EL
\99‘]

This is to certify that the

dissertation entitled

THE EFFECTS OF LEARNER CONTROL AND
INDIVIDUAL DIFFERENCES 0N LEARNING OUTCOMES

presented by

Marcia Jeanette Simmering

has been accepted towards fulfillment
of the requirements for

Ph . D . degree in Management

 

 

 

. JR

 

 

r Major professor

7/???

MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

LIBRARY
Michigan State
University

 

 

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

we chlRClDdoOquS-p.“

THE EFFECTS OF LEARNER CONTROL AND
INDIVIDUAL DIFFERENCES ON LEARNING OUTCOMES

By

Marcia Jeanette Simmering

A DISSERTATION

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

1 999

 

ABSTRACT

THE EFFECTS OF LEARNER CONTROL AND
INDIVIDUAL DIFFERENCES ON LEARNING OUTCOMES

By

Marcia Jeanette Simmering

The degree to which individuals learn in a training situation is an important
consideration for research and practice. In this dissertation, I examine the effects
of learner control and individual differences on learning outcomes. Specifically, I
predicted that learners’ personality characteristics and attitudes would have
differing effects on motivation to learn and learning depending on the degree to
which learners have control over the training. I examined this fundamental
question of what leads to learning in the context of a relatively new instructional
medium—web—based training.

One-hundred eighty-nine students participated in web-based training to
improve grammar skills and were evaluated on post-tests and subsequent
performance on a writing assignment. Results of this study indicated that mastery
oriented individuals were more motivated to learn. Additionally, perceptions of the
legitimacy of training were related to motivation to learn and were moderated by
the amount of learner control in training. A predicted relationship between
motivation to learn and learning was not found with self-report measures of
motivation to learn. However, two unobtrusive measures of motivation to learn
were positively related to training. Finally, learning was significantly related to

transfer of training.

ACKNOWLEDGEMENTS

There are many people who contributed to this dissertation who I would
like to thank. First, my committee members, Ray Noe (chair), Alison Barber, Paul
Rubin, and Linn Van Dyne, provided excellent feedback and guidance throughout
this project. Additionally, each of my committee members has individually had a
hand in helping me develop as a researcher and/or teacher, and for that I am
very thankful.

I would also like to thank those people who assisted with data collection or
preparation of this document. Thank you to Jason Coleman for the expertise and
patience in teaching me how to think like a programmer. Thank you to Alison
Barber for funding the purchase of the Wonderlic Personnel Tests and for access
to the MGT 310 course for data collection. Thank you also to Michael Wesson
and Jay Cook for assistance in data collection in the MGT 310 classes. Thank
you to Dr. Vernon Miller and Rebecca Chory for allowing me to include their
communications students as participants in this study. Thank you to Jennifer
Alberico for assistance in data entry. Finally, thank you to Shari Martin for the
thorough and thoughtful editing of this document.

Lastly, I would like to thank two people who have offered useful advice
and social support during this process. Thank you to Jason Colquitt, who is not
only an excellent researcher, but a dear friend. Thank you to Becky Luce, whose

support and understanding have made her a priceless friend and colleague.

TABLE OF CONTENTS

LIST OF TABLES ..................................................................................... vi
LIST OF FIGURES .................................................................................... x
INTRODUCTION ....................................................................................... 1
LITERATURE REVIEW ............................................................................. 4
Learning and Learning Outcomes .................................................. 7

The Learning Treatment: Learner Control ...................................... 8
Individual Differences in Training .................................................... 13
Personality ........................................................................... 16

Attitudes ............................................................................... 25

A New Training Medium: Web-Based Training .................... 27

Additional Outcomes ....................................................................... 33

Other Factors .................................................................................. 35
METHOD ................................................................................................... 38
Participants ..................................................................................... 38
Power Analysis ............................................................................... 39
Web-Based Training ....................................................................... 40
Procedure ....................................................................................... 41

Pilot Testing .................................................................................... 47
Measures ........................................................................................ 47
Independent Variables .......................................................... 48

Dependent Variables ........................................................... 53

RESULTS .................................................................................................. 59
Data Analysis Strategy .................................................................... 59
Descriptive Statistics ....................................................................... 59
Handling of Missing Data ................................................................ 62

Test of Overall Significance of ........................................................ 63
Analysis of Hypotheses ................................................................. 64
Additional Hypotheses .................................................................... 92
Exploratory Analyses ...................................................................... 94
Mediation Analysis ............................................................... 94

