian rhythms and anxiety) on speeded test performance.

The preceding hypotheses encompass the ﬁrst objective of the research

investigating the relationships of important factors with measures of speed. The second

60

to m we 85322 5:5 .9. E

 

mecca—om
g

.3593 can :85 .«o mEmEmnuoE

 

 

 

868223; .
3:58qu .

3on mo 8:532

b2x=<

 

 

\\

632
255
568:0

 

 

\

 

8am

 

 

 

 

 

 

 

 

 

 

v: 6:5 m:

I:

 

 

Marcie—2
.:om

 

 

h:

 

:23: :0
25.530;
msmuu>
2:028:02

 

 

 

 

EoEo>qu< -
3083258200

 

 

 

 

 

9.3.3:.
382.38.

 

sewage—2
may;
rumOr—t 0

36:;
8am
quoBom .

ow< .

woumcgou

 

 

 

 

 

 

 

33¢.
95500
.8280

 

 

61

objective of the research is to examine the incremental validity of speeded test scores
above and beyond cognitive ability on job performance and will be expanded upon in the

following paragraphs.

Research Study: Incremental Validity of Speed on Job Performance

A second major question being asked by the current research is the relationship of
measures of speed to measures of job performance. In other words, do differences in the
speed by which individuals complete tasks inﬂuence overall job performance? Much
research has demonstrated the relationship of cognitive ability and job performance
across a wide range of jobs (e.g., Hunter & Hunter, 1984) with some researchers
indicating that other variables might add little to the overall prediction of job
performance above and beyond the predictability associated with general cognitive ability
(e.g., Ree & Earles, 1992). While not directly the focus of the study, a relationship
between cognitive ability and performance is expected.

Hypothesis 11: Individuals with higher measures of general cognitive ability should
score higher on job performance measures.

However, the major question asked by the current research is whether or not
measures of speed can add incremental validity after cognitive ability has been accounted
for in predicting job performance. The criterion of performance is multidimensional in
that certain dimensions of performance might be better predicted by cognitive ability
while other dimensions of performance might be better predicted by other variables
(Schneider & Hough, 1995). For instance, in Project A, criterion dimensions such as core

technical proﬁciency were highly related to measures of cognitive ability, while criterion

62

dimensions such as effort and discipline were highly related to personality constructs
(McHenry et al, 1990). Measures of speed require individuals to work at a quick pace
over a certain period of time. Criterion dimensions that might be related to speed include
an individual’s proﬁciency in meeting tight deadlines or working under pressure,
demonstrating effort, and maintaining personal discipline. For these types of criterion
dimensions of job performance, measures of speed might have the highest potential of
adding incremental validity over cognitive ability. In scenarios measuring speed,
individuals are completing tasks “against the clock” in such a way that time pressure is
evident in the testing situations. Individuals who do well on these tasks can maintain
effort and focus on the task and work at a quick pace. The tasks in a speeded test are at a
low level of difﬁculty such that problem-solving activities and decision-making aspects
of performance that are typically highly related to measures of general cognitive ability
might be limited or reduced. When the level component (cognitive ability) is accounted
for in performance measures, speed should still have an incremental effect on
performance in these contexts, particularly with the criterion dimensions mentioned
above that deal with meeting deadlines, maintaining deadlines, demonstrating effort, etc.
In addition to objective measures, subjective measures of performance provided
by raters could be equally affected by individual differences in measures of speed. Some
research has demonstrated that an individual’s “implicit” theories of intelligence contain
a notion that individuals who complete tasks faster are deemed more intelligent
(Stemberg, 1985; Stemberg, Conway, Ketron, & Bernstein, 1981). Ratings of
performance could be driven by the salience of viewing individuals working at a faster

pace than others, even if they are performing the task incorrectly, or of viewing

63

individuals developing solutions to problems faster than others, even if others would have
come up with similar solutions. Therefore, speed might be viewed positively from raters
in such a way that individuals who perform and discuss issues more quickly than others
may be viewed more poSitively. While it is expected that measures representing effort
would have a higher relationship with speed than other dimensions, general impressions
and method effects from the assessment center will probably increase the relationship
with speed as well (Gaugler et al, 1987).

Hypothesis 12: Measures of speed should add positive incremental validity, particularly
with job performance measures that emphasize demonstrating eﬂort, working under
pressure, and meeting deadlines in job performance (work orientation ratings, number of
units completed, and proportion of errors-quality) as opposed to performance measures

which are non-effort based (e. g., team skills and problem solving ratings)

64

METHOD

Participants

The original sample consisted of 912 job applicants for an entry-level production
position at a large manufacturing company. This sample was reduced using various data
cleaning steps to eliminate individuals who did not appear to take the data collection
seriously. First, three individuals were eliminated for failing to respond to more than
10% of the items on the additional research questionnaire. An additional 37 individuals
were eliminated for not providing a response to the question involving age, a critical

icovariate in the study. An additional 18 individuals were eliminated for indicating that
they were either neutral or had little interest in attaining the position for which they were
applying. While failing to fully complete the items on the additional research
questionnaire or indicating they had little interest in the job eliminated these applicants
from this research, it did not have any effect on their employment eligibility. Finally, one
individual was eliminated for not completing the measure of psychomotor speed to bring
the ﬁnal sample to 853 job applicants. In terms of the demographic characteristics of the
sample, 54% were men, and 67% were white with the next largest minority group being
Aﬁican American (25%). In terms of age, 50% were less than 30 years old, 30% were
between the ages of 31 and 40, 16% were between the ages of 41 and 50, and 4% were
between the ages of 51 and 60. This sample exceeded the requirement of 348 applicants
(see Appendix A) needed to detect small effect sizes (AR2 = .03) for the interactions at a
power level of .80 and or = .05 (Cohen, 1988). Of the 853 applicants, only 531 passed the

cutoff established in the ﬁrst phase of the selection process. Therefore, questions

65

concerning incremental validity were based on that sample rather than the full sample of
853.
Measures Investigating Relationships with greed

General cogpitive ability. General cognitive ability was a composite of two tests
measuring quantitative and reasoning abilities. The two tests, Practical Arithmetic (24
items) and Following Policies and Procedures (25 items), are each administered under
time limits. Due to copyright and proprietary restrictions, these tests are not available for
reproduction in the appendices. However, a controlled review of the instruments is
available through the author. Such administration conditions are a concern in
relationship to examining speed versus power issues by questioning the extent to which
the relationship between cognitive ability and speed might be elevated due to the speeded
nature of the cognitive ability tests. To alleviate such concerns, the Educational Testing
Service (ETS) uses a “rule of thumb” approach whereby a test is considered speeded if
less than 80% of the examinees ﬁnish the items (Peterson, 1993). In a study examining
the speededness of the tests, it was found that 16% of a sample of examinees for an entry-p
level production job failed to ﬁnish the Practical Arithmetic test, and 20% failed to ﬁnish
the Following Policies and Procedures test (n=2615). However, 96% of the examinees
ﬁnished the ﬁrst 18 items of Practical Arithmetic test and 95% of the examinees ﬁnished
the ﬁrst 20 items of the Following Policies and Procedures test. Using a conservative
estimate in relationship to ETS’s standard, the general cognitive ability measure will be a
total score of those 38 items. 63 undergraduates from a large midwestem university
completed both instruments along with the Wonderlic Personnel Test (1992), a well

documented measure of general cognitive ability. A substantial relationship (r = .83; p <

66

.05) between the 38 item total score and scores on the Wonderlic were found after
correcting for restriction of range due to differences between university (n=63) and job
applicant (n=2615) samples. Further corrections for measurement error using a
conservative estimate of reliability (.94) for the Wonderlic (Wonderlic, 1992) and a
coefﬁcient alpha of .88 for cognitive ability in the current research sample, found a “true”
relationship of r =.91; p < .05.

anatience/Irritability. A ﬁve-item measure based on the work by Spence and
colleagues (1987) was used (see Appendix B). A coefﬁcient alpha of .63 was found in
the current research sample.

Conscientiousness. Two scales were used to assess conscientiousness. The ﬁrst
scale is a seven-item measure of achievement strivings (see Appendix C) based on the
work of Spence and colleagues (1987) with a coefﬁcient alpha of .64 in the current
research. The second scale is a 23-item measure of conscientiousness using background
information related to work from the Work Experience and Interest Inventory, a product
of Aon Consulting which generated a coefﬁcient alpha of .60 in the current research
sample. Due to copyright and proprietary restrictions, these tests are not available for
reproduction in the appendices. However, a controlled review of the instruments is
available through the author.

Polychronic vs. monochronic orientation. A ﬁve-item scale based on the work of
Kaufman and colleagues (1991) and Tuttle (1996) was used with higher scores
representing polychronic orientation and lower scores representing monochronic

orientation (see Appendix D). Current research found a coefﬁcient alpha of .81.

67

Page. A ﬁve-item scale based on the work of Tuttle (1996) was used with higher
scores representing faster pace of life (see Appendix E). The current research sample
found a coefﬁcient alpha of .67.

Self-monitoring. Two options are present in measuring self-monitoring as
conceptualized by Snyder: the original 25-item scale (Snyder, 1974) and a revised scale
(Gangestad & Snyder, 1985) that contained 18 of the original items. Gangestad and
Snyder (1985) removed seven items ﬁom the original scale that had low item-total
correlations. Since all the items in the revised scale are contained in the original scale,
the original scale will be used to assess proper coverage of the self-monitoring construct
domain (see Appendix F). The current research found similar mean interitem
intercorrelations between the two scales, therefore, the 25 item scale was used
(coefﬁcient alpha = .62).

Circadian rhythms match. A l3-item scale constructed by Smith and colleagues
(1989) using the best characteristics of other circadian rhythm scales will be used in the
current study (see Appendix G) High internal consistency estimates (0t= .87, .83 on
different samples of college students) and relationships to external criteria associated
with morningness and eveningness have been reported for this scale. The current
research sample found a coefﬁcient alpha of .85. Using suggestions by Smith et a1
(1989), the scores of the scale were divided into three categories based on a 10-90 split.
The top 10th percentile was equated with being a Morning Type, the bottom 10th
percentile was equated with being an Evening Type, and the middle 80% were deemed
Intermediate. Based on that categorization and the time of day individuals took the test

(7:30 am, 1:30 pm) a circadian rhythms match variable was created. If a person was a

68

Morning Type and took the various measures in the morning or was an Evening Type and
took the various measures in the afternoon, they were given a score of 3. If the reverse
occurred (Morning & pm; Evening & am), individuals were given a score of 1. All other
combinations received a score of 2.

A__n_xﬁy. The ten-item scale measuring comparative anxiety that was developed
by Arvey and colleagues (1990) was used in the current study (see Appendix H). The
scale displayed a high coefﬁcient alpha (.85) in the current sample.

Covariate Measures

In addition to the above measures, 17 items (see Appendix I) were used to reﬂect
various covariates including the following: (1) age (one item), (2) motivation consisting
of interest in obtaining the job (one item) and test motivation (ten items from Arvey et al,
1990), and (3) perceived face validity (ﬁve items ﬁom Smither and colleagues (1993)).
The test motivation and perceived face validity scales displayed high coefﬁcient alphas
(.83 and .81, respectively) in the current sample.

Measures of Speed

Anentivemercwmm. Individuals completed the 60 item Data Comparison
Test. Due to copyright and proprietary restrictions, these tests are not available for
reproduction in the appendices. However, a controlled review of the instruments is
available through the author. Individuals were presented with two lists, a list to be
checked and a correct list. Each test contained four lists with 15 items per list.
Participants determined whether any of the information in the two lists did not match
exactly. Individuals had 10 minutes to work on the test. In a study examining the

speededness of the tests, it was found that 98% of examinees for an entry-level

69

production job failed to ﬁnish the test, providing sufﬁcient basis to conclude that the test
is highly speeded (n=2615). Scores were obtained on both how many items the
participants ﬁnished and the proportion of correct answers obtained on those items (see
Hypothesis 4).

Psychomotor speed. Participants took two similar forms of a test measuring
psychomotor speed. In the ﬁrst test (parts transfer), individuals were timed when moving
a series of parts from one side of the testing kit to the other. In the second test (small
parts), individuals were required to switch the order of a series of small pegs, so the
bottom peg was on the top peg and vice versa. Individuals practiced both tests prior to
testing to assure that they understand the nature of the exercise. Scores were obtained on
both tests in terms of how long it took the participants to ﬁnish and how many errors
were committed. However, errors on these tests were minimal. An average score
representing the amount of time required to complete the two tests was used (coefﬁcient
alpha = .60).

Mums of Job Performm

In an assessment center exercise, participants worked individually to assemble
“cam boxes” to supply a ﬁctitious Install Team. Each participant produced cam box
assemblies according to pre-determined instructions and put their completed units into a
common bin. The group then discussed how to improve their efﬁciency. Then,
participants began a second assembly session to test improvements. In this procedure,
several performance dimensions were assessed. A more complete overview of the

assessment phase is found in the procedure section that follows.

70

Subjective measures. Two raters provided ratings on behavioral checklists for
three dimensions: work orientation, problem solving, and team skills, factors that are
similar to those commonly found in assessment center activities (Schmitt, 1977). Due to
copyright and proprietary restrictions, these tests are not available for reproduction in the
appendices. However, a controlled review of the instruments is available through the
author. Work orientation is deﬁned as the willingness to accept personal responsibility
for one’s actions and performance and to recognize the importance of productivity and
speciﬁc procedures. It also includes the ability to perform repetitive tasks and Operations,
including the ability to work without close supervision. Problem solving is the ability to
learn, understand, and improve on ideas, systems, or processes, and to generate new ideas
and techniques to solve problems or improve operations. Team skills represent the ability
to interact effectively with others whether in a team or in one-to-one situations. This
includes the ability and willingness to create and maintain cooperative relationships and
to be sensitive to others’ points of view as well as to tactfully deal with disagreements
and demonstrate self-assurance and a positive point of view. All raters took a two-day
course on how to administer and make ratings on the exercise and all passed a
certiﬁcation test to ensure they were able to correctly perform the task. In terms of the
actual ratings, most applicants completed the assessment phase in groups of six. Each
rater rated only four people, thereby allowing an overlap rating on 33% of the candidates.
A ﬁnal measure of each dimension was based on an average of the two ratings. If only
one rating was provided, that rating was used as the ﬁnal measure of the dimension.
Interrater reliability estimates for work orientation, problem solving, and team skills were

.40, .48, and .56 respectively (n=232).

71

Objective measures. In addition to the subjective measures, two objective
measures were assessed. Participants were given instructions and a modeling display on
how to create “cam boxes” and practiced constructing two of these devices following
speciﬁc guidelines. To build the product, participants had to follow speciﬁc instructions.
Aﬁer that time, participants had 10 minutes to construct as many “cam boxes” as they
could without having any construction errors or including defected parts in the box. In
the exercise, participants placed their ﬁnished products in a common bin, unaware that
their boxes were marked to identify the individual who constructed the box. Measures
were assessed both in terms of how many assembles were created and how many errors
were made in the construction of the boxes. Measures of objective output were only
taken after the ﬁrst assembly period to control for any group effects that might have
enhanced productivity in subsequent assembly periods.

Procedure

Testing phase. Participants applying for entry-level jobs with a large
manufacturing company are part of a multiple hurdle selection process. At the ﬁrst
phase, applicants took a series of paper-and-pencil tests in the following order: (1)
psychomotor speed, (2) Following Policies and Procedures, (3) Practical Arithmetic, and
(4) Data Comparison Test. Halfway through the above tests, participants received a short
break. Aﬁer those tests were completed, paper-and-pencil questionnaires containing
many of the above measures were administered. Participants were instructed that the
results of the questionnaire are for research purposes only and that the results would not
affect their status as a candidate. In addition, they completed a consent form before the

questionnaire was administered explaining the nature of the study. Testing occurred in a

72

group setting either in the morning or in the aﬁemoon. Participants were randomly
assigned to sessions or in cases of time conﬂicts, assigned to sessions based on their
availability. In addition, applicants completed a series of applications and reference
check forms as required by the manufacturing company and for system tracking reasons.

Assessment phase. After the testing phase was completed, individuals proceeded
to the assessment phase of the selection process if their test scores were above the cutoffs
set. Only 62% of the 853 applicants passed the cut scores of the testing phase, thereby,
reducing the sample to 531 applicants. The assessment phase took place several days
after the testing phase.

Assessment took place in groups of four to six individuals. Participants were
seated in an assigned work station containing an assembly stand, a parts bin containing
unassembled parts, a defects bin to place defected parts, and an example cam box.
Participants were told not to handle any of the exercise materials. To begin the exercise,
participants were welcomed and given a description of the day's activities. They were
told that they would be given exercises that would simulate situations they might
encounter on the job. In addition, applicants were told that the exercises were not a
competition, so everyone could potentially do well. Next, participants were told that it
was important to participate in the exercise and that by not participating they could be
hurting their chance for employment. Participants were then asked to reply that they
understand this. They also were told that all instructions would be read from a script by
the raters and other than that, the observers would not participate during the exercise. At

the conclusion of the introduction, participants were given a chance to ask questions.