Use of Practice Exercises and Hyperlinks ........................... 96

iv

DISCUSSION ............................................................................................ 99

Overview of Discussion .................................................................. 99
Summary of Results and Implications for Theory Building ............. 99
Limitations ............................................................................. ‘ ......... 1 05
Future Research ............................................................................. 109
Motivation to Learn ..... , ......................................................... 109
Learner Control .................................................................... 1 10
Goal Orientation ...................................................... ' ............. 1 1 1
Legitimacy ............................................................................ 113
Transfer of Training .............................................................. 1 14
Training Media ..................................................................... 115
Practical Application ....................................................................... 1 16
APPENDIX A: WEB-BASED TRAINING ................................................... 118
APPENDIX B: INFORMED CONSENT FORM .......................................... 143
APPENDIX C: MEASURES ....................................................................... 145
APPENDIX D: PRE-TESTS/POST-TESTS ............................................... 153
APPENDIX E: INDIVIDUAL PAPER ASSIGNMENT ................................. 160
REFERENCES ......................................................................................... 162

LIST OF TABLES

TABLE 1. Comparison of Common Training Media .................................. 32

TABLE 2. Means, Standard Deviations, and Zero-Order
lntercorrelations ............................................................ ' ............ 6 0

TABLE 3. Regression Analysis of Motivation to Learn Predicting
Post-test Scores (Hypothesis 1) ................................................ 66

TABLE 4. Regression Analysis of Motivation to Learn Predicting
Post-test Paragraph Scores (Hypothesis 1) .............................. 66

TABLE 5. Regression Analysis of VIE Predicting Post-test Scores
(Hypothesis 1) ........................................................................... 67

TABLE 6. Regression Analysis of VIE Predicting Post-test Paragraph
Scores (Hypothesis 1) ............................................................... 67

TABLE 7. Regression Analysis of Conscientiousness Predicting
Motivation to Learn (Hypothesis 2) ............................................ 68

TABLE 8. Regression Analysis of Conscientiousness Predicting
VIE (Hypothesis 2) ..................................................................... 68

TABLE 9. Regression Analysis of Conscientiousness Predicting .
Post-test Scores (Hypothesis 3) ................................................. 69

TABLE 10. Regression Analysis of Conscientiousness Predicting
Post-test Paragraph Scores (Hypothesis 3) ............................ 69

TABLE 11. Regression Analysis of Conscientiousness X
Learner Control Predicting Motivation to Learn
(Hypothesis 4) ......................................................................... 70

TABLE 12. Regression Analysis of Conscientiousness X Learner
Control Predicting VIE (Hypothesis 4) .................................... 71

TABLE 13. Regression Analysis of Openness to Experience
Predicting Motivation to Learn (Hypothesis 5) ......................... 72

TABLE 14. Regression Analysis of Openness to Experience
Predicting VIE (Hypothesis 5) .................................................. 72

vi

TABLE 15.

TABLE 16.

TABLE 17.

TABLE 18.

TABLE 19.

TABLE 20.

TABLE 21.

TABLE 22.

TABLE 23.

TABLE 24.

TABLE 25.

TABLE 26.

TABLE 27.

Regression Analysis of Openness to Experience
Predicting Post-test Scores (Hypothesis 6) ............................. 73

Regression Analysis of Openness to Experience
Predicting Post-test Paragraph Scores
(Hypothesis 6) ............ 73

Regression Analysis of Openness to Experience X
Learner Control Predicting Motivation to Learn
(Hypothesis 7) ......................................................................... 74

Regression Analysis of Openness to Experience X
Learner Control Predicting VIE (Hypothesis 7) ...................... 75

Regression Analysis of Mastery Orientation Predicting
Motivation to Learn (Hypothesis 8) ........................................ 76

Regression Analysis of Mastery Orientation Predicting
VIE (Hypothesis 8) ................................................................ 76

Regression Analysis of Mastery Orientation Predicting
Post-test Scores (Hypothesis 9) ............................................ 76

Regression Analysis of Mastery Orientation Predicting
Post-test Paragraph Scores (Hypothesis 9) ........................... 77

Regression Analysis of Mastery Orientation X
Learner Control Predicting Motivation to Learn
(Hypothesis 10) ....................................................................... 78

Regression Analysis of Mastery Orientation X
Learner Control Predicting VIE (Hypothesis 10) ..................... 78

Regression Analysis of Performance Orientation
Predicting Motivation to Learn (Hypothesis 11) ..................... 79

Regression Analysis of Performance Orientation
Predicting VIE (Hypothesis 11) .............................................. 80

Regression Analysis of Performance Orientation
Predicting Post-test Scores (Hypothesis 12) ......................... 81

vii

TABLE 28. Regression Analysis of Performance Orientation
Predicting Post-test Paragraph Scores (Hypothesis 12) .......... 81

TABLE 29. Regression Analysis of Performance Orientation X
Learner Control Predicting Motivation to Learn
(Hypothesis 13) ......................................................................... 82

TABLE 30. Regression Analysis of Performance Orientation X
Learner Control Predicting VIE
(Hypothesis 13) ......................................................................... 82

TABLE 31. Regression Analysis of Perceptions of Legitimacy
of Learning Grammar Predicting Motivation to Learn
(Hypothesis 14) ........................................................................ 83

TABLE 32. Regression Analysis of Perceptions of Legitimacy
of Learning Grammar Predicting VIE (Hypothesis 14) .............. 84

TABLE 33. Regression Analysis of Perceptions of Legitimacy
of WBT Predicting Motivation to Learn (Hypothesis 14) .......... 84

TABLE 34. Regression Analysis of Perceptions of Legitimacy
of WBT Predicting VIE (Hypothesis 14) .................................... 84

TABLE 35. Regression Analysis of Perceptions of Legitimacy of
Learning Grammar Predicting Post-test Scores
(Hypothesis 15) ......................................................................... 85

TABLE 36. Regression Analysis of Perceptions of Legitimacy of
Learning Grammar Predicting Post-test Paragraph Scores
(Hypothesis 15) ......................................................................... 86

TABLE 37. Regression Analysis of Perceptions of Legitimacy
of WBT Predicting Post-test Scores (Hypothesis 15) ............... 86

TABLE 38. Regression Analysis of Perceptions of Legitimacy
of WBT Predicting Post-test Paragraph Scores
(Hypothesis 15) ......................................................................... 87

TABLE 39. Regression Analysis of Perceptions of Legitimacy

of Learning Grammar X Learner Control Predicting
Motivation to Learn (Hypothesis 16) ......................................... 88

viii

TABLE 40. Regression Analysis of Perceptions of Legitimacy
of Learning Grammar X Learner Control

Predicting VIE (Hypothesis 16) .............................................

TABLE 41. Regression Analysis of Perceptions of Legitimacy
of WBT X Learner Control Predicting

Motivation to Learn (Hypothesis 16) ......................................

TABLE 42. Regression Analysis of Perceptions of Legitimacy
of WBT X Learner Control Predicting VIE

(Hypothesis 16) ......................................................................

TABLE 43. Regression Analysis of Post-test Scores Predicting

Transfer of Training (Hypothesis 19) ......................................

TABLE 44. Regression Analysis of Post-test Paragraph Scores

Predicting Transfer of Training (Hypothesis 19) ....................
TABLE 45. Summary of Hypothesized Results ........................................

TABLE 46. Regression Analysis of Frequency of Practice Predicting

Post-test Scores .....................................................................

TABLE 47. Regression Analysis of Frequency of Practice Predicting

Post-test Paragraph Scores ...................................................

TABLE 48. Regression Analysis of Frequency of Use of Hyperlinks

Predicting Post-test Scores .................................................... ,

TABLE 49. Regression Analysis of Frequency of Use of Hyperlinks

Predicting Post-test Paragraph Scores ..................................

89

89

90

93

93

95

97

97

98

98

LIST OF FIGURES

FIGURE 1. Conceptual Model ................................................................... 3
FIGURE 2. Expected Interaction between Individual Differences

and Treatment ........................................................... ; ............ 18
FIGURE 3. Study Time Line ...................................................................... 43
FIGURE 4. Web-Based Training Structure ................................................ 45

FIGURE 5. Plot of Perceptions of Legitimacy of WBT X Learner
Control Predicting Motivation to Learn ................................... 91

INTRODUCTION

Learning is a key issue in understanding the effectiveness of training in
business organizations. The Bureau of Labor Statistics estimated that US.
companies spent over $55 billion in training and development in 1996
(http://www.astd.org/virtual_community/comm_trends/98soi_slide_train_
expenditures. html). With such an investment in training and developing
employees, evaluation is a crucial issue. Evaluation of training is an attempt to
answer two questions: (1) Did learning occur? and (2) Did transfer of training
occur (Kraiger, Ford, and Salas, 1993)? The principal concern of training
evaluation is whether or not learning occurred.