73

After the instructions phase, participants were given an introduction to the nature
of the exercise. They were told that they were a member of an assembly team for a
company that manufactures "cam boxes," and that the company is facing tough
competition and needs to improve its productivity and quality to stay proﬁtable and the
company is counting on the employees to help with this. They were also told that the
company knows some of its assembly processes are inefﬁcient and that the employees
needed to suggest better ways of doing things. In the exercise, they were told that they
must ﬁrst assemble products according to speciﬁc instructions. Then the group would
meet to come up with better ways of assembling boxes. They were told that they could
completely change the way they were ﬁrst told to work unless the instructions
speciﬁcally tell them they can't. So, for the ﬁrst part of the exercise, they must follow
instructions exactly and for the second part they can change and improve the way they do
work. Participants were then asked if they understand this. Some speciﬁc instructions
they were told to follow for the exercise include the following: (1) immediately pick up
any parts that you drop, (2) keep your work area neat, and (3) no running. They were
also told that they could talk to each other anytime instructions were not being given or
when questions were being answered. Finally, they were told that they did not need to
know anything about assembly to do well in the exercise and that they would be provided
with all the instructions they need. In addition, they were told that they would be
provided handouts that they must read and follow closely. If a decision is not covered in
the instructions or handouts, they were free to make it. At this point, participants were

given time to ask questions.

74

After the introduction, participants were taken through a mini-training course on
how to perform the task. They were introduced to their work station, the parts that would
be used to assemble the boxes, the work aid sheet in front of them that outlined the
procedures to follow in building cam boxes, and an example cam box that was correctly
constructed. Next, participants were told that they must examine parts for defects and
were visually given examples of each kind of defect. An outline of the defects was
provided on their handout as well. Next, one rater provided a visual example of how to
construct a box by going through all the required steps. Participants followed the process
exactly and assembled a cam box while the rater demonstrated the process. At the end of
the assembly demonstration, every participant placed their completed units in the same
common bin in the center of their work stations. At the end of the demonstration, the
participants were asked if they had any questions. After all questions were answered,
participants were given a summary of the steps required to build cam boxes and then
were told to complete two more cam boxes according to the process that was just
demonstrated to them. They were also provided with a reproduction of the steps on the
handout at each station. While building the two extra units, raters walked around the
stations providing feedback to ensure that the participants understood the proper
assembly process.

At the end of the training, participants began the exercise and were told they had
10 minutes to assemble boxes. They were told that they had been requested to make 60
boxes if they were a group of six participants, 50 boxes if they were a group of ﬁve

Participants, and 40 boxes if they were a group of four participants. Once again, they

were told they must follow the assembly process steps exactly and were asked if they had

75

any questions. They were also told that they would be notiﬁed when there were two
minutes left and that they were free to talk to each other. At the end of ten minutes,
participants were told to stop and disassemble any partially assembled units and place the
parts back in their parts bin. When ﬁnished, each individual was given a suggestion sheet
and three minutes to write down their own ideas for improving the cam box production.
At the end of the three minutes, they were to hand the sheet to the rater.

During the ﬁrst team meeting, the participants were told that they needed to report
about their production during the ﬁrst assembly session. They also were given a memo
that stated exactly what they needed to do during the meeting. The rater passed out the
memo and went over each step verbally with the participants. During the meeting, the
1 participants checked the quality of the cam boxes produced using gauges, ﬁlled out a-

productivity chart outlining how many units were correctly assembled, and calculated the
ratio of correctly built cam boxes compared to total cam boxes. In addition, the
participants discussed how they would change their assembly methods to build more and
better cam boxes. They were told that they could change anything to improve quality and
productivity except one thing: each participant had to assemble their own cam boxes at
their own station using their own parts. At the end of the discussion, the participants
decided on a goal for the next 10-minute assembly session and using the productivity
chart, they circled the number of units they would correctly assemble in the next session.
They were also told that the company would use all of the cam boxes they could get.
During the team meeting, they were notiﬁed that there were two things they could not do:
(1 ) they could not move the unassembled parts, and (2) they could not disassemble or

Change any of the completed assemblies. They were told they would have 10 minutes to

76

conduct the meeting, asked if they had any questions, and informed that they would be
reminded when there were two minutes left. At the end of 10 minutes, they handed in all
completed materials and cam boxes from the ﬁrst assembly period.

After the team meeting, participants were given three minutes to set up their work
stations for the next assembly period and asked if they had any questions. The only
things they could not do were to begin to assemble units and check parts for defects.
Once again, they were asked if they had any questions. At the end of the three minute
set-up, the next ten-minute assembly period began with participants placing all completed
boxes during the assembly period in the common bin again. Once again, participants
were reminded when there were two minutes left and at the end of ten minutes, they were
instructed to stop and disassemble any partially assembled cam boxes and put all unused
parts away.

At the end of the second assembly period, participants conducted another team
meeting where they repeated the activities that occurred in the ﬁrst meeting. Once again
they were reminded not to move unassembled parts or to disassemble or change
completed assemblies. Participants were given ten minutes to conduct the meeting, asked
if they had any questions, and told to begin. With two minutes remaining in the exercise,
participants were reminded that they only had two minutes left. At the end of the ten
minutes, participants were told to stop and return all the completed cam boxes and
materials in the center bin. Participants at that point were told that the exercise was
completed and thanked for their time.

After the participants left, raters scored the assembly units by placing each

completed cam box assembly at the appropriate work station. Each cam box was

77

165ml!
defect
rich :
static

of:
mar

-- ..
.dIsL

I}?

identiﬁed using a UV light to read the station marker on the cam box. In addition, each
defected part included in the cam boxes was identiﬁed using the UV light. Each box for
each assembly session was assessed in terms of how many boxes were produced at each
station, how many boxes were incorrectly assembled, and how many defected parts were
included in cam boxes for the objective measures of performance (quantity and quality).
Both quantity and quality measures were assessed for each assembly period, so separate
measures could be obtained. At the end of the exercise, the two raters also individually
rated the four candidates in the group on the three subjective dependent variables

(problem solving, work orientation, and team skills) using behavioral checklists.

Data Analysis

To test and analyze the hypotheses in the current research, the results were broken
down into two parts corresponding with the dual purposes of the present research. For
the ﬁrst part of the research, investigating relationships with measures of speed, the
investigation of various relationships of cognitive and noncognitive factors on measures
of speed is examined both for measures of psychomotor and perceptual speed (Data
Comparison). This analysis was conducted in a series of steps. First, zero-order
correlations and simple regressions (minus the covariates) were conducted to examine if
there were any relationships between the measures of speed and the noncognitive
measures and which of the noncognitive measures added the most predictive efﬁciency to
the measures of speed. Second, the ﬁrst group of covariates (age, test motivation, and
perceived face validity) were added to the analysis to control for demographic and

motivational differences of applicants. Partial correlations investigating bivariate

78

relationships between the noncognitive measures and measures of speed with the
covariates as controls were computed. Then hierarchical regression was used to control
for the covariates prior to examining the noncognitive relationships. Finally, general
cognitive ability, a well proven measure of speed, was added as a covariate in a second
step to determine which noncognitive measures maintained their relationship to measures
of speed after variance was accounted for by general cognitive ability and age, test
motivation, and perceived face validity. Once again, three analyses were conducted
using partial correlations and hierarchical regression with the covariates entered in the
ﬁrst two steps. These analyses satisﬁed the requirements for tests for hypotheses 1-3, 5-
10. In each of these analyses, the product term between conscientiousness and self-
monitoring (H7) and the product term between polychronic orientation and self-
monitoring (H8) were added as the ﬁnal step of the hierarchical regressions. This
allowed the research to detect the role of self-monitoring as a potential moderator.
Hypothesis 4, examining the relationship between measures of achievement and
proportion correct in perceptual speed was examined using a similar series of analyses
using proportion correct as the dependent variable rather than time or number completed.
The second section examines the research question investigating how much
incremental validity do measures of speed add to the prediction of job performance above
and beyond that of cognitive ability. To test the second major research question (Hll
and H12), partial correlations and hierarchical multiple regression were used. The
contribution of each individual measure of speed, perceptual and psychomotor, was
assessed using partial correlations. Next, separate hierarchical multiple regressions were

used to examine the contribution of both speed measures on the ﬁve measures of job

79

performance. Larger relationships were expected on measures requiring effort, arousal,
and maintaining discipline, performance constructs hypothesized to be most related to
measures of speed. These measures include the total number of assemblies produced, the
proportion of defects built into the assemblies, and work orientation. Two hierarchical
multiple regressions were also performed on performance ratings of problem solving and

teamvskills, dimensions that should not be as highly related to measures of speed.

80

RESULTS

B_e11ation,s_hips with Lipasures of Speed

Mm relationships. Descriptive statistics and zero-order correlations between
variables are presented in Table 2. In examining the matrix, signiﬁcant positive
relationships were found between general cognitive ability (Hl), conscientiousness (H3),
and pace (H6), with perceptual speed such that higher levels of measures of these
constructs led to completing more items. A signiﬁcant negative relationship was found
between polychronic orientation (H3) and comparative anxiety (H10) with perceptual
speed such that someone who was more oriented toward multiple activities and had more
comparative anxiety completed less items than those with a more monochronic, singular
focus and less anxiety. In addition, signiﬁcant negative relationships were found between
general cognitive ability (H1), conscientiousness (H3), and pace (H6), with psychomotor
speed such that higher levels of these constructs were related to more time required to
complete the task. A signiﬁcant positive relationship was found between polychronic
orientation (H3) and comparative anxiety (HlO) with psychomotor speed such that someone
who oriented toward multiple tasks and someone with higher levels of anxiety took more
time to complete the task providing some initial support for those hypotheses.

Next stepwise hierarchical regression was utilized on perceptual speed and the
results are presented in Table 3. Given the strong relationship between general cognitive
ability and measures of speed, it was omitted from this analysis to determine the nature of

the relationship of noncognitive measures with speed in the absence of such a factor.

81

Table 2

Summm of Means Standard Deviations, Reliabilities, and Zero-Order Intercorrelations
Investigating Relationships to Measures of Smed

 

 

Variables Mean SD 01 02 03 04 05 06 07
01 General Cognitive 26.56 6.63 (.88)
Ability '
02 Irnpatience/ 2.27 .53 -.02 (.63)
Irritability
03 Achievement 3 .90 .46 . 12* -.29* (.64)
04 Conscientiousness 20.68 2.06 .28* -.22* .31* (.60)
05 Polychronic 3.14 .77 -.l6* .11* -.21* -.18* (.81)
Orientation
O6 Pace 3.75 .55 .10* -.02 .36* . 14* -.26* (.67)
07 Self-Monitoring 3 .35 .31 -.04 -.21* .06 . 14* .07* -.08* (.62)
08 Momingness 41.72 6.51 -.11* -.3 1* .32* .21* -. 12* .08* .19*
09 Circad. Rhythm 2.06 .47 -.02 .03 .06 .00 .06 .02 .05
Match
10 Comparative Anxiety 2.57 .68 -.45* -.31* -.18* -20* .10* -.15* .02
11 Age (Categorical) 1.74 .88 -.25* .04 -. 10* .07* .01 -.23* .20*
12 Test Motivation 4.47 .43 .13* -. 19* .34* .23* -.01 .22* .10*
13 Perceived Face 3.86 .67 .21* -.20* .28* .23* -.15* .19* .03
Validity
l4 Perceptual Speed 31.19 7.69 .41* -.01 .07* .10* -.11* .12* -.05
(Data Comparison)
15 Perceptual Speed - .80 .13 .53* -.02 .06 .20* -. 12* .04 -.03
Proportion Correct
16 Psychomotor Speed 121.1 17.6 -.41* .05 -.13 -. 14* .12* -.l6* 04
(Average Time)
17 Total Assemblies 8.24 1.24 .01 -.03 -.02 .07 -.09* .03 -.02
18 Proportion of .088 .049 -.17* -.06 .00 .05 -.07 -.03 .00
Defective Parts
19 Problem Solving 4.98 1.15 .16* .02 .07 -.03 -.05 .00 -.O9*
20 Team Skills 5.45 1.17 .12* -.02 .15* .06 -.19* .05 -.04
21 Work Orientation 5.29 .87 .19* .06 .08 .03 -.04 .05 -.06

 

*p<.05; Reliabilities in the diagonal
n=853 for variables 1-16 (n=531 for means, SDs, & intercorrelations with variables 17-21)

82

Table 2 (cont’d)

 

Variables Mean SD 08 09 10 ll 12 13 14

01 General Cognitive 26.56 6.63
Ability

02 Impatience/ 2.27 .53
Irritability

03 Achievement 3.90 .46

04 Conscientiousness 20.68 2.06

05 Polychronic 3. 14 .77
Orientation

06 Pace 3.75 .55

07 Self-Monitoring 3.35 .3 l

08 Morrringness 41.72 6.51 (85)

09 Circad. Rhythm 2.06 .47 .09* ---
Match

10, Comparative Anxiety 2.57 .68 -.O7* .00 (.85)

11 Age (Categorical) 1.74 .88 .23* .O7* .20* «-

12 Test Motivation 4.47 .43 .09* .07* -.22* -08* (.83)

13 PerceivedFace 3.86 .67 .12* .08* -.39* -.10* .33* (.81)
Validity

14 Perceptual Speed 31.19 7.69 -.06 -.02 -.31* -.28* .10* .13* ---
(Data Comparison)

15 PerceptualSpeed- .80 .13 -.07* .07 -.26* -.18* .13* .15* .17*
Proportion Correct

16 PsychomotorSpeed 121.1 17.6 .08* -.05 .24* .39* -09* -.08* -.29*
(Average Time)

17 Total Assemblies 8.24 1.24 -.06 -.09* -.O3 -.13* .02 .02 .22*

18 Proportion of .088 .049 .04 .00 -.04 .13* .01 .03 .05
Defective Parts

19 Problem Solving 4.98 1.15 .04 .03 -.04 .05 .00 .00 .05

20 Team Skills 5.45 1.17 .09* .00 -.05 .07 .00 .00 .08

21 WorkOrientation 5.29 .87 .01 -.12* -.02 -.O3 -.09* -.05 .ll

 

*p<.05;

n=853 for variables 1-16 (n=53l for means, SDs, & intercorrelations with variables 17-21)
NOTE: reliabilities are in the diagonal

83

Table 2 (cont’d)

 

 

Variables Mean SD 15 16 17 18 19 20 21
01 General Cognitive 26.56 6.63
Ability
02 Impatience/ 2.27 .53
Irritability
03 Achievement 3.90 .46
04 Conscientiousness 20.68 2.06
05 Polychronic 3. 14 .77
Orientation
06 Pace 3.75 .55
07 Self-Monitoring 3.35 .3 1
08 Morningness 41 .72 6.5 l
09 Circad. Rhythm 2.06 .47
Match
10 Comparative Anxiety 2.57 .68
11 Age (Categorical) 1.74 .88
12 Test Motivation 4.47 .43
13 Perceived Face 3.86 .67
Validity
14 Perceptual Speed 3 1. 19 7.69
(Data Comparison)
15 Perceptual Speed - .80 .13 ---
Proportion Correct
16 Psychomotor Speed 121.1 17.6 -.32* ---
(Average Time)
17 Total Assemblies 8.24 1.24 .08 -.20* ---
18 Proportion of .088 .049 -.07 .18* .12* ---
Defective Parts
19 Problem Solving 4.98 1.15 -.01 -.11* -.08 .00 (.48)
20 Team Skills 5.45 1.17 .02 —.04 .00 .04 .39* (.56)
21 WorkOrientation 5.29 .87 -.04 -.15* -.O6 —.05 .27* .31* (.40)
*p<.05;

n=853 for variables 1-16 (n=53l for means, SDs, & intercorrelations with variables 17-21)
NOTE: reliabilities are in the diagonal

84

Table 3
Result of Simple Reggession Predicting Perceptual Speed (No Covariates)

 

 

. Step Final 2 2
Variable Beta Beta F R R AR
Step 1
Impatience/Irritability . 10* . 10
Conscientiousness .04 -.32
Polychronic Orientation -.07* .09
Pace .05 .05
Circadian Rhythm Match -.02 -.02
Comparative Anxiety -.32* -.32
Self-Monitoring -.02 -.30 16.12 .34 . 12*
Step 2
Self Monitoring X .53 .53
Conscientiousness
Self-Monitoring X Polychronic -.17 -.17 12.66 .34 .12* .00
Orientation
*p < .05
n=853

NOTE: Step Beta weights refer to the standardized regression weights assigned for each
step of the regression equation

85

In the ﬁrst step, impatience/irritability, conscientiousness, polychronic orientation, pace,
circadian rhythm match, comparative anxiety, and self monitoring were entered as a block
resulting in a signiﬁcant R2 of .12 (p < .05). At this step, impatience/irritability (H2)
displayed a signiﬁcant positive relationship with perceptual speed such that individuals who
were more impatient/irritable completed more items on the Data Comparison test. Also,
polychronic orientation (HS) and comparative anxiety (HlO) displayed signiﬁcant negative
relationships with number of items completed on the perceptual speed measure such that a
person who typically focused on multiple activities or had higher levels of anxiety
completed less items on a perceptual speed measure than those with a more monochronic
orientation or possessing lower levels of anxiety. At the second step, the interactions
between self-monitoring and conscientiousness (H9) and self-monitoring and polychronic
orientation (HlO) failed to provide a signiﬁcant change in R2 providing no support for these
hypotheses.

Next stepwise hierarchical regression was utilized on psychomotor speed and the
results are presented in Table 4. Once again, general cognitive ability was omitted from this
analysis to determine the nature of the relationship of noncognitive measures with speed in
the absence of such a strong factor. In the ﬁrst step, impatience/irritability,
conscientiousness, polychronic orientation, pace, circadian rhythm match, comparative
anxiety, and self monitoring were entered as a block resulting in a signiﬁcant R2 of .09 (p <
.05). At this step, conscientiousness (H3), and pace (H6) and displayed signiﬁcant negative

relationships with psychomotor speed such that higher levels of these constructs resulted in

86

Pat

Cir

C0

Se]

Table 4
Result of Simple Reggession Predicting Psychomotor Speed (No Covariates)

 

 

, Step Final 2 2
Variable Beta Beta F R R AR
Step 1
Impatience/Irritability -.03 -.03
Conscientiousness -.08* -.13
Polychronic Orientation .06 .73
Pace -.10* -.10
Circadian Rhythm Match -.05 -.05
Comparative Anxiety .21* .21
Self-Monitoring .03 .24 11.39 .29 .09*

Step 2
Self Monitoring X .06 .06
Conscientiousness
Self-Monitoring X Polychronic -.74 -.74 9.27 .30 .O9* .00
Orientation
*p < .05
n=853

NOTE: Step Beta weights refer to the standardized regression weights assigned for each
step of the regression equation

87

less time required to complete the psychomotor task. Comparative anxiety (HlO) displayed
a signiﬁcant positive relationship with psychomotor speed such that the more anxiety
experienced, the longer it took to complete the psychomotor tasks. At the second step, the
interactions between self-monitoring and conscientiousness (H9) and self-monitoring and
polychronic orientation (HlO) failed to provide a signiﬁcant change in R2 providing no
support for these hypotheses.

Covgying out age. test motivation and perceived face gliditv. Partial

 

correlations between variables are presented in Table 5 after controlling for variance
accounted for by the applicant’s age, level of test motivation, and perceived face validity.
In examining the matrix, signiﬁcant positive relationships were found between
conscientiousness (H3) with perceptual speed such that higher levels of conscientiousness
was related to completing more items. Signiﬁcant negative relationships were found
between polychronic orientation (H3) and comparative anxiety (H10) with perceptual
speed such that someone who was more oriented toward multiple activities or had higher
with levels of anxiety completed less items than those with a more monochronic, singular
focus or lower anxiety. In addition, signiﬁcant negative relationships were found between
conscientiousness (H3), pace (H6), and circadian rhythm match (H9) with psychomotor
speed such that higher levels of these constructs were related to less time required to
complete the task. Signiﬁcant positive relationships were found between polychronic
orientation (H3) and comparative anxiety (HlO) with psychomotor speed such that someone
who oriented toward multiple tasks or had higher anxiety took more time to complete the

task, providing some initial support for those hypotheses.

88

Table 5

Summary of Partial Correlations Investigating Relatioghips to Measures of Speed
Controlling for AgeLTest Motivation and Perceived Face Validity

 

 

Variables 01 02 03 04 05 06 07 08 09
0 Impatience/ (.63)
1 Irritability
0 Conscientiousness -.16* (.60)
2
0 Polychronic .O9* -.16* (.81)
3 Orientation
0 Pace .04 .10* -.25* (.67)
4
0 Self-Monitoring -.21* .10* .07* -.06* (.62)
5
0_ Circadian Rhythm .05 -.O4 .07 .02 .03 «-
Match
6
0 Comparative .24* -.12* .05 -.04 .01 .02 (.85)
7 Anxiety
0 Perceptual Speed .03 .10* -.10* .03 .00 -.Ol -.25* ---
8
0 Psychomotor .02 -.17* .12* -.07* -.04 -.07* .17* -.19* ---
Speed (Average
9 .
Time)
*p<.05
n=853

NOTE: Reliabilities are displayed in the diagonal

89

Next stepwise hierarchical regression was utilized on perceptual speed and the
results are presented in Table 6. Given the strong relationship between general cognitive
ability and measures of speed, it was omitted from this analysis to determine the nature of
the relationship of noncognitive measures with speed in the absence of such a strong factor.
In the ﬁrst step, age, test motivation and perceived face validity were entered as a block
resulting in a signiﬁcant R2 of .09 (p < .05). At this step, age displayed a signiﬁcant negative
relationship such that the younger the applicant the more items completed on the Data
Comparison test. Perceived face validity also displayed a signiﬁcant positive relationship
with perceptual speed such that people who thought the tests were more face valid
completed more items on the Data Comparison test. At the second step,
impatience/irritability, conscientiousness, polychronic orientation, pace, circadian rhythm
match, comparative anxiety, and self monitoring were entered as a block resulting in a
signiﬁcant AR2 of .08 (p < .05). At this step, impatience/irritability (HZ), and
conscientiousness (H3) displayed signiﬁcant positive relationships with perceptual speed
such that higher levels of these constructs resulted in the completion of more items on the
Data Comparison test. Also, polychronic orientation (H5) and comparative anxiety (H10)
displayed signiﬁcant negative relationships with number of items completed on the
perceptual speed measure such that a person who typically focused on multiple activities or
who had higher levels of anxiety completed less items on a perceptual speed measure than
those with a more monochronic orientation or lower levels of anxiety. At the third step, the

interactions between self-monitoring and conscientiousness (H9) and self-monitoring and

90

Table 6

Result of Simple Regpession Predicting Percaptual Speed Controlling for Age, Test
Motivation and Perceived Face Validity

 

 

 

, Step Final 2 2
Variable Beta Beta F R R AR
Step 1
Age -.27* -.24
Test Motivation .04 .03
Perceived Face Validity .O9* -.01 29.05 .30 .09*

Step 2
Impatience/Irritability .1 1 * .1 1
Conscientiousness .07* -.34
Polychronic Orientation -.08* .02
Pace -.01 -.01
Circadian Rhythm Match .00 .00
Comparative Anxiety -.28* -.28
Self-Monitoring .02 -.32 17.15 .41 .17* .08*
Step 3
Self Monitoring X .60 .60
Conscientiousness
Self-Monitoring X Polychronic -.ll -.11 14.39 .41 .17* .00
Orientation
*p < .05
n=853

NOTE: Step Beta weights refer to the standardized regression weights assigned for each
step of the regression equation

91

polychronic orientation (H10) failed to provide a signiﬁcant change in R2 providing no
support for these hypotheses.

Next stepwise hierarchical regression was utilized on psychomotor speed and the
results are presented in Table 7. Given the strong relationship between general cognitive
ability and measures of speed, it was again omitted from this analysis. In the ﬁrst step,
age, test motivation and perceived face validity were entered as a block resulting in a
signiﬁcant R2 of .16 (p < .05). At this step, age displayed a signiﬁcant positive relationship
such that the older the applicant the more time was required to complete the psychomotor
task. At the second step, impatience/irritability, conscientiousness, polychronic orientation,
pace, circadian rhythm match, comparative anxiety, and self monitoring were entered as a
block resulting in a signiﬁcant AR2 of .06 (p < .05). At this step, conscientiousness (H3) and
circadian rhythm match (H9) displayed signiﬁcant negative relationships with psychomotor
speed such that higher levels of these constructs resulted in the less time required to
complete the psychomotor task. Also, polychronic orientation (H5) and comparative
anxiety (H10) displayed signiﬁcant positive relationships with the time required to complete
the psychomotor task such that a person who typically focused on multiple activities or with
higher levels of anxiety required more time to complete the task than those with a more
monochronic orientation or lower levels of anxiety. At the third step, the interactions
between self-monitoring and conscientiousness (H9) and self-monitoring and polychronic
orientation (HI 0) failed to provide a signiﬁcant change in R2. However, the product term
between self-monitoring and polychronic orientation demonstrated a signiﬁcant negative

beta weight ([3 = -.82, p < .05). The interaction term

92

Table 7

Result of Simple Regpession Predicting Psychomotor Speed Controlling for Age, Test

Motivatiop, and Perceived Fge Validity

Step

Final

 

 

. 2 2
Variable Beta Beta R R AR
Step 1
Age .38* .38
Test Motivation -.05 -.01
Perceived Face Validity -.02 .08 51.98 .39 .16*

Step 2
Impatience/Irritability -.05 -.05
Conscientiousness -.13* -.13
Polychronic Orientation 09* .83
Pace -.02 -.02
Circadian Rhythm Match -.08* -.08
Comparative Anxiety .16* .16
Self-Monitoring -.04 .24 23.08 .46 .22* .06*
Step 3
Self Monitoring X Conscientiousness .00 .00
Self-Monitoring X Polychronic -.82* -.82 19.71 .47 .22* .00
Orientation
*p < .05
n=853

NOTE: Step Beta weights refer to the standardized regression weights assigned for each

step of the regression equation

93

(HlO) demonstrated that performance differences on the psychomotor task attributable to
polychronic orientation (favoring monochronic orientation) decreased as levels of self-
monitoring increased providing some support for the hypothesis (see Figure 3).

Cﬂfarving out age, test motivation, perceived gee vaLliditv.and general cognitive
aim. Partial correlations between variables are presented in Table 8 after controlling for
variance accounted for by the applicant’s age, level of test motivation, perceived face
validity, and general cognitive ability. In examining the matrix, signiﬁcant negative
relationships were found only between comparative anxiety (HlO) with perceptual speed
such that lower levels of anxiety led to completing more items. However, signiﬁcant
negative relationships were found between conscientiousness (H3), pace (H6), and circadian
rhythm match (H9) with psychomotor speed such that higher levels of these constructs were
related to less time required to complete the task and a signiﬁcant positive relationship was
found between polychronic orientation (H3) and psychomotor speed such that someone who
oriented toward multiple tasks took more time to complete the task, providing some initial
support for those hypotheses.

Next, stepwise hierarchical regression was utilized on perceptual speed and the
results are presented in Table 9. In the ﬁrst step, age, test motivation and perceived face
validity were entered as a block resulting in a signiﬁcant R2 of .09 (p < .05). At this step,
age displayed a signiﬁcant negative relationship such that the younger the applicant the
more items completed on the Data Comparison test. Perceived face validity also displayed
signiﬁcant a positive relationship with perceptual speed such that people who thought the

tests were more face valid completed more items on the Data Comparison test. At the

94

 

 

160/

1504 i

140.

131.05

130;

 

   

l

1204 p .250 Polychronlcj

l
I I I +ZSD Polychronic]
1104 l

100.

90;

Psychomotor Speed (Average Time-seconds)

 

-ZSD +ZSD

SELF-MON ITORING

 

 

 

F'gge 3:
Interaction Between Polychronic Oriengtion and Self-Monitoring on Psychomotor Speed

95

Table 8
Summg of Partial Correlations Investigating Relationships to Measures of Speed

Controlling for Age, Test Motivatiop, Perceived Face Validig, and Cogjtive Abilig

 

 

Variables 01 02 03 04 05 06 07 08 09
0 Impatience/ (.63)
1 Irritability
0 Conscientiousness -.18* (.60)
2
0 Polychronic .10* -.13* (.81)
3 Orientation
0 Pace .04 .11* -.25* (.67)
4
0 Self-Monitoring -.21* .10* .07* -.06* (.62)
5 _
0 Circadian Rhythm .05 -.03 .06 .02 .03 «-
Match
6
0 Comparative .28* -.02 .00 -.04 .01 .01 (.85)
7 Anxiety
0 Perceptual Speed .01 .00 -.05 .03 .00 .00 -.14* «-
8
O Psychomotor .04 -.08* .08* -.07* -.04 -.09* .05 -.08* ---
Speed (Average
9 .
Time)
*p< .05
n=853

NOTE: Reliabilities are displayed in the diagonal

96

Table 9
13_e_sult of Simple Regression Predicting Perceptual Speed Controlling for Age, Test
Motivatiop, Perceived F acelalidity, and GenerMognitive Ability

 

 

. Step Final 2 2
Variable Beta Beta F R R AR
Step 1
Age -.27* -.18
Test Motivation .04 .02
Perceived Face Validity .09* -.02 29.05 .30 .09*

Step 2
General Cognitive Ability .36* .29 55.38 .46 .21* .12*
Step 3
Impatience/Irritability .06 .06
Conscientiousness .00 -.29
Polychronic Orientation -.05 .16
Pace .01 .01
Circadian Rhythm Match .00 .00
Comparative Anxiety -.16* -.16
Self-Monitoring .02 -.17 22.43 .48 .23* .02*
Step 4
Self Monitoring X .42 .42
Conscientiousness
Self-Monitoring X Polychronic -.24 -.24 19.05 .48 .23* .00
Orientation
*p < .05
n=853

NOTE: Step Beta weights refer to the standardized regression weights assigned for each
step of the regression equation

97

second step, general cognitive ability (H1) was entered resulting in a signiﬁcant AR2 of .12
(p < .05), such that individuals with higher levels of general cognitive ability completed
more items on the perceptual speed test. At the third step, impatience/irritability,
conscientiousness, polychronic orientation, pace, circadian rhythm match, comparative
anxiety, and self monitoring were entered as a block resulting in a signiﬁcant AR2 of .02 (p
< .05). At this step, only comparative anxiety (HlO) displayed a signiﬁcant negative
relationship with perceptual speed such that those individuals with lower levels of anxiety
completed more items on the Data Comparison test. At the fourth step, the interactions
between self-monitoring and conscientiousness (H9) and self-monitoring and polychronic
orientation (HI 0) failed to provide a signiﬁcant change in R2, providing little support for
these hypotheses.

Next stepwise hierarchical regression was utilized on psychomotor speed and the
results are presented in Table 10. In the ﬁrst step, age, test motivation and perceived face
validity were entered as a block resulting in a signiﬁcant R2 of .16 (p < .05). At this step,
age displayed a signiﬁcant positive relationship such that the older the applicant the more
time was required to complete the psychomotor task. At the second step, general cognitive
ability (Hl) was entered resulting in a signiﬁcant AR2 of .10 (p < .05) such that individuals
with higher levels of general cognitive ability required less time to complete the
psychomotor task. At the third step, impatience/irritability, conscientiousness, polychronic
orientation, pace, circadian rhythm match, comparative anxiety, and self monitoring were
entered as a block resulting in a signiﬁcant AR2 of .02 (p < .05). At this step, only circadian

rhythm match (H9) displayed a signiﬁcant negative relationship with psychomotor speed

98

Table 10

Result of Simple Reggession Predicting Psychomotor Speed Controlling for Age, Test
Motivation, Perceived Face Validity, and GeneraLCognitive Ability

 

 

 

. Step Final 2 2
Variable Beta Beta F R R AR
Step 1
Age .38* .32
Test Motivation -.05 .00
Perceived Face Validity -.02 .08 51.98 .39 .16*

Step 2
General Cognitive Ability -.34* -.30 73.53 .51 .26* .10*
Step 3
Impatience/Irritability .01 .01
Conscientiousness -.06 -.19
Polychronic Orientation .06 .69
Pace 1 -.04 -.04
Circadian Rhythm Match -.08* -.08
Comparative Anxiety .04 .04
Self-Monitoring .04 .08 29.29 .53 .28* .02*
Step 4
Self Monitoring X .18 .18
Conscientiousness
Self-Monitoring X Polychronic -.69 -.69 25.17 .53 .28* .00
Orientation
*p < .05
n=853

NOTE: Step Beta weights refer to the standardized regression weights assigned for each
step of the regression equation

99

mil

COII

CO:

it

[57‘

such that those individuals whose circadian rhythm style better matched the time they
completed the testing phase required less time to complete the measure of psychomotor
speed. At the fourth step, the interactions between self-monitoring and conscientiousness
(H9) and self-monitoring and polychronic orientation (H10) failed to provide a signiﬁcant
change in R2 providing no support for these hypotheses. A summary of the various analyses
conducted to investigate the relationships of factors to measures of speed is displayed in
Table 11 and will be discussed more fully in the Discussion section.