Gagne (1985) defines learning as “a change in human disposition or
capability that persists over a period of time and is not simply ascribable to
processes of growth” (p. 2). Gagne further notes that learning exhibits itself as a
change in behavior that is evidenced after an individual has been put in a
learning situation. In order to improve future training effectiveness, it is important
to understand what characteristics of training and trainees contribute to learning.

My main research question in this dissertation is aimed at understanding
the extent to which individual differences of learners impact motivation to learn
and learning in different training conditions. Specifically, I expect personality
characteristics and attitudes of learners to have differing impact on motivation to
learn depending on the degree to which learners have control over the training. I
also expect that motivation to learn will be positively related to learning. I

examine this fundamental question of what leads to learning in the context of a

1

relatively new instructional medium—web-based training. I discuss a theoretical
framework, my research model, features of web-based training, and hypotheses.

The overview of my research model is presented in Figure 1. Learners’
individual differences are expected to have direct effects on motivation to learn,
which in turn will have a positive relationship with learning. The individual
differences of learners are also hypothesized to have direct effects on learning.
The learning context, or more specifically, learner control of training, is expected
to have a moderating effect on the relationship between the individual differences
and motivation to learn, such that in situations of high learner control, individual
differences will have a stronger effect.

There are three expected contributions of this dissertation to the existing
literature on learning. First, I examine individual differences in the learning
situation, which expands on existing research in personality, especially in the
context of person-situation interaction. Second, I investigate levels of learner
control, a topic that has received attention in educational literature but has been
neglected in organizational training literature. Third, I examine learning in the
context of a new instructional medium, web-based training, which will provide
empirical data regarding the effectiveness of this technologically advanced

training medium.

Figure 1

Conceptual Model

 

 

 

Individual
Differences

 

 

 

 

 

Motivation
- To Learn

 

 

 
   

Treatment
(Learner Control)

 

 

 

 

Y

 

 

 

Learning

 

 

 

LITERATURE REVIEW

Two theories guide my investigation of individual differences and learner
control in a web-based training context. First, a general aptitude theory explains
the relationship between different levels of learner control and the effects of
individual differences on learning. Specifically, aptitude theory emphasizes the
importance of investigating aptitude-treatment interactions (ATI) (Snow &
Lohman, 1984). Second, Mischel’s (1977) work on person-situation interaction
guides the hypotheses regarding how the individual differences and treatment
will interact. Within this theoretical framework, I draw from models of instructional
design and organizational training to explain the roles of specific constructs.

An aptitude theory describes inner environments (individual differences of
learners) and outer environments in which the aptitude operates (Snow &
Lohman, 1984). Snow (1978) defines aptitudes as “student characteristics that
predict response to instruction under a given instructional treatment" (p. 227).
Although the focus of aptitudes has traditionally been on cognitive processes,
they may also encompass personality and attitudes that contribute to or detract
from learning (Snow, 1978; Snow & Lohman, 1984). In the context of learning
from instruction, aptitude theory specifies that individual aptitudes must be
matched to instructional demands in order for learning to occur. Accordingly, an
aptitude-treatment interaction, in which aptitudes are directly related to learning
outcomes and are also moderated by the training treatment, is implied (Snow &

Lohman, 1984).

Aptitude-treatment interactions occur through aptitude processes, which
are defined as predictable, directed changes in psychological functioning by
which some individual learners adapt to the demands of instructional conditions,
while other learners do not (Snow & Lohman, 1984). For example, a learner who
is ovenrvhelmed with information may not have the cognitive capabilities to sort
and order information in a way that facilitates learning. However, if the learning
situation (or treatment) is such that the Information is presented in a very
structured manner, the individual may have increased learning.

The relevance of aptitudes Is in their effect on information processing
which is related to learning outcomes (Snow, 1978). These processes are
influenced not only by individual differences, but also by situational demands (the
treatment). Differences in aptitude processes could occur due to the speed or
efficiency, the sequencing, and the higher order strategic processes used to
digest information presented in a learning situation (Snow & Lohman, 1984).
Therefore, if the treatment can accommodate these differences in information
processing, the situation can facilitate learning. In conclusion, Snow (1978) notes
that the treatment should be designed to capitalize on a learner’s aptitudes and
compensate for the learner's inaptitudes.