In addition, the relationship between achievement (H4) and number of errors
completed was also examined. Results found that the zero-order intercorrelation between
achievement and percentage correct (r = .06), the partial correlation controlling for age,
test motivation, and perceived face validity (r = -.02), and the partial correlation
controlling for age, test motivation, perceived face validity, and cognitive ability (r = -
.03) all were nonsigniﬁcant (p > .05), thereby providing no support for the fourth
hypothesis that achievement would be negatively related to proportion of correct items in
the perceptual speed measure.

Incremental Validity of Measures of Speed

The second major purpose of this research was to investigate the incremental
validity of these measures of speed over that of the relationship between general
cognitive ability and job performance. Table 12 summarizes the descriptive statistics and
zero-order correlations between variables. General cognitive ability displayed signiﬁcant
relationships with all measures of job performance with the exception of total assemblies

completed, thus providing support for hypothesis 11. Individuals with higher levels of

100

Table 11

Hypothesis Summag Investigating Relationships to Measures of Speed

 

Zero-Order Correlations and Hierarchical Regression

 

 

 

 

 

 

 

 

 

 

 

Without Covariates
Table 2 Table 3 Table 4

Perceptual Psychomotor
H2: Cognitive Ability Support N/A N/A
H2: Impatience/Irritability No Support Support No Support
H3:Conscientiousnes Support No Support Support
H5: Polychronic Orientation Support Support No Support
H6: Pace Support No Support Support
H7: Self-Monitoring X N/A No Support No Support
Conscientiousness
H8: Self-Monitoring X N/A No Support No Support
Polychronic Orientation
H9: Circadian Rhythm Match No Support No Support No Support
H10: Comparative Anxiety Support Support Support

 

 

 

 

101

 

Table 11 (cont’d)

 

Partial Correlations and Hierarchical Regression
Covarying Out Age Test Motivation, and Perceived

 

 

 

 

 

 

 

 

 

 

 

Face Validity
Table 5 Table 6 Table 7
Perceptual Psychomotor

H1: Cognitive Ability N/A N/A N/A
H2: Impatience/Irritability No Support Support No Support
H3:Conscientiousnes Support Support Support
H5: Polychronic Orientation Support Support Support
H6: Pace Partial Support No Support Support

(Psychomotor

Only
H7: Self-Monitoring X N/A No Support No Support
Conscientiousness
H8: Self-Monitoring X N/A No Support Support
Polychronic Orientation
H9: Circadian Rhythm Match Partial Support No Support No Support

(Psychomotor

Only
H10: Comparative Anxiety Support Support Support

 

 

 

 

102

 

Table 11 (cont’d)

 

Partial Correlations and Hierarchical Regression
Covarying Out Age, Test Motivation, Perceived Face
Validity, and Cognitive Ability

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 8 Table 9 Table 10
Perceptual Psychomotor
H1: Cognitive Ability N/A Support Support
H2: Impatience/Irritability No Support No Support No Support
H3:Conscientiousnes Partial Support No Support No Support
(Psychomotor
Only)
H5: Polychronic Orientation Partial Support No Support No Support
(Psychomotor
Only)
H6: Pace Partial Support No Support No Support
(Psychomotor
- Only)
H7 : Self-Monitoring X N/A No Support No Support
Conscientiousness
H8: Self-Monitoring X N/A No Support No Support
Polychronic Orientation
H9: Circadian Rhythm Match Partial Support No Support Support
(Psychomotor
Only)
H10: Comparative Anxiety Partial Support Support No Support
(Perceptual
Only)

 

103

 

Table 12

Summary of Means, Standard Deviations, Reliabilities, and Zero-Order Intercorrelations

Investigating the Incremental Validity of Speed

Variables

Mean

SD

01

02

 

03

O4

05

06

07

08

 

01

02

03

O4

05

06

07

08

General
Cognitive
Ability
Perceptual
Speed (Data
Comparison)

Psychomotor
Speed
(Average
Time)

Total
Assemblies

Proportion of
Defective
Parts
Problem
Solving
Rating

Team Skills
Rating

Work
Orientation
Rating

29.35

33.06

113.20

8.24

.088

4.98

5 .45

5.29

4.23

6.81

11.21

1.24

.049

1.15

1.17

.87

(.88)

.17*

-.08

.01

-.l7*

.16*

.12*

.19*

-.12*

.22*

.05

.05

.08

.ll*

(.60)

—.20*

.18*

-.ll*

-.04

-.15*

.12*

-.08

.00

-.O6

.00

.04

-.05

(.48)

.39*

.27*

(.56)

.31*

(.40)

 

*p < .05
n=53l
NOTE: Reliabilities are displayed in the diagonal

104

general cognitive abilityhad a lower proportion of defective parts, and higher ratings of
problem solving, team skills, and work orientation. Perceptual speed displayed signiﬁcant
validity with two measures of performance, total ntunber of assemblies completed and work
orientation. However, these two measures were ones that most emphasized demonstrating
effort, working under pressure, and meeting deadlines. Psychomotor speed displayed
signiﬁcant validity with all the measures of job performance with the exception of team
skills. All but problem solving were dimensions that most emphasized demonstrating
effort, working under pressure, and meeting deadlines.

Partial correlations between variables are presented in Table 13 after controlling
for the variance accounted for by general cognitive ability. In examining the matrix,
perceptual speed only demonstrated signiﬁcant positive relationship with total units built
such that individuals who completed more items on the Data Comparison test were able
to build more assemblies. However, perceptual speed failed to display signiﬁcant
relationships with proportion of defective parts and work orientation, thereby only
providing partial support for the incremental validity of perceptual speed. Psychomotor
speed continued to display signiﬁcant relationships with all the performance measures with
the exception of team skills. In addition, the relationship between psychomotor speed and
the rating of problem solving displayed the lowest validity (r = -.10, p < .05) of the four
signiﬁcant validities. Therefore validity was maintained even after variability incognitive
ability was controlled with the strongest relationships being exhibited in performance

dimensions related to effort and discipline.

105

Table 13

Summag of Partial Correlations Investigating the Incremental Validigy of Speed
Controlling for Cogm'tive Ability

Variables Ol 02 03 04 05 O6 07

 

Perceptual Speed ---
(Data

Comparison)

Psychomotor -.11* (.60)
Speed (Average

Time)

Total Assemblies .22* -.20* ---

Proportion of .08 .17* .12* «-
Defective Parts

Problem Solving .02 -.10* -.08 .03 (.48)
Rating

Team Skills .06 -.03 .00 .07 .38* (.56)
Rating

Work Orientation .08 -.14* -.07 -.02 .25* .30* (.40)
Rating

\IOOOMOAOWONOF‘O

 

*p < .05
n=53 1
NOTE: Reliabilities are displayed in the diagonal

106

Next, hierarchical multiple regression were conducted on each of the ﬁve measures
of job performance. First, regressions were performed on the three measures where
signiﬁcant incremental validity was expected (proportion of defects, total assemblies built,
and work orientation). Second, regressions were performed on two measures where
signiﬁcant incremental validity was not expected (problem solving, team skills).

Relationships with perfonpance dimensipns related to effort and discipline. The ﬁrst
stepwise hierarchical multiple regression examined the incremental validity in the prediction
of proportion of defects and is displayed in Table 14. In the ﬁrst step, general cognitive
ability resulted in a signiﬁcant R2 of .03 (p < .05). At this step, general cognitive ability
displayed a signiﬁcant negative relationship such that the higher level of general cognitive
ability, the lower the proportion of defects in built units, thus providing support for
hypothesis 11. At the second step, perceptual speed and psychomotor speed were entered
resulting in a signiﬁcant AR2 of .04 (p < .05) At this step, perceptual speed displayed
signiﬁcant positive relationships with proportion of defects such that higher levels of
perceptual speed resulted in a higher proportion of defects. Also the time required to
complete the psychomotor task also displayed a signiﬁcant positive relationship with
proportion of defects such that people who took less time to complete the psychomotor
tasks also had a lower proportion of defects in their built units, providing some support
for the twelfth hypothesis.

The second stepwise hierarchical multiple regression examined the incremental
validity in the prediction of total number of assemblies and is displayed in Table 15. In the

ﬁrst step, general cognitive ability resulted in a nonsigniﬁcant R2, thus providing a lack of

107

Table 14

Result of Simple Reggssion Predicting Proportion of DefecLs Built Controlling for

Cogm'tive Abilig;

 

 

, Step Final 2 2
Varrable Beta Beta F R R AR
Step 1
Cognitive Ability -.17* -.18 15.95 .17 .03*

Step 2

Psychomotor Speed .18* .18

Perceptual Speed .10* .10 12.77 .26 .07* .04*
*p < .05
n=531

Note: Step Beta weights refer to the standardized regression weights assigned for each step

of the regression equation

108

Table 15
Result of Simple Regrgession Predicting Total Assemblies Built Controlling for Cogpjtive

Abilig

 

 

Variable Beta Fm” F R R2 AR2
Beta

Step 1

Cognitive Ability .01 -.04 .073 .01 .00

Step 2

Psychomotor Speed -.18* -. 18

Perceptual Speed .21* .21 15.68 .29 .08* .08*
*p < .05
n=531

Note: Step Beta weights refer to the standardized regression weights assigned for each step
of the regression equation

109

support for hypothesis 11 in relationship to the prediction of total number of assemblies
built. At the second step, perceptual speed and psychomotor speed were entered resulting in
a signiﬁcant AR2 of .08 (p < .05) At this step, perceptual speed displayed signiﬁcant
relationships with total number of assemblies such that higher levels of perceptual speed
resulted in a higher number of assemblies built. Also the time required to complete the
psychomotor task displayed a negative relationship with total number of assemblies built
such that pe0ple who took less time to complete the psychomotor tasks were able to build
more assemblies, providing some support for the twelfth hypothesis.

The third stepwise hierarchical multiple regression examined the incremental
validity in the prediction of work orientation and is displayed in Table 16. In the ﬁrst
step, general cognitive ability resulted in a signiﬁcant R2 of .04 (p < .05). At this step,
general cognitive ability displayed a signiﬁcant positive relationship such that the higher
level of general cognitive ability, the higher the rating of work orientation, providing
support for hypothesis 11. At the second step, perceptual speed and psychomotor speed
were entered resulting in a signiﬁcant AR2 of .02 (p < .05) At this step, perceptual speed
did not display a signiﬁcant relationship with the rating of work orientation.
Psychomotor speed displayed a signiﬁcant negative relationship with the rating of work
orientation such that people who took less time to complete the psychomotor tasks also
had a higher rating of work orientation, providing some support for the twelfth
hypothesis.

Relationships with perforraance dimensions unrelated to effort and discipline. The

fourth stepwise hierarchical multiple regression examined the incremental validity in the

110

Table 1

Result of Simple Regzession Predicting Performance Rating of Work Orientation

Controlling for Cogr_11_'tive Abilig

Final

 

 

. Step 2 2
Vanable Beta Beta F R R AR
Step 1
Cognitive Ability .19* .16 18.99 .18 .04*

Step 2 ’

Psychomotor Speed -. 13* -. 13

Perceptual Speed .06 .06 10.61 .24 .06* .02*
*p < .05
n=531

Note: Step Beta weights refer to the standardized regression weights assigned for each step

of the regression equation

111

prediction of problem solving and is displayed in Table 17. In the ﬁrst step, general
cognitive ability resulted in a signiﬁcant R2 of .03 (p < .05). At this step, general
cognitive ability displayed a signiﬁcant positive relationship such that the higher level of
general cognitive ability, the higher the rating of problem solving, providing support for
hypothesis 11. At the second step, perceptual speed and psychomotor speed were entered
resulting in a signiﬁcant AR2 of .01 (p < .05) At this step, perceptual speed did not
display a signiﬁcant relationship with the rating of problem solving. Psychomotor speed
displayed a signiﬁcant relationship with the rating of problem solving such that the
peOple who took less time to complete the psychomotor tasks also had a higher rating of
problem solving.

The ﬁfth stepwise hierarchical multiple regression examined the incremental validity
in the prediction of team skills and is displayed in Table 18. In the ﬁrst step, general
cognitive ability resulted in a signiﬁcant R2 of .01 (p < .05). At this step, general cognitive
ability displayed a signiﬁcant positive relationship such that the higher level of general
cognitive ability, the higher the rating of work orientation, thus providing support for
hypothesis 11. At the second step, perceptual speed and psychomotor speed were entered
resulting in a nonsigniﬁcant ARZ. In general, changes in R2 were signiﬁcant for the step
involved with speed only for those dimensions hypothesized to involve effort and discipline
(i.e., proportion of defects, total assemblies built, and work orientation rating). Five
signiﬁcant differences comparing incremental validity (see Table 12) for effort-based versus
non-effort-based correlations by speed measure (see Table 19). In this table, tests of

signiﬁcance were performed between each row partial correlation and each column partial

112

Table 17

Result of Simple Regression Predicting Performancﬂating of Problem Solving Controlling

 

 

 

for Cognitive Ability
. Step Final 2 2
Vanable Beta Beta F R R AR
Step 1
Cognitive Ability .16* .15 13.96 .16 .03*
Step 2
Psychomotor Speed -.10* -.10
Perceptual Speed .01 .01 6.48 .19 .04* .01
*p < .05
n=53l

Note: Step Beta weights refer to the standardized regression weights assigned for each step

of the regression equation

113

Table 18

Result of Simple Regression Predicting Performance Ratig of Team Skills Controlling for
Cogm'tive Abilig

 

 

Step Frnal F R R2 ARZ

 

 

Vanable Beta Beta

Step 1

Cognitive Ability .12* .10 7.49 .12 .01*

Step 2

Psychomotor Speed -.02 -.02

Perceptual Speed .06 .06 3.33 .14 .02* .01
*p < .05
n=531

Note: Step Beta weights refer to the standardized regression weights assigned for each step
of the regression equation

114

Table 19

Signiﬁcant Differences Incremental Validity Coefﬁcients for Effort-Based versus Non-
Effort Based Peﬁonmce Dimensionsl

 

 

 

 

 

 

 

 

 

 

 

Perceptual Speed Proportion of . . .

Partial Defective Parts Totairtssgmnbhes Work (ingrgratron
Correlations (r = .08) ' (r -

Team Skills n s

(r = .06) ' ° s.d. (t = -2.6, p < .05) us.

Problem Solving n s n

(r = .02) ° ' s.d. (t = -3.2, p < .05) .s.

Psychomotor Proportion of . . .

Speed Partial Defective Parts T031258??? hes Work Texan”

Correlations (r = .17) " (1' '- )

Team Skills

(r = -03) s.d. (t = -2.4, p < .05) s.d. (t = 2.8, p < .05) s.d. (t = 2.2, p < .05)

Problem Solving n s n s

(r= -.10) . . . . n.s.

 

 

 

 

 

 

n.s. = no signiﬁcant difference (p > .05)
s.d. = signiﬁcant difference (p < .05)

' based on formula by Steiger (1980) to test the signiﬁcance of the difference between
dependent r’s:

t = (5,, — rvy) \l[(n-1)(1 + rxv)

 

\l[2((n-l)/(n-3)) I R I + 804,03]

wherer=(rxy+rVy)/2and |R|=1—r2,,y—r2,,y —12v,,+2r,,yrvyrxv

115

 

correlation. These tests represent twelve separate tests of Hypothesis 12. These ﬁve
signiﬁcant differences provide additional support for Hypothesis 12.

Correction for criterion unreliabilig and restriction of range. Based on interrater
reliability estimates of the ratings of work orientation, team skills, and problem solving and
the selection ratio of .62 between the testing and assessment phase, various corrections were
made to the partial intercorrelations (see Table 13) of perceptual and psychomotor speed
with the various performance measures (see Table 20). Perceptual speed estimates for
related dimensions (work orientation, proportion of defects, and total assemblies) ranged
from .09 to .27 whereas estimates for unrelated dimensions (team skills, problem solving)
ranged from .03 to .09. Psychomotor speed estimates were higher for proportion of defects
(r = .28), total assemblies (r = -.33), and work orientation (r=-.36) and lower for unrelated
dimensions such as team skills (r = -.06) and problem solving (r= - .23), providing
additional evidence to the types of performance dimensions most likely to be related to

measures of speed.

116

mane S 223 e 8.258 £322 sense wee: H3.82

 

 

 

 

 

 

 

 

 

 

 

Emu:
mo. V a...
museum
on: Q...- *3 .- 8. 2. mo. 5:8:on 833
8.- 3.- 8.- so. mo. 8. 2:2 23m 889
MN. 3 .- 2.2.- 8. no. no. mere”— wEZom 350$
mm.- 8: seat R. 8: Km. mom—£822 30,—-
mm. 3: en ﬂ. mo. S: we. $330.3 5382A
Acmemm Aowsam
Co 88582 33285 as 88.53 3:332:
a. 3332:: 3.33-00 29669 3:3 a. bang—2:3 532-200 .321; Eng 2:32
actuate €223 33m 362825 8.3300 206:2; Ema-am 38228.5 ooﬁgotom new
56:5» 33m 36280 £21; 323 36280
382.50 38280
82m 82m
Sueﬁegimm 3533.8.—

 

 

 

 

 

m..2:. 32 3.13. .E. ..o...311 m :2. vacuum we 3.5302 coca/“om basin 9,2300 3556:”; 338205 382.50
cm 23,—-

117

DISCUSSION

The purpose of the present study was to examine both the relationships of a
variety of variables to measures of speed and to examine the incremental validity of
speed on broader measures of job performance. The discussion is divided up into three
sections. The ﬁrst section provides a summary of the ﬁndings and the implications of the
results. The second section highlights some of the limitations of the study and the impact

of those limitations on the results. The third section addresses future research directions.

Summgy and Implications of Major Finding§

 

Relationships with measures of speed. Table 11 presents a summary of the
evidence examining the relationships of cognitive and noncognitive factors to measures
of speed and this summary will follow both that evidence as well as the order of the
results. Three covariates used in the study but not direct focus of the hypotheses included
age, test motivation, and perceived face validity. Of these three, age displayed a strong
relationship with an individual’s speed, both perceptual and psychomotor. This ﬁnding
was consistent with previous research (e.g., Birren & Fisher, 1995). Even though a large
variation in terms of age wasn’t expected with a predominantly younger applicant group,
the strong effect continued to be present. Test motivation displayed little effect on
measures of speed, probably due to the fact that most applicants were highly motivated to
do well in the testing phase (Mean = 4.47, SD = .43). Perceived face validity added
incremental prediction over age with the measure of perceptual speed such that the more
the individual thought the testing phase was face valid, the more items completed on the

perceptual speed measure (Data Comparison Test). However, such a relationship may be

118

stronger with the perceptual speed measure than the psychomotor speed measure because
the position being applied for is in manufacturing where psychomotor skills are likely to
be deemed more important. The effort applied in the paper-and-pencil test may have
been reduced because individuals may have viewed the perceptual speed measure
containing text as somewhat irrelevant.

The ﬁrst hypothesis of the study postulated that general cognitive ability would be
related to measures of speed and the research study found strong support across both
speed measures for such an effect. Such a relationship was not only pervasive across
multiple stages of analysis, but is well documented in the research literature (i.e., Carroll,
1993; Vernon, 1987).

The next series of hypotheses focused upon noncognitive factors in relationship to
measures of speed. Going in order of hypotheses, the ﬁrst factor examined was
Impatience/Irritability (hypothesis 2). It was hypothesized that individuals who had a
higher level of Impatience/Irritability would subsequently have higher levels of arousal
and display higher levels of speed. However, results provided limited support for the
construct in relationship to measures of speed. Using hierarchical regression, Impatience/
Irritability added some signiﬁcant incremental prediction in relationship to measures of
perceptual speed, both without the covariates and covarying out age, perceived face
validity, and test motivation. However, bivariate relationships in those two situations
(zero-order and partial correlations) did not demonstrate any signiﬁcant relationships
until variance was accounted for by the other noncognitive factors in the regression
equations. The predictive setting of the study could have had an effect on this

relationship. Applicants attempting to do well in the selection system might not want to

119

claim to be impatient and irritable which may have resulted in a slightly lower mean and
standard deviation (Mean = 2.27, SD = .53) than has been found in non-selection studies
using undergraduates (e.g., Conte et al, 1995).

The third hypothesis investigated the relationship between conscientiousness and
measures of speed. It was hypothesized that individuals who were higher in
conscientiousness would devote more effort over time to completing the measures of
speed. Results provide some reasonable support for this hypothesis, primarily for
measures of psychomotor speed where signiﬁcant bivariate, and partial correlation
relationships were discovered both with and without covariates. In addition, signiﬁcant
incremental prediction was found when age, perceived face validity, and test motivation
were controlled for and other noncognitive variables were included in the regression
equation. However, this relationship was nonsigniﬁcant once cognitive ability was added
to the regression equation even though the partial correlation between psychomotor speed
and conscientiousness was signiﬁcant. Relatedly, the fourth hypothesis examined if
achievement would be related to number of errors on the perceptual speed measure (the
psychomotor speed measures had virtually no errors) such that achievement-oriented
individuals would think that success was based on speed and, therefore, trade accuracy
for speed. However, no relationship was discovered either for zero-order correlations or
after controlling for the variance attributed to covariates. In terms of the prediction of
perceptual and psychomotor speed by achievement, a similar series of analyses were run
as those listed in Table 10 but achievement was added as a subsequent step after the
group of noncognitive variables were entered. Only one signiﬁcant ﬁnding was

discovered: a negative signiﬁcant relationship between achievement and psychomotor

120

speed after controlling for conscientiousness (r = -.09, p < .05). All other analyses failed
to provide support that achievement would add anything to the prediction of measures of
speed above that of a general measure of conscientiousness.

The ﬁfth hypothesis pr0posed that individuals who held a more polychronic
orientation towards activities would have lower levels of speed than those who had a
monochronic orientation. People with a monochronic orientation were proposed to have
higher levels of speed because they could focus their effort to completing the task at hand
rather than an orientation where effort is often divided among multiple activities which is
more characteristic of individuals with a polychronic orientation. The hypothesis
received reasonable support such that polychronic orientation displayed signiﬁcant
negative relationships across multiple lines of analysis including zero-order correlations,
hierarchical regressions without any covariates (perceptual speed only), and when the
covariates of age, perceived face validity, and test motivation were accounted for (both
partial-correlations and regressions). When variance associated with cognitive ability
was accounted for, only the partial-correlation between polychronic orientation and
psychomotor speed was signiﬁcant. However, overall evidence tended to support the
notion that the orientation (single task versus multiple task) generally had a fairly stable
polychronic relationship with measures of speed and could serve as a meaningful
variable.

The sixth hypothesis proposed that an individual whose orientation towards
activities was faster paced would have higher levels of speed through higher levels of
arousal. Analyses, found that pace displayed signiﬁcant positive relationships with

measures of speed, particularly psychomotor speed. This bivariate relationship with

121

psychomotor speed continued after controlling for age, test motivation, and perceived
face validity, and it continued when cognitive ability was controlled for as well.
However, in hierarchical regressions with the other noncognitive variables, pace failed to
account for any signiﬁcant incremental efﬁciency in predicting measures of speed. So,
while a relationship seems to exists between self-reports of pace and measures of speed,
its incremental prediction in light of other noncognitive factors proved weaker.

The seventh and eighth hypotheses examined self-monitoring as a moderator
between the relationship of conscientiousness with measures of speed (H7) and
polychronic orientation with measures of speed (H8). Hypothesis seven proposed that
the relationship between conscientiousness and measures of speed would be strongest for
individuals who had lower levels of self-monitoring. At higher levels of self-monitoring,
individuals with lower levels of conscientiousness would change their orientation to meet
the speed demands of the task, thereby reducing the relationship between
conscientiousness and speed. However, no relationships were discovered in any analyses
to support this hypothesis. Hypothesis eight proposed a similar relationship occurring
with self-monitoring acting as a moderator between the polychronic orientation and speed
such that the relationship between polychronic orientation and measures of speed would
be strongest for individuals who had lower levels of self-monitoring. At higher levels of
self-monitoring, individuals with a more polychronic orientation would change their
orientation to meet the speed demands of the task, thereby reducing the relationship
between polychronic orientation and speed. Results provided some evidence for the
hypothesis. After controlling for age, test motivation, and perceived face validity, the

interaction added signiﬁcant incremental prediction (Figure 3) that matched the

122

hypothesis when psychomotor speed was the dependent variable. This interaction was
nonsigniﬁcant but approached signiﬁcance once cognitive ability was added as a
covariate to the hierarchical regression equation (p = .054) predicting psychomotor speed.
Perhaps the relationship was stronger with polychronic orientation than with
conscientiousness due to impression management constraints. As such, the items
measuring conscientiousness may have been more socially desirable in terms of what
response would most likely result in employment. Items measuring monochronic (tend to
one task before starting another) and polychronic (tend to do multiple activities) may
have been less transparent, leading to more accurate responding.

Hypothesis nine proposed that individuals whose circadian rhythm style better
matched the actual time they took the speed measures would perform at higher levels of
speed due to higher levels of arousal. Little support was found for this relationship until
the various covariates were added to the prediction of speed. Both when controlling for
age, test motivation, and perceived face validity, and when controlling for cognitive
ability as well, relationShips began to emerge between the matching of circadian rhythms
and actual testing time with psychomotor speed. Both partial correlations displayed
signiﬁcant relationships with psychomotor speed and after controlling for all covariates,
the variable added incremental prediction to the other noncognitive factors in predicting
psychomotor speed.

Hypothesis ten proposed that the comparative anxiety felt by individuals in the
testing phase would be negatively related to measures of speed due to a deﬂection of
effort away from the task at hand toward comparing one’s efforts to others. Results

provided ample support for this hypothesis across multiple steps in the analysis. The

123

only area where support was not found was in relationship to psychomotor speed when
all the variance attributed to the covariates (age, test motivation, perceived face validity,
and cognitive ability) was accounted for.

In summary, all the hypotheses with the exception of the interaction between self-
monitoring and conscientiousness received some support in the analyses. However,
certain variables tended to display more consistent and more incremental evidence across
various analyses. In terms of covariates, age and general cognitive ability continued to
display their well-documented relationship with measures of speed. In terms of the
noncognitive relationships, the results demonstrated that the hypotheses involving
conscientiousness, polychronic orientation, and comparative anxiety, variables
inﬂuencing the measurement of speed primarily through a proposed effort mechanism,
tended to have the most stable and most consistent relationships with measures of speed.
In particular, relationships with psychomotor speed tended to be more prevalent than
those with perceptual speed even though the Multiple R examining the prediction of
speed with noncognitive predictors (.34 perceptual (Table 3) & .30 psychomotor (Table
4)) and the correlation with general cognitive ability (.41 perceptual & -.41 psychomotor)
tended to be similar. Perhaps, the psychomotor speed measure with its two trials and lack
of errors provided a more robust measure in relationship to noncognitive relationships
whereas perceptual speed relationships were dominated by its relationship with
comparative anxiety (r = -.31).

Incremental validity of measures pfapeed over cognitive ability. The second
major purpose of the research was to examine whether or not measures of speed were

related to broader measures of job performance, particularly job performance dimensions

124

emphasizing behaviors such as demonstrating effort, maintaining discipline, etc. This
research hypothesized that speed would provide signiﬁcant prediction beyond the strong
association between cognitive ability and job performance which is often reported to
encompass any predictive power that speed may have (e. g., Ree & Earles, 1992). The
analyses conducted to examine this hypothesis used ﬁve measures of performance.
Speed was hypothesized to provide signiﬁcant incremental validity over general
cognitive ability in three of the job performance measures that more emphasized
dimensions involving effort, and discipline: Number of assemblies built, the number of
errors built into the assemblies (quality), and a rating on the dimension of work
orientation. The other two measures, ratings of problem solving and team skills, were
examined to see if the null hypothesis was found since they did not emphasize such
dimensions involving demonstrating effort. First, the 11th hypothesis proposed that
general cognitive ability would have a signiﬁcant relationship with the job performance
measures and it did with all but number of assemblies built, providing support for this
hypothesis consistent with previous literature (e.g., Hunter and Hunter, 1984).

Hypothesis 12 proposed that speed would add incremental validity over cognitive
ability and was examined for the ﬁve measures of performance in two ways. First, partial
correlations between each measure of speed (perceptual and psychomotor) and the
measures of performance were examined after controlling for cognitive ability. Second,
hierarchical regression was used with each measure of performance to control general
cognitive ability prior to examining the incremental validity of the two speed measures
when added as a group. In terms of partial correlations, results supported the 12th

hypothesis. Perceptual speed displayed its highest correlations with the three

125

hypothesized dimensions (.22, .08, .08) and its lowest correlations with the two measures
proposed to be most unlikely to affected by speed (.02, .06). However, only the
incremental validity with total assemblies proved to be signiﬁcant. In terms of
psychomotor speed, the pattern of correlations was consistent with perceptual speed in
that the dimensions proposed to receive incremental validity from measures of speed
were higher (-.20, .17, -. 14) than those not proposed to receive incremental validity (-.03,
-. 10). In addition, all three dimensions involving effort and discipline were signiﬁcant.
In terms of hierarchical regression results, similar results were obtained with
signiﬁcant incremental validity for the step where perceptual and psychomotor speed
were entered only occurring with total assemblies built, number of errors in assemblies,
and work orientation. In terms of individual patterns of relationships, psychomotor speed
added incremental validity to all three dimensions and perceptual speed added
incremental validity to total assemblies built and proportion of errors in assemblies. In
addition, psychomotor speed added incremental validity to the problem solving rating
even though the step as a whole failed to add to overall validity. The most unexpected
result occurred with the measure of perceptual speed displaying signiﬁcance in the
opposite direction in relationship to proportion of errors such that the higher one’s level
of perceptual speed, the higher the percentage of errors built into assemblies. This was
opposite to what was expected. While the relationship was small, one potential rationale
is that the perceptual speed test had a higher than desired rate of error (Mean item
difﬁculty = .8) in the testing phase and speed accuracy tradeoffs were transferred to the
next phase. However, the correlation between the number of perceptual speed items

completed in the testing phase was positive with percentage of items correct (r = .17, p <

126

.05) in the same phase. A further examination into the inﬂuence of error levels for a
speed test of this type could produce more meaningful results.

Overall, the incremental validity evidence for speed measures in the prediction of
job performance over general cognitive ability was strong, particularly for psychomotor
speed and particularly for dimensions where relationships were proposed to occur (total
assemblies built, proportion of errors in assemblies, and work orientation rating). A
potential argument against this result was that the performance dimensions only provided
a parallel test for the speed measures. While the argument can be made that total number
of assemblies is somewhat of a parallel measure, this would not explain the incremental
validity with the work orientation rating, which was virtually uncorrelated with total
number of assemblies. In addition, work orientation represents a more “typical” measure
of performance rather than a more “maximum” measure of performance such as total
assemblies built, yet it displayed some of the strongest relationships with the speed
measures.

PgacticaL and theoretical implications. An important hypothesis in the work by
Carroll (1993) was that speed was a viable construct and not just a measurement
byproduct of cognitive ability tests. He encouraged research that would continue to
demonstrate the separateness of the constructs and the current research demonstrates that
this statement appears to be true. Both incremental validity in the prediction of job
performance and relationships between speed and noncognitive factors were displayed in
the current research. Particularly stable relationships between speed and different types
of noncognitive measures including those associated with stable personality traits (e.g.,

conscientiousness), temporal orientations (e. g., monochronic orientation) and situational

127

factors (e.g., comparative anxiety) were also found. Primarily, these relationships were
proposed to occur through an increase in concentrated effort on the task at hand rather
than higher arousal levels. However, further research should be conducted to better
specify that it is effort that helps to mediate the relationship with speed.

Some research has even begun to explore the relationship of constructs across
three trait domains: personality, ability, and interests (Ackerrnan & Heggestad, 1997) in
search of sources of commonality. In particular, Ackerrnan and Heggestad (1997) have
proposed that a trait complex entitled Clerical/Convential show positive commonalities
among dimensions such as perceptual speed, control, and conscientiousness. In their
proposed research, the development of these individual differences may fall along causal
lines such that the successful performance of certain tasks may increase one’s level of
interest or personality. Future research can continue to explore the nature of relationships
among these constructs.

Effort is a critical component to examine in assessing the viability of speed as a
useful construct. Two of the three second-order factors of speed in Carroll's (1993)
research, psychomotor and perceptual speed, were utilized in the current research. Such
tasks required not only speed in completing task, but persistence to complete the task
over the time period required. As such, these speed tasks may be construed as
motivational tasks rather than measures of ability. However, the rank order of scores on
speed measures should be consistent in highly motivated situations (such as the current
research) as they are in less highly motivated situations. In highly motivated situations,
the level of the scores may just be higher. More simple measures of speed such as the

third second-order speed factor examined in Carroll's research, reaction time, may not

128

display similar relationships as those found in the current study, yet these measures may
represent even purer measures of speed than those used in the current research. However,
the maintenance of speed over time may be a more important predictor of overall job
performance. Whereas quickly reacting to a problem in a work setting is important, the
series of steps that follow that reaction and the speed at which those steps are completed
and maintained can become the critical component to success in the job situation.
Beyond the incremental validity of speed measures is a question of their practical
utility. Table 19 presents some corrected partial-correlations after controlling for
variance attributed to cognitive abilities and correcting for criterion unreliability and
range restriction. Correlations between psychomotor speed and total assemblies,
proportion of errors in assemblies, and work orientation rating were -.33, .28, and -.36
respectively. The levels of these relationships are not only statistically signiﬁcant, but
practically important in relevance to the prediction of job performance both with
measures that are considered indicies of maximum performance (total assemblies built)
and those that are considered measures of typical performance (work orientation rating).
In addition, the practical utility of the results slightly favors using measures of
speed that are more closely aligned with the deﬁnition of a pure speed measure, one in
which the difﬁculty of attempted items is near 1.0 (Anastasi, 1988). In the current
research, the measure of psychomotor speed tended to produce more favorable results
than the measure of perceptual speed. This may be due to an absence of errors in its
measurement. Perhaps establishing a measure where everyone ﬁnishes a simple task and
time is the criterion may provide less of an opportunity for other factors to inﬂuence the

measurement of speed, such as guessing. Completing a number of items completed in a

129

set time period may provide a higher chance of guessing, thereby raising the difﬁculty
level of the test. Therefore, choosing a more stable and pure measure of speed that still
requires displaying speed over a time period can have a positive impact on incremental
predictive efﬁciency.