Aptitude theory provides insight into why individual differences constructs
and treatment conditions will interact. Aptitude theory is similar to Mischel's
(1977) work on person-situation interaction. Mischel (1977) theorizes that
different environments, or situations, influence characteristics of the person such

that the individual may think and act differently in various situations. The situation

can also be conceptualized as a stimulus or treatment, which can apply to
training research.

Although not specifically aimed at training research, this broad theory
applies to many work situations, which can include organizational training.
Mischel (1977) describes two categories of situations: powerful and weak.
Powerful (or strong) situations are those in which demands placed on individuals
induce conformity, whereas weak situations are those in which there are few
situational pressures to conform (Mischel, 1977). In the context of training, a
powerful situation is one in which there is considerable pressure to put forth effort
or to become cognitively engaged. For example, a classroom presentation that
requires active participation of members could be a powerful situation. A weak
situation is one where there is less pressure to put forth effort and become
cognitively engaged. Self-paced learning, in which an individual is expected to
learn without direct supervision, is an example of a weak situation. Mischel
(1977) predicts that individual differences exert the greatest influence in the weak
situations and have the smallest effect in the powerful ones. Therefore, I predict
that when the training treatment presents a weak situation, variation in individual
differences will have more impact on outcomes than when there is a powerful
sfluafion.

In addition to these two general theories, instructional design theories from
the educational psychology literature and models of training from the
organizational behavior and human resource management literature address

these concepts. These two literatures often examine similar constructs and

relationships, but don’t necessarily draw from one another. A strength of my
dissertation is the use of themes and research from organizational behavior,
human resource management, industrial/organizational psychology, and
educational psychology. If organizational training researchers ignore the wealth
of information present in the educational psychology literature, important theories

will be overlooked.

Learning and Learning Outcomes

Learning occurs in a learning event, which Gagne (1985) characterizes as
having four elements. The first element is the learner, or in the case of
organizational training, the trainee. The trainee’s abilities, skills, attitudes, and
personality characteristics are important aspects of the individual learner that
must be considered. The second element, the stimulus situation, is a situation
that stimulates the leamer’s senses. A training program and materials are
categorized asstimulus situations. The third element is the information recovered
from the learner's memory. Tapping into this memory can enhance learning
effectiveness. The fourth element is the response, or in some cases,
performance. A response is any action taken as a result of the stimulus situation
inputs. A result is identified as performance if the effect of that action is
considered.

Learning is the major dependent variable in my dissertation; specifically, I
evaluate learning with measures of cognitive and affective outcomes. Kraiger et

al. (1993) developed a theoretical taxonomy of learning outcomes in three

categories: cognitive, skill-based, and attitudinal. Cognitive learning outcomes
account for declarative knowledge, procedural knowledge, and cognitive
strategies. Skill-based learning outcomes refer to development of technical and
motor skills. Affective learning outcomes involve a change in motivation or
attitudes.

Cognitive learning outcomes are common in organizations and have been
used in training research to assess learning (e.g., Colquitt & Simmering, 1998;
Fisher & Ford, 1998; Phillips & Gully, 1997; Quinones, 1995). Cognitive
outcomes are appropriate for examining skills such as verbal ability, and are
often assessed using written tests (Kraiger et al., 1993).

I also investigate affective outcomes and time spent on training. These
additional outcomes are relevant to investigations of web-based training because
they are related to proposed strengths of web-based training. These outcomes

are explained more fully in the description of web-based training.

The Learning Treatment: Learner Control

Situational factors of training, or more specifically, the treatments involved
in training, may play important roles in learning outcomes. The strength of the
training situation in my model is the level of learner control of training. Learner
control is relevant to today’s organizations because there is an emphasized need
for individualized continuous learning (London & Bassman, 1989). Trainees often
have different knowledge, skills, and abilities, which makes individualized

learning experiences effective (Fleishman & Mumford, 1989). Additionally,

employees are oflen expected to keep their skills up to date through self-directed
professional training and development activities (London & Bassman, (1989).

Learner control has been studied in the educational literature, specifically
in terms of its role in computer-based training (CBT), or computer-assisted
instruction (CAI). In studies of learner control, the results are not consistent. One
area that has been overlooked in this area of research is the degree to which
individual differences of learners interact with learner control to affect training
outcomes. The role of individual differences will be discussed later.