Other practical implications of the results include the nature of the criterion,
namely job performance. In the current research setting, participants were keenly aware
that they were working "against the clock." Speed demonstrated incremental validity in
the prediction of performance in this setting, primarily for those dimensions expressing
the demonstration of effort. However, the current research doesn't proclaim that
measures of speed will be practically useful in every type of job. In those positions
where time pressure becomes salient, a measure of speed could provide additional
validity. Other positions where speed is not critical to completing job tasks, speed may
not provide incremental validity over other measures. However, given that job forecasts
predict a change from a mass production to a service economy (Heneman & Heneman,
1994), an increase pressure on providing goods and services in a timely manner
(Camevale, 1991), and an increased workload due to downsizing, measures of speed
could become more important in identifying successful candidates for an increased
number of positions.

Studv Limitation_s

The potential limitations with current research fall into four major categories: (a)

determination of causality, (b) global perceived face validity, (c) errors in perceptual

speed, (c) measurement of typical performance.

130

Determination of causalig. To move beyond investigating relationships of
variables with measures of speed would be to begin to make causal attributions.
However, sample constraints prevented the research from making strong causal
statements concerning the relationships because the completion of the survey containing
many of the measures occurred after the performance on the measures of speed. For
instance, the strongest noncognitive factor, comparative anxiety, may have been seen as
higher by individuals because they were attributing poor performance in the measures of
speed to outside pressure. Further research similar to that by Chan and colleagues (1997)
that examines test-taking motivation on subsequent performance could lend more support
to causal relationships.

Perceived face validity. The current measurement of perceived face validity also
could fall into the causality trap mentioned above. This could result from individuals
who perceive themselves doing poor on the test attributing their poor performance to lack
of face validity. In addition, another limitation of the current study was that only a global
measure of perceived face validity was obtained for the testing phase. A more speciﬁc
measure of perceived face validity strictly for the measures of speed might have been
more appropriate. Perhaps perceived face validity was higher for the psychomotor speed
measure given the manufacturing context than the perceptual speed measure and could
have produced different results. I

Errors in percpptual speed measure. The number of errors produced in the
perceptual speed measure were higher than desired. While a speed-accuracy tradeoff
effect did not seem to occur (i.e., number of items ﬁnished was positively correlated with

percentage of items correct), a cleaner measure of speed could have produced different

131

results. However, relationships with measures of general cognitive ability were virtually
identical with perceptual speed as they were with psychomotor speed even though
virtually no errors were produced in this measure of speed.

Measurement of typical performan_c_e, Maximum performance has three main
characteristics (Sacket et al, 1988): (a) there must be explicit awareness that one is being
evaluated, (b) there must be an awareness of and acceptance of instructions to maximize
effort, and (c) performance must be measured over a short enough time duration that the
perforrner’s attention remains focused on the accepted goal of maximum performance.
Whereas assessment center ratings over the length of the exercise are more typical than a
quick ten-minute assembly process, other performance measures could be construed as
better representing “typical” performance. That is, participants in this research were all
keenly aware that they were being evaluated throughout the assessment center task, hence
their performance may be more accurately characterized as maximal than, typical.
Therefore the general ability of the researcher’s results to prime measures of “typical”
performance could still use further investigation. In addition, the task in the current
design was fairly speciﬁc to manufacturing jobs and may not be fully applicable to every
different measure of job performance across different types of jobs.

Directions for Future Resﬁeagh

Future research could take a number of directions. First, subsequent research
could begin to better specify causality in terms of relationships of factors to speed. To
better deﬁne whether these factors lead to differential performance on measures of speed
through different samples and research designs could bring further clarity to these

relationships. While the current research situation likely ensured maximized motivation

132

to do well on the job, subsequent research samples could provide more opportunity to
explore these casualty issues. Other research situations might be devised to create more
variance on factors such as impatience/ irritability; it is unlikely that participants in this
study would admit to 11. Additional research should also measure and investigate the
mechanisms of effort and arousal. While the current research used these constructs as
theoretical mechanisms to explain various relationships with measures of speed,
subsequent research could build these concepts into research models and measure their
actual mediating effect. A possible model for such research includes that of Barrick and
colleagues (1993) who investigated the mediation effects of goal setting between the
conscientiousness-job performance relationship.

Second, research on the performance side of the incremental validity equation
could expand. Future research investigating the incremental validity of measures of
speed against performance measures that are further towards the “typical” end of the
maximum vs. typical continuum could provide additional evidence as to the practical
utility of using measures of speed to predict important organizational behaviors. In
addition, how speed is taken into account in a performance rating context could also use
further research. Marr and Stemberg (1987) found some evidence that people who
perform activities quickly are often perceived as being more intelligent. Research using
policy capturing methodology could provide some evidence as to how individuals weigh
speed in relationship to overall performance judgments. Perhaps research in impression
management could provide a useful framework for how subordinates “look busy” to

obtain higher ratings of performance.

133

Third, if speed has incremental validity and is considered an important aspect of
performance, various adverse impact concerns could come into play. For instance, age
had a dramatic effect on measures of speed such that if speed is an integral part of
organizational decision-making, older individuals could be adversely impacted. In
addition, research on other minority groups, either race or gender, could be conducted to
determine if speed could have an adverse impact in a similar way to how general

cognitive ability impacts Aﬁican Americans.

134

APPENDICES

APPENDIX A

Power Analyses

For each of the following power analyses, the desired power was ﬁxed at .80 and or was
ﬁxed at .05. Expected effect sizes were construed as "small" effect sizes (Cohen, 1988).
H7-H8:

H7-H8 tests the unique variance accounted for by the Self-Monitoring X
Personality Characteristics over and above the set of covariates, cognitive ability,
and noncognitive inﬂuences. A small AR2 of .03 was arbitrarily expected for
interaction effects. The expected R2 of .15 (Set A + Interactions) was arbitrarily
ﬁxed at a value of .15. Using Cohen and Cohen's (1983) formula for effect size:

r? = ARz/(l- R2)
= .03/(1- .15)
= .035

According to Cohen (1988), a f‘2 value of .035 is construed as a small effect size.
Cohen and Cohen's (1983) formula for required sample size n* is as follows:

n*=(L/ﬁ)+k+1

k refers to df for unique source of variance. The current proposal has k = 4. From
the table of L values in Cohen (1988), we have a L = 12.00. Therefore, the
required 11 is: *

n*=(L/f‘2)+k+l

= (12/.035) + 4 + 1
= 347.86

135

APPENDIX B

Impatience/Irritability Measure - Spence, Pred, & Helmreich (1987)

. When a person is talking and takes too long to come to the point, how often do you
feel like hurrying the person along? (very frequently to almost never)

. Typically, how easily do you get irritated? (extremely easily to not at all easily)
. Do you tend to do most things in a hurry (deﬁnitely true to not at all true)
. How is your “temper” these days? (very hard to control to I seldom get angry)

. When you have to wait in line such as at a restaurant, the movies, or the post ofﬁce,
how do you usually feel (accept calmly to feel very impatient and refuse to stay long)

136

APPENDIX C

Achievement Strivings Measure - Spence, Pred, & Helmreich (1987)
. How much does work “stir you into action?” (much less to much more than others)

. Nowadays, do you consider yourself to be (very hard-driving to very relaxed and easy
going)

. How would your best ﬁiends or others who know you well rate your general level of
activity? (too slow to very active, should slow down)

. How seriously do you take your work? (much more to much less than most)

. How often do you set deadlines or quotas for yourself in courses or other activities
(very often to almost never)

. Compared with my peers, the amount of effort I put forth is (much more to much
less)

. Compared with my peers, I approach life in general (much more to much less
seriously)

137

APPENDIX D

Monochronic versus Polychronic Orientation - Kaufman, Lane, & Lindquist (1991);
Tuttle (1996)

l. I prefer to do one thing at a time

2. When I work by myself, I usually work on one project at a time
3. I am comfortable doing several things at the same time

4. When I sit down to do work, I usually do one project at a time

5. I do not like to juggle several activities at a time

Scored on a 1 - 5 scale (strongly disagree - strongly agree)

138

APPENDIX E

Pace of Life - T uttle (1996)
1. I prefer to move quickly throughout the day
2. Fast moving activities are my favorite
3. Slow moving activities are my favorite
4. I do things faster than most people

5. I prefer to move slowly throughout the day

Scored on a 1 - 5 scale (strongly disagree - strongly agree)

139

10.

11.

12.

13.

14.

15.

16.

17.

18.

APPENDIX F

Self-Monitoring - Snyder (1974)
I ﬁnd it hard to imitate the behavior of other people
My behavior is usually an expression of my true inner feelings, attitudes, and beliefs

At parties and social gatherings, I do not attempt to do or say things that others will
like

I can only argue for ideas which I already believe

I can make impromptu speeches even on topics about which I have almost no
information

I guess I put on a show to impress or entertain people

When I am uncertain how to act in a social situation, I look to the behavior of others
for cues ‘

I would probably make a good actor

I rarely need the advice of my ﬁiends to choose movies, books, or music

I sometimes appear to others to be experiencing deeper emotions than I actually am.

I laugh more when I watch comedy with others than when alone

In a group of people I am rarely the center of attention

In different situations and with different people, I often act like very different persons
I am not particularly good at making other people like me

Even if I am not enjoying myself, I often pretend to have a good time

I’m not always the person I appear to be

I would not change my opinions (or the way I do things) in order to please someone
else or win their favor

I have considered being an entertainer

140

19. In order to get along and be liked, I tend to be what people expect me to be rather
than anything else.

20. I have never been good at games like charades or improvisational acting

21. I have trouble changing my behavior to suit different people and different situations
22. At a party I let others keep the jokes and stories going

23. I feel a bit awkward in company and do not show up quite so well as I should

24. I can look anyone in the eye and tell a lie with a straight face (if for a right end)

25. I may deceive people by being friendly when I really dislike them

Scored on a l - 5 scale (strongly disagree - strongly agree)

141

APPENDIX G

Circadian Rhythm Scale - Smith, Reilly, & Midkiff (1989)

1. Considering only your own “feeling best” rhythm, at what time would get up if
you were entirely free to plan your day

A. 5:00-6:30 a.m. (5)
B. 6:30-7:45 a.m. (4)
C. 7:45-9:45 a.m. (3)
D. 9:45-11:00 a.m. (2)
E. 11:00a.m.-12:00 (noon) (1)
2. Considering your own “feeling best” rhythm, at what time would you go to bed if

you were entirely free to plan your evening?

A. 8:00-9:00 pm. (5)

B. 9:00-10:15 pm. (4)

C. 10:15-12:30 a.m. (3)

D. 12:30-1 :45 a.m. (2)

E. 1:45am-3:00 a.m. (1)
3. Assuming normal circumstances, how easy do you ﬁnd getting up in the
morning? (check one)

A. Not at all easy (1)
B. Slightly easy (2)
C. Fairly easy (3)

D. Very easy (4)

4. How alert do you feel during the ﬁrst half hour after having awakened in the
morning? (check one)
A. not at all alert (1)
B. slightly alert (2)
C. fairly alert (3)
D. very alert (4)

5. During the ﬁrst half hour after having awakened in the morning, how tired do you
feel? (check one)
A. very tired (I)
B. fairly tired (2)
C. fairly refreshed (3)
D very refreshed (4)

142

four

10.

You have decided to engage in some physical exercise. A ﬁiend suggests that
you do this one hour twice a week and the best time for him/her is 7:00-8:00am.
Bearing in mind nothing else but your own “feeling best” rhythm, how do you

think you would perform?
A. would be in good form (4)
B. would be in reasonable form (3)

C. would ﬁnd it difﬁcult (2)
D. would ﬁnd it very difﬁcult (1)

At what time in the evening do you feel tired and, as a result, in need of sleep?
8:00-9:00 pm (5)

9:00-10:15 pm (4)

10:15-12:30 am (3)

12:30-lz45 am (2)

1:45am-3:00 am (1)

mean?

You wish to be at your peak performance for a test which you know is going to be
mentally exhausting and lasting for two hours. You are entirely free to plan your
day, and considering only your own “feeling best” rhythm, which ONE of the
testing time would you choose?

8:00-10:00 am (4)

11:00am-1 :00pm (3)

3:00-5:00 pm (2)

7:00-9:00 pm (1)

POP”?

One hears about “morning” and “evening” types of people. Which ONE of these
types do you consider yourself to be?

A. deﬁnitely a morning type (4)

B. more a morning than an evening type (3)

C. more an evening than a morning type (2)

D. deﬁnitely an evening type (1)

When would you prefer to rise (provided you have a full day’s work - 8 hours) if
you were totally free to manage your time?

A. before 6:30 am (4)

B. 6:30-7:30 am (3)

C. 7:30-8:30 am (2)

D. 8:30 am or later (1)

143

ll.

12.

13.

If you always had to rise at 6:00 am, what do you think it would be like?
very difﬁcult and unpleasant (1)

rather difﬁcult and unpleasant (2)

a little unpleasant but no great problem (3)

easy and not unpleasant (4)

93.09”?

How long a time does it usually take before you “recover your senses” in the
morning after rising from a night’s sleep?

A. 0-10 minutes (4)

B. 11-20 minutes (3)

C. 21-40 minutes (2)

D. more than 40 minutes (1)

Please indicate to what extent you are a morning or evening active individual.
pronounced moming active (morning alert and evening tired) (4)
B to some extent, morning active (3)

C. to some extent, evening active (2)

D pronounced evening active (morning tired and evening alert) (1)

,3-

144

8.

9.

APPENDIX H

Comparative Anxiety Scale - Arvey, Strickland, Drauden, & Martin (1990)

. I probably didn’t do as well as most of the other people who took these tests

I am not good at taking tests

During the testing, I often thought about how poorly I was doing

' - ".J’A‘B'

I usually get very anxious about taking tests
I usually do pretty well on tests

I expect to be among the people who score really well on this test

 

‘3 l . -1

My test scores don’t usually reﬂect my true abilities
I very much dislike taking tests of this type

During the test or tests, I found myself thinking of the consequences of failing

10. During the testing I got so nervous I couldn’t do as well as I should have

Scored on a l - 5 scale (strongly disagree - strongly agree)

145

APPENDIX I

Covariate Measures

Backggound
What is your age?
A. Younger than 30
B. 31-40
C. 41-50
D. 51-60
E. 61 or older

I really am interested in getting hired for the job for which I am applying for?

W909”?

Strongly Agree
Agree

Neutral

Disagree

Strongly Disagree

How many paper-and-pencil tests like the sort used here have you taken in the last three

years?

W909”?

Zero

One

Two

Three

Four or more

On the tests where there was time pressure, what describes your approach in trying to
complete the test?

POP”?

Completely ignore quality in favor of speed

Check quality occasionally

Slow down pace a bit to double-check items

Ensure that every item has the best answer before moving to the next one

146

Perceived Face Validim - Smither, Reilly, Millsap, Pearhnan, & Stoffey. (1993).

l.

2.

I did not understand what the tests I just took had to do with the job

I could not see any relationship between the tests I just took and what is required on
thejob

It would be obvious to anyone that the tests I just took are related to the job
The actual content of the tests I just took was clearly related to the job

There was no real connection between the tests that I went through and the job

Scored on a l - 5 scale (strongly disagree - strongly agree)

Test Motivation - - Arvey, Strickland, Drauden, & Martin (1990)

l.

2.

8.

9.

Doing well on this test (or these tests) is important to me.

I wanted to do well on this test or tests.

I tried my best on this test or tests.

I tried to do the very best I could to on this test or tests.

While taking this test or tests, I concentrated and tried to do well
I want to be among the top scores on these tests.

I pushed myself to work hard on this test or these tests.

I was extremely motivated to do well on this test or tests.

I just didn’t care how I did on this test or tests.

10. I didn’t put much effort into this test or tests.

Scored on a l - 5 scale (strongly disagree - strongly agree)

147

Social Desirability Items - Paulhus (1986)

1. I sometimes tell lies if I have to

2. I never cover up my mistakes.

3. There have been occasions when I have taken advantage of someone
4. I never swear.

5. I sometimes try to get even rather than forgive and forget.

6. I always obey laws, even if I’m unlikely to get caught.

7. I have said something bad about a ﬁiend behind his or her back

8. When I hear people talking privately, I avoid listening

 

9. I have never dropped litter on the street.

171'

10. I sometimes drive faster than the speed limit.

11. I have done things that I don’t tell other people about.

12. I never take things that don’t belong to me

13. I have never damaged a library book or store merchandise without reporting it.
14. I have some pretty awful habits.

15. I don’t gossip about other people’s business.

Scored on a l - 5 scale (very true - very untrue)

 

148

BIBLIOGRAPHY

K‘Lﬁ"

 

BIBLIOGRAPHY

Ackerrnan, PL. (1987). Individual differences in skill learning: An integration of
psychometric and information processing perspectives. Psychologicpl Bulletin. 102. 3-27.

Ackerrnan, PL. (1989). Individual differences in skill acquisition. In P.L.
Ackerrnan, R.J. Stemberg, & R. Glaser (Eds), Learning and individuﬂifferences (pp.
164-217). New York: Freeman.

Ackerrnan, P.L., & Heggestad, ED. (1997). Intelligence, personality, and f
interests: Evidence of overlapping traits. Psychological Bulletin. 121, 219-245.

Amato, PR. (1983). The effects of urbanization on interpersonal behavior.
Journal of Cross-Cultural Psychology, 14, 353-367.

Anastasi, A. (1988). Psychological Testing (6th Ed.). Macmillan: New York.

 

Fm .mrx
7'

Arvey, R.D., Strickland, W., Drauden, G., & Martin, C. (1990). Motivational
components of test taking. Personnel Psychology, 43, 695-716.

Audley, R]. (1973). Some observations on theories of choice reaction time:
Tutorial review. In S. Komblum (Ed), Attention & Performance IV. New York:
Academic Press.

Baehr, M.E., & Orban, J.A. (1989). The role of intellectual abilities and
personality characteristics in determining success in higher-level positions. Journal of
Vocational BehaLior, 35, 270-287.

Barrett, G., Phillips, J ., & Alexander, R. (1981). Concurrent and predictive
validity designs: A critical analysis. Journal of Applied Psychology. 67. 239-244.

Barrick, M.R., & Mount, MK. (1991). The Big Five personality dimensions and
job performance: A meta-analysis. Personnel Psychology, 44, 1-26.

Barrick, M.R., & Mount, MK. (1993). Autonomy as a moderator of the
relationships between the Big Five and job performance. Journal of Applied Psychology,
7_8, 111-118.

Barrick, M.R., Mount, M.K., Strauss, J .P. (1993). Conscientiousness and the
performance of sales representatives: Tests of the mediating effects of goal setting.

Jouraal of Applied Psychology, 78. 715-722.

Bates, T.C., & Eysenck, H]. (1993). Intelligence, inspection time, and decision
time. Intelligence, 17, 523-531.

149

Baxter, RN. (1941). An experimental analysis of the contributions of speed and
level in an intelligence test. Joumalgof Educational Psychological Bulletip, 96, 72.

Birren, J .E. (1965). Age changes in speeded behavior: Its central nature and
physiological correlates. In A.T. Welford & J .E. Birren (Eds.), Behavior, aging and the
nervous system (pp. 191-216). Springﬁeld, IL: Charles C. Thomas.

Birren, J .E., & Fisher, L.M. (1995). Aging and speed of behavior: Possible
consequences for psychological functioning. Annual Review of Psychology. 46. 329-
353.

Birren, J .E., Woods, A.M., & Williams, M.V. (1980). Behavioral slowing with I
age: Causes, organization, and consequences. In L.W. Poon (Ed.) Aging in the 1980s:
Psychological issues (pp. 293-308). Washington DC: American Psychological
Association.

_ at" .a-

Blake, M.J.F. (1967). Time of day effects on performance in range of tasks.
Psychonomic Science. 9. 349-350.

 

l.l. .' _Eu"f\l' ‘MH
a

Bluedorn, A.C., & Denhardt, RB. (1988). Time and organizations. Journal of
Management, 14, 299-320.

Bluedorn, A.C., Kauﬁnan, C.F., & Lane, PM. (1992). How many things do you
like to do at once? An introduction to monochronic and polychronic time. Academy of
Management Executive, 6, 17-26.

Bodenhausen, G.V. (1990). Stereotypes as judgmental heuristics: Evidence of
circadian rhythms in discrimination. Psychological Science, 1, 319-322.

Borman, W.C., & Brush, DH. (1993). More progress toward a taxonomy of
managerial performance requirements. Human Performance 6, 1-21.

 

Borman, W.C., & Motowidlo, SJ. (1993). Expanding the criterion space to
include elements of contextual performance. In N. Schmitt & W.C. Borman (eds.),
Personnel Selection in Organizations (pp. 71-98). San Francisco, CA: Jossey Bass.

Borman, W.C., White, L.A., Pulakos, E.D., & Oppler, SH. (1991). Models of
supervisory job performance ratings. Journal of Applied Psychology. 76, 863-872.

Bomstein, M.H. (1979). The pace of life: Revisited. lntemational Journal of
Psychology, 14, 83-90.

Brand, V.C. (1987). The importance of general intelligence. In S. Modgil & C.
Modgil (Eds.), Arthur J en_sen: Consensus and controversy. New York: Falmer.

 

150

 

Briggs, S.R., Cheek, J.M., & Buss, AH. (1980). An analysis of the self-
monitoring scale. Journal of Personality an_d Social Psychology, 38, 679-686.

Briggs, S.R., & Cheek, J .M. (1988). On the nature of self-monitoring: Problems
with assessment, problems with validity. Journal of Personality and Social Psychology,
3, 663-678.

Briggs, S.R., & Cheek, J .M. (1986). The role of factor analysis in the
development and evaluation of personality scales. Journal of Personalig, 54, 106-147.

Brinley, J .F . (1965). Cognitive sets, speed and accuracy of performance in the __
elderly. In A.T. Welford & J .E. Birren (eds.) Behavior, aging, and the nervous system I.-
(pp. 114-149). Springﬁeld, IL: Thomas.

Bumarn, M.A., Pennebaker, J .W., & Glass, DC. (1975). Time consciousness,
achievement striving, and the Type A cOronary prone behavior pattern. Journal of .:
AbnormAPsychology. 84. 76-79. :

 

Campbell, J .P. (1990). Modeling the performance prediction problem in industrial 5:.
and organizational psychology. In M.D. Dunnette & L.M. Hough (Eds), Handbook of
industrial and organizational psychology (2nd ed., Vol. 1, pp. 687-732). Palo Alto, CA:
Consulting Psychologists Press.

Carlson, J .S., Jensen, C.M., & Widaman, KR (1983). Reaction time,
intelligence, and attention. Intelligence. 7. 329-344.

Carlson & Widaman (1987). Elementary cognitive correlates of G: Progress and
prospects. In P.A. Vernon (Ed.), Speed of information-processing and intelmrence (pp.
69-99). Norwood, NJ: Ablex.

Camevale, AP. (1991). America and the New Economy (pp. 24-33). American
Society for Training and Development, US. Department of Labor: Washington, DC.

Carpenter, P.A., Just, M.A., & Shell, P. (1990). What one intelligence test
measures: A theoretical account of the processing in the Raven Progressive Matrices
Test. Psychological Review. 97. 404-431.

Carroll, J .B. (1980). Discussion. In DJ. Weiss (Ed.), Proceedings of the 1979
Computerized Adaptive Testing Conference (pp. 278-283, 449-452). Minneapolis, MN:
Computerized Adaptive Testing Laboratory, Psychometric Methods Program,
Department of Psychology, University of Minnesota.

Carroll, J .B. (1993). Human ggnitive abilities: A survey of factor-analy_tic
studies (pp. 440-509). Cambridge: New York.

151

Ceci, SJ. (1990). On intelligence...more or less: A bioecologjcal treamie on
intellectual development. Englewood Cliffs, NJ: Prentice-Hall.

Ceralla, J. (1985). Information processing rates in the elderly. Psychological
Bulletin 98, 67-83.

 

Ceralla, J. (1991). Age effects may be global, not local: Comment on Fisk and
Rogers (1991). Journal of Experimental Psychology: General, 120, 215-223.

Cerella, J ., Poon, L.W., & Williams, D.M., (1980). Age and the complexity

hypothesis. In L.W. Poon (Ed.), Aging in the 19805: Psychological issues (pp. 332-340).
Washington DC: American Psychological Association.

Chan, D. (1996). Video-based versus paper-and-pencil method of assessment in
situational judgment tests: Subgroup differences in performance and examinee reactions.
Unpublished master’s thesis. Michigan State University.

Chan, D., Schmitt, N., DeShon, R.P., Clause, C.S., & Delbridge, K. (1997).
Reactions to cognitive ability tests: The relationships between race, test performance,
face validity perceptions, and test-taking motivation. Journal of Applied Psychology, _8_2,
300-310.

Clausen, J. (1966). Abilig structure and subggoups in mentmetardation.
Washington, DC: Spartan Books.

 

Cohen, J. (1988). Statistical power analysis for the beﬂvioral sciences (2nd ed.)-
Hillsdale, NJ: Erlbaum.

Cohen, J ., & Cohen, P. (1983). Applied multiple regression/corregtion analysia
for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.

Colquhoun, WP. (1971). Circadian variations in mental efﬁciency. In W.P.
Colquhoun (Ed.), Biological rhythms and human performance. London: Academic Press.

Conte, J .M., Landy, F.J., & Mathieu, J.E. (1995). Time urgency: Conceptual and
construct development. Journal of Applied Psychology, 80, 178-185.

Cottle, T.J. (1976). Perceiving time: A psychological invesgation with men an_d
women. New York: Wiley.

Cronbach, L.J., & Warrington, W.G. (1951). Time limit tests: Estimating their
reliability and degree of speeding. Psychometrika, 14, 167-188.

Data comparison. Detroit, MI: Aon Consulting.

152

Davidson, W.M., & Carroll, J .B. (1945). Speed and level components in time-
limit scores. A factor analysis. Educational & Paychologjcal Measurement, 5, 411-427.

Day, D.V., & Silverrnan, SB. (1989). Personality and job performance: Evidence
of incremental validity. Personnel Psychology, 42, 25-36.

Deary, II. (1993). Inspection time and WAIS-R IQ subtypes: A conﬁrmatory
factor analysis study. Intelligence, 17, 223-236.

Deary, I.J., & Stough, C. (1996). Intelligence and inspection time: Achievements,
prospects, and problems. American Psychologist, 51, 599-608.

Digrnan, J .M. (1990). Personality structure: Emergence of the ﬁve-factor model.
. Annual Review of Paychology. 41. 417-440.

 

DuBois, PH. (1932). A speed factor in mental tests. Archives of Psychology, 22
(Whole No. 141).

Edwards, J .R., Baglioni, A.J., Jr., & Cooper, CL. (1990). Examining relationships
among self-report measures of Type A behavior pattern: The effects of dimensionality,
measurement error, and differences in underlying constructs. Journal of Applied
Psychology, 3, 440-454.

F olkard, S., Monk, T., & Lobban, M. (1979). Towards a predictive test of
adjustment to shift work. Ergonomics, 22, 79-91.

F lokard, S., Wever, R.A., & Wildgruber, CM. (1983). Multi-oscillatory control
of circadian rhythms in human performance. Nature 305, 223-226.

 

Following policies and procedures. Detroit, MI: Aon Consulting.

Gangestad, S., & Snyder, M. (1985). On the nature of self-monitoring. In P.
Shaver (Ed) Review of pers_opality and social psychologiaVol. 6: Self, situations and
social belaryior (pp. 65-86). Beverly Hills, CA: Sage.

 

Gaugler, B.B., Rosenthal, D.B., Thornton, GO 111, & Bentson, C. (1987). Meta-
analysis of assessment center validity. Journal of Applied Paychologyr 72. 493-511.

Geary, D.C., French, P.A., & Wiley, J .G. (1993). Simple and complex mental
subtraction: Strategy choice and speed-of-processing differences in younger an older
adults. Psychology and Aging, 8, 242-256.

Goldberg, LR. (1993). The structure of phenotypic personality traits. American
Psghologiat. 48, 26-34.

153

 

Gorsuch, R.L. (1988). Exploratory factor analysis. In J .R. Nesselroade & R.B.
Cattell (Eds), Handbook of Multivariate Experimental Psychology. New York: Plenum.

Gregory, R.J. (1996). Psychological Testing (2nd ed.). Allyn & Bacon: Needharn
Heights, MA.

Guion, RM. (1991). Personnel assessment, selection, and placement. In M.D.
Dunnette & L.M. Hough (eds), H_andbool_< of industrial and organizational psychology,
(2nd ed., Vol. 2, pp. 327-397). Palo Alto, CA: Consulting Psychologists Press.

Guion, R.M., & Cranny, C]. (1982). A note on concurrent and predictive validity T"
designs: A critical reanalysis. Journal of Applied Psychology, 66, 1-6.

Hale, S., Myerson, J ., & Wagstaff, D. (1987), General slowing and nonverbal
information processing: Evidence for a power law. Journal of Gerontology, 42, 131-136.

Hall, ET. (1983). The daace of life: The other dimer1_sion of time. Garden City,
NY: Anchor Press.

mm _u

 

Hartigan, J .A., & Wigdor, AK. (1989). Fairness in Emploment Testing.
Washington: National Academy Press.

Hashemian, K. (1978). An experimental investigation of the relationship between
test anxiety and memory processes under different motivation conditions (University of
Florida, 1977). Dissertation Abstracts International, 38, (7-A), 4046.

 

Haynes, S.G., Levine, S., Scotch, N., F einleib, M., Kannel, W.B. (1978). The
relationship of psychosocial factors to coronary heart disease in the Framingham study:
1. Methods and risk factors. American Journal of Epidemiology, 107, 362-383.

Helmreich, R.L., Sawin, L.L., & Carsrud, AL. (1986). The honeymoon effect in job
performance: Temporal increases in the predictive power of achievement motivation.

Journal of Applied Psychology, ﬂ, 185-188.

Heneman, H.G., III, & Heneman, R.L. (1994). Stafﬁng Organizations.
Middleton, WI: Mendota House.

Himmilweit, HT. (1946). Speed and accuracy of work as related to temperament.
British Journal of Psychology, 36, 132-144.

Horn, J .L. (1978). Individual differences in cognitive development: Final report.
Betheseda, Maryland: National Science Foundation.

Horn, J .L. (1981). Age differences in human abilities: Final report. Betheseda,
Maryland: National Institute of Aging.

154

Horn, J .L. (1988). Thinking about human abilities. In J .R. Nesselroade & R.B.

Cattell (Eds), Handbook of multivariate experimental psychology (2nd ed., pp. 645-685).
New York: Plenum.

Home, J .A., & Ostberg, O. (1976). A self-assessment questionnaire to determine
morningness-eveningness in human circadian rhythms. International Journal of
Chronobiology, 4. 97-1 10.

Horne, J .A., & Ostberg, O. (1977). Individual differences in human circadian
rhythms. Biological Psychology, 5, 179-190.

Hough, L.M. (1989). Development of personality measures to supplement
selection decisions. In B]. Fallon, H.P. Pﬁster, & J. Brebner (Eds), Advances in

industrﬁrl organizational psychology (pp. 365-375).

Hough, L.M. (1992). The “Big Five” personality variables-construct confusion:
Description versus prediction. Human Performance 5, 139-155.

 

Hough, L.M., Eaton, N.K., Dunnette, M.D., Kamp, J .D., & McCloy, RA. (1990).
Criterion-related validities of personality constructs and the effect of response distortion
on those validities. Journal of Applied Psychology, 75, 581-595.

Humphreys, M.S., & Revelle, W. (1984). Personality, motivation, and
performance: A theory of the relationship between individual differences and information
processing. Psychological Review,91 , 153-184.

Hunt, E. (1978). Mechanics of verbal ability. Psychological Review. 85. 109-130.

Hunter, J .E. (1986). Cognitive ability, cognitive aptitudes, job knowledge, and job
performance. Journal of Vocational Behjavior, 29. 340-362.

Hunter, J .E., & Hunter, RF. (1984). Validity and utility of alternative predictors
of job performance. Psychological Bulletin. 96. 72-98.

Jackson, M., & McClelland, J. (1979). Processing determinants of reading speed.
Journal of Experimental Psychology: General. 108. 151-181.

Jenkins, C.D., Zyzanski, S.J., & Rosenman, RH. (1979). Jenkins Activity Survey
manual (Form C). New York: Psychological Corporation.

 

Jensen, AR. (1981). Bias in mental testing. New York: The Free Press.
Jensen, AR. (1987). Individual differences in the Hick paradigm. In P.A. Vernon

(Ed.), Speed of information processing and intelligence (pp. 101-175). Norwood, NJ:
Ablex.

155

. .- as.”
f

 

 

Jensen, AR. (1992). Commentary: Vehicles of g. Psychological Science, 3, 275-
278. "

Jensen, AR. (1993). Why is reaction time correlated with psychometric g?
Current Directions in Psychological Science, 2, 53-56.

Kanfer, R. (1992). Work motivation: New directions in theory and research. In C.L.
Cooper & T. Robertson (Eds), International Review of Industriaiand Organizational
Psychology (V 01. 7, pp. 1-53). London: John Wiley & Sons, Ltd.

Kanfer, R., & Ackerrnan, PL. (1989). Motivation and cognitive abilities: An
integrative/aptitude-treatrnent interaction approach to skill acquisition [Monograph].

Journal of Applied Psychology, 74, 657-690.

Kauﬁnan, C.F., Lane, P.M., & Lindquist, J .D. (1991). Exploring more than 24
hours a day: A preliminary investigation of polychronic time use. Journal of Consumer
Research 18, 392-401.

 

Kelly, J .R. (1988). Entrainment in individual and group behavior. In J .E.
McGrath’s (Ed.) The social psychology of time (pp. 89-112) Newbury Park, CA: Sage.

Kendall, L.M. (1964). The effects of varying time limits on test validity.
Educationﬂand Psychological Measurement, 24, 789-800.

Kerkhof, G. (1985). Inter-individual differences in the human circadian system: A
review. Biological Psychology, 20, 83-112.

Kranzler, J .H. (1990). The nature of intelligence: A unitary process or a number
of independent processes? Unpublished doctoral dissertation. University of California at
Berkeley.

Kranzler, J.H., & Jensen, AR. (1989). Inspection time and intelligence: A meta-
analysis. Intelligence, 13, 329-347.

Kyllonen. PC. (1985). Dimensions of information processing speed. Brooks Air
Force Base, TX: Air Force Systems Command, AF HRL-TP-84-56.

Landy, F.J., Rastegary, H., Thayer, J ., & Colvin, C. (1991). Time urgency: The
construct and its measurement. Journal of Applied Psychology, 7_6, 644-657.

Lauer, RH. (1981). Temporal man: The meaning and uses of social time. New
York: Praeger.

Lennox, R.D., & Wolfe, RN. (1984). Revision of the self-monitoring scale.
Jouraal of Personality and Social Psychology, 46, 1349-1364.

156

Levine, G., Preddy, D., & Thorndike, R.L. (1987). Speed of information
processing and level of cognitive ability. Persoaalityand Individual Differences, 8, 599-
607.

Levine, R.V. (1988). The pace of lie across cultures. In J .E. McGrath’s (Ed.) Tm
social psychology of time (pp. 39-60). Newbury Park, CA: Sage.

Levine, R.V., & Bartlett, K. (1984). Pace of life, punctuality, and coronary heart
disease in six countries. Journal of Cross-Cultural Psychology, 15, 233-255.

Lima, S.D., Hale, S., & Myerson J. (1991). How general is general slowing? 'F
Evidence from the lexical domain. Psychology and Aging, 6, 416-425.

Lohman, D. (1979). Spatial ability: A review and reanalysis of the correlational
literature (Tech. Rep. No. 8). Aptitude Research Project, Stanford University, School of
Education, Stanford, CA.

 

Lohman, D. (1986). The effect of speed-accuracy tradeoff on sex differences in i
mental rotation. Perception & Psychophysics, 39, 427-436.

Lohman, D. (1994). Implications of cognitive psychology for ability testing:
Three critical assumptions. In M.G. Rumsey, C.B. Walker, & J .H. Harris (Eds),
Personnel selection and classiﬁcation (pp. 145-171). Hillsdale, NJ: Erlbaum.

Lord, FM. (1956). A study of speed factors in tests and academic grades.
Psychometrika, 21, 31-50.

Lord, F .M. (1971). Testing if two measuring procedures measure the same
psychological dimension. RB-71—36. Educational Testing Service: Princeton, NJ.

Lucia, M. (1985). Effect of season, day of week, and the degree of urbanization
omce of life. Unpublished master’s thesis. California State University, Fresno.

Macan, T.H., Avedon, M.J., Paese, M., & Smith, DE. (1994). The effects of
applicants’ reactions to cognitive ability tests and an assessment center. Personnel
Psychologya47, 715-738.

Mangan, G.L. (1959). A factorial study of speed, power, and related temperament
variables. British Journal of Educational Paychology, _2_9_, 144-154.

Marr & Stemberg (1987). The role of mental speed in intelligence: A triarchic

perspective. In P.A. Vernon (Ed.), med of information-processing and intelligence (pp.
271-294). Norwood, NJ: Ablex.

157

McGrath, J.E. (1976). Stress and behavior in organization. In M.D. Dunnette

(Ed.), Handboolaof industrial and organizational psychology (pp. 1351-1395). New
York: John Wiley.

McGrath, J .E., & Rothchrod, N.L. (1983). Time and behavior in organizations.
Research in Organizational BehaﬁomS. 57-101.

McHenry, Hough, Toquam, Hanson, & Ashworth, (1990). Project A validity
results: The relationship between predictor and criterion domains. Personnel Psychology,
4_3_, 335-3 54.

Miller, L.T., & Vernon, RA. (1983). The general factor in short-term memory,
intelligence, and reaction time. Intelligence, E, 5-29.

 

Morrison, E]. (1960). On test variance and the dimensions of the measurement
situation. Educational and Psychological Measurement. 20. 231-250.

Muchinsky, PM. (1997). Psychology applied to work (5th ed.). Brooks/Cole:
Paciﬁc Grove, CA.

Mumford, M.D., Baughman, W.A., Threlfall, K.V., Uhlman, C.E., & Constanza.
DP. (1993). Personality, adaptability, and performance: Performance on well-deﬁned
and ill-deﬁned problem-solving tasks. Human Performance 6, 241-285.

 

Myers, CT. (1952). The factorial composition and validity of differentially
speeded tests. Psychometrika, 17, 347-352.

Naylor, J .C., Pritchard, R.D., & Ilgen, DR. (1980). A theory of behavior in
organizations. New York: Academic Press.

Neisser, U., Boodoo, G., Bouchard, T.J., Jr., Boykin, A.W., Brody, N., Ceci, S.J.,
Halpem, D.F., Loehlin, J .C., Perloff, R., Stemberg, R.J., Urbina, S. (1996). Intelligence:
Knowns and unknowns. American Psychologist, 51, 77-101.

Nettelbeck, T. (1987). Inspection time and intelligence. In P.A. Vernon (Ed.),
Speed of processing and intelligence (pp. 295-346) Norwood, NJ: Ablex.

Nevo, B., & Sfez, J. (1985). Exarninees’ feedback questionnaires. Assessment a_n_c_1
Evaluation in Higher Education, 10, 236-249.

Nunnally, J .C., & Bernstein, I.H. (1994). Psychometric theog (3rd ed.). New
York: McGraw-Hill.

Parks, R.W., Loewenstein, D.A., Dodrill, K.L., Barker, W.W., Yoshii, F., Chang,
J .Y., Emran, A., Apicella, A., Sherrnata, W.A., & Duara, R. (1988). Cerebral metabolic

158

effects of a verbal ﬂuency test: A PET scan study. Journal of Clinical and Experimental
Neurop§ychologv. 10, 565-575.

Paulhus, D.L. (1986). Self-deception and impression management in test
responses. In A. Angleitner & J.S. Wiggins (Eds), Personalig assessment via
questionnaires (pp. 143-165). Berlin: Springer-Verlag.

Peterson, N.G. (1993). Review of issues associated with speededness of GATB
tests. American Institutes for Research: Washington, DC.

Practical arithmetic. Detroit, MI: Aon Consulting.

Raven, J .C. (1965). Advanced Progessive Matrices, Sets I and 11. London: H.K.
Lewis.

Ree, M.J., & Earles, J .A. (1992). Intelligence is the best predictor of job
performance, Current Directions in Psychological Science, 1, 86-89.

Reed, T.E., & Jensen, AR. (1992). Conduction velocity in a brain nerve pathway
of normal adults correlates with intelligence. Intelligence, 16, 259-272.

Revelle, W. (1989). Personality, motivation, and cognitive performance. In R
Kanfer, P.L. Ackerrnan, R. Cudeck (Eds), Abilities, motivation, and methodology: The
Minnesota symposium on learning and individual differences (pp. 297-341). Hillsdale, NJ:

Erlbaum.

Rindler, SE. (1979). Pitfalls in assessing test speediness. Journal of Educational
Measurement, _1_6, 261-270.

Rynes, S.L. (1993). Who’s selecting whom? Effects of selection practices on
applicant attitudes and behavior. In N. Schmitt & W.C. Borman (Eds), Personnel
selection in organizations (pp. 240-274). San Francisco: Jossey-Bass.

Sackett, P.R., Burris, L.R., & Ryan, AM. (1989). Coaching and practice effects in
personnel selection. In C.I. Cooper & I. Robertson (Eds), International review of

industrial and organization psychology (pp. 145-163).

Sackett, P.R., Zedeck, S., & Fogli, L. (1988). Relations between typical and
maximum job performance. Journal of Applied Psghology, 3, 482-486.

Salthouse, TA. (1985). Speed of behavior and its implications for cognition. In
J .E. Birren & K.W. Schaie (Eds), _H_andbooﬁk of the psychology of aging (2nd ed., pp.
400-426). New York: Van Nostrand Reinhold.

Salthouse, TA. (1993). Attentional blocks are not responsible for age-related
slowing. Journal of Gerontolcgg chhologic_al Science. 48, 263-270.

159

 

Schaie, K.W., & Willis, S.L. (1993). Age difference patterns of psychometric
intelligence in adulthood: Generalizability within and across ability domains.

Psychology and Aging, 8, 44-55.

Schmidt, F .L., Greenthal, A.L., Hunter, J .E., Bemer, J.G., & Seaton, F.W. (1977).
Job samples vs. paper-and-pencil trade and technical tests: Adverse impact and examinee
attitudes. Personnel Psychology, 30, 187-197. ‘

Schmit, M.J., & Ryan, AM. (1992). Test-taking dispositions: A missing link?
Journal of Applied PsychomV. 77, 629-637. F

Schmitt, N. (1977). Interrater agreement in dimensionality and combination of
assessment center judgments. Journal of Applied Psychology, 62, 171-176.

Schneider, J .R. (1992). The effects of group variables on evaluations of !
individuals in assessment center group exercises. Unpublished doctoral dissertation.
Michigan State University: East Lansing, MI.

 

F'-

Schneider, R.J., & Hough, L.M. (1995). Personality and industrial/organizational
psychology. Intemrﬂrnal Review of I/O Psychology, 10, 75-129.

Schuler, R.S., & Jackson, SE. (1987). Linking human resource practices with
competitive strategies. Academy of Management Executive, 1, 207-219.

Secretary’s Commission on Achieving Necessary Skills (SCANS) (1992). Skills
and Tasks for Jobs. US. Government Printing Ofﬁce: Washington, DC.

Smith, C.S., Reilly, C., & Midkiff, K. (1989). Evaluation of three circadian
rhythm questionnaires with suggestions for an improved measure of morningness.
Journal of Applied Psychology, 74, 728-738.

Smith, FD. (1991). Work samples as measures of performance. In A.K. Wigdor
and BF. Green Jr. (Eds), Performance a_ssessment for the workplace (pp. 27-52).

Smith, P.C., Kendall, L.M., & Hulin, CL. (1969). The measurement of
satisfaction in work_and retirement: A strategy for the study of attitudes. Chicago: Rand-
McNally.

 

Smither, J .W., Reilly, R.R., Millsap, R.E., Pearlman, K., & Stoffey, R.W. (1993).
Applicant reactions to selection procedures. Personnel Psychology. 46, 49-76.

Snyder, M. (1974). The self-monitoring of expressive behavior. Journal of
Personality and Social Psychology, 30, 526-537.

160

Snyder, M. (1979). Self-monitoring processes. In L. Berkowitz (Ed.), Advances in
experimental social psychology (V 01. 12). New York: Academic Press.

Snyder, M., & Gangestad, S. (1986). On the nature of self-monitoring: Matters of
assessment, matters of validity. Journal of Personality and Social Psychology, 51, 125-
139.

Snyder, M., & Ickes, W. (1985). Personality and social behavior. In G. Lindzey
& E. Aronson (eds), The Handbook of Social Psycholagy (pp. 883-947). New York:
Random House.

Snyder, M., & Kendzierski, D. (1982). Acting on one’s attitudes: Procedures for
linking attitude and behavior. Journal of Experimental Social Psychology. 18, 165-183. 1-

Spearrnan, C. (1904). “General Intelligence,” objectively determined and
measured. American Journal of Paychology, 15, 201-293. 1;

 

Spence, J.T., & Pred, R.S., & Helmreich, R.L. (1987). Achievement strivings, E
scholastic aptitude, and academic performance: A follow-up to "Impatience versus
achievement strivings in the Type A pattern." Journal of Applied Psychology, 25, 176-
178. '

Spielberger, C.D., Jacobs, G.A., Crane, R.G., Russell, S.F., Westberry, L., Barker,
L., Johnson, E.H., Knight, J ., & Marks, E. (1979). Preliminm manual for the Sane-Tait
Personality Inventory (STPI). Tampa, University of South Florida Human Resources
Institute.

Steiger, J .H. (1980). Tests for comparing elements of a correlation matrix.
Psychological Bulletin, 87. 245-251.

Stemberg, R.J. (1985). Implicit theories of intelligence, creativity, and wisdom.

Journal of Personalig and Social Psychology, 49, 607-627.

Stemberg, R.J., Conway, B.E., Ketron, J .L., & Bernstein, M. (1981). People’s
conceptions of intelligence. Journal of Personalig and $0031 Psychology. 41. 37-55.

Task Force on Assessment Center Standards (1989). Guidelines and ethical
considerations for assessment center operations. Public Personnel Management, 18, 457-
470.

Tett, R.P., Jackson, D.N., & Rothstein, M. (1991). Personality measures as
predictors of job performance: A meta-analytic review. Personnel Psychology. 44. 703-
742.

Thorndike, E.L., Bregman, E.O., Cobb, M.V., Woodyard, E., & the Staff of the
Division of Psychology of the Institute of Educational Research of Teachers College,

161

Columbia University (1926). The measurement of intelligence. New York: Bureau of
Publications, Teachers College, Columbia University.

Thurstone, LL. (1937). Ability, motivation, and speed. Psychometrika, 2, 249-
254.

Torsvall, L., & Akerstedt, T. (1980). A diurnal type scale. Scandinavian Journal
of Work_and Environmental Health, 6, 283-290.

Tuttle, DB. (1996). Measuring individuals’ temporal orientation. Paper presented
at the American Psychological Association Meetings, Toronto, Ontario, Canada.

Vernon, PA. (1987). New developments in reaction time research. In P.A.

Vernon (Ed.), Speed of information processing and intelligence (pp. 1-20). Norwood,
NJ: Ablex.

Vernon, RA. (1987). Speed of information processinggd intelligence.
Norwood, NJ: Ablex.

Vernon, RA. (1991). Studying intelligence the hard way. Intelligence, 15, 389-
395. .

Verster, J .M. (1983). The structure, organization, and correlates of cognitive
speed and accuracy: A cross-cultural study using computerized tests. In S.H. Irvine &
J .W. Berry (Eds), Human assessment and cultural factors (pp. 275-292).

Warner, R. (1988). Rhythm in social interaction. In J.E. McGrath’s (Ed.) 111;
social psychology of time (pp. 63-88). Newbury Park, CA: Sage.

Weiss, H.M., & Adler, S. (1984). Personality and organizational behavior. In B.M.
Staw & L.L. Cummings (eds), Research in Organizational Behavior (Vol. 6, pp. 1-50).
Greenwich, CT: JAI Press.

Wesman, AG. (1960). Some effects of speed in test use. Educational and
Psychological Megurement, 20, 267-274.

Wickelgren, W.A. (1977). Speed-accuracy tradeoff and information processing
dynamics. Acta Psychologica, 41, 67-85.

Wonderlic personnel test & scholastic level exam (1992). Libertyville, IL:
Wonderlic Personnel Test, Inc.

Work experience and interest inventog. Detroit, MI: Aon Consulting.

Zinbarg, R., & Revelle, W. (1989). Personality and conditioning: A test of four
models. Journal of Persoralitpyand Socii Psychology, _5_Z, 301-314.

162