Learner control has three components: pacing, sequence, and content
(Milhelm & Martin, 1991). Pacing is the speed at which the training occurs.
Learners often have varied levels of reading or comprehension speed, and
leamer-controlled pacing should accommodate these differences. If a learner can
progress at a pace that is comfortable, he or she is likely to retain more of the
learning content. Sequence is the order in which the trainee learns the material.
Sequence can vary on a macro level, where larger training modules maybe
taken in any order, or it can vary on a micro level, where material presented
within fixed modules may be presented in a non-structured sequence (Milhelm &
Martin, 1991). If a learner is able to control the sequence of the material, he or
she is more likely to maintain motivation by choosing the most interesting
material to cover first. The final component of learner control in computer training
is content. There is high learner control of content when the learner is able to
choose which information he or she sees or reviews. This element of learner

control contributes not only to the efficiency of the learning, but to the learner's

motivation (Milhelm & Martin, 1991). Individuals can ignore content that is
already known and spend more time on content that is unfamiliar. Furthermore, if
the learner does not have to spend time on familiar content, he or she is more
likely to have a positive reaction to the training.

The levels of learner control of these three components can take several
forms, varying by the degree to which the learner has control. Complete learner
control affords the user all decisions regarding content, pacing, and sequence
(Murphy & Davidson, 1991). In this situation, a learner could simply decide not to
engage in any of the training. However, the goal is for the learner to engage in
the training following his or her own strategy in order to maximize learning and
efficiency. The inherent assumption with complete learner control, however, is
that the learner is the most effective at formulating and executing that strategy
(Murphy & Davidson, 1991).

Program control, or adaptive control, allows the user to continue with
training material or a post-test only when the program has taken the learner
through a certain amount of material or a set level of performance is achieved in
practice (Murphy & Davidson, 1991). With these conditions, the learner is
required to do at least the basics, and is unable to avoid necessary training. This
contributes to creating a more uniform level of competency between learners,
and may be most effective when there are minimum standards for proficiency.

Learner advisement retains the learner’s control of content, pacing, and
sequence, but informs the learner of topics that should be covered for mastery

(Murphy & Davidson, 1991). At that point, the learner can ignore the advice or

10

follow it, based on his or her perceived need for training. Again, there is the
implicit assumption that the learner is able to determine the necessary learning
strategy and effectively follow it.

The research on learner control has produced mixed results (Kinzie, 1990;
Milhelm & Martin, 1991; Steinberg, 1977). Some studies have found that
increased learner control results in higher performance on post-tests (e.g., Avner,
Moore, & Smith, 1980). However, others have found opposite results (e.g.,
Pollock & Sullivan, 1990). Additionally, many studies have found no differences
in post-test scores due to amount of learner control (Lee & Lee, 1991; Murphy &
Davidson, 1991; Pridemore 8 Klein, 1991). There are several possible reasons
for discrepant results in this stream of literature.

Differences in the results of these studies may be due to the neglect of
relevant individual differences of learners. Several studies included no measure
of any person-specific factors, not even demographic characteristics (e.g., Avner
et al., 1980; Murphy & Davidson, 1991; Pridemore & Klein, 1991). Other studies
(e.g., Lee & Lee, 1991; Kinzie et al., 1988) measured Iearners' ability or prior
knowledge, but did not assess any personality characteristics of learners.
Another possible reason for different results is the small sample sizes generally
used in these studies. The majority of these studies have sample sizes below
100. One notable exception is Avner et al., (1980) which had a sample size of
approximately 700 participants. Finally, these studies were conducted on a
variety of samples, ranging from junior high school students (e.g., Kinzie et al.,

1988) to nursing students (e.g., Murphy & Davidson, 1991). These groups may

11

differ in their mean level or variance in relevant personality characteristics.
Therefore, individual differences are important constructs to investigate in the
context of learner control.

When there is high learner control, it is a weak situation: there are fewer
pressures for the learner to behave in a certain way. In this situation, the learner
may choose not to take certain modules of the training or engage in optional
exercises or practice. A leamer’s individual differences are more likely to dictate
his or her actions. Therefore, personality variables should be strong indicators of
motivation to learn and learning outcomes. In high learner control treatments,
there is more opportunity for learners’ aptitudes to have an impact on learning
outcomes. If individual learners possess the aptitudes necessary to learn, they
are likely to have positive outcomes. However, if the necessary aptitudes are not
present or prominent enough, the learning outcome will be less positive.

Low learner control is a strong situation because the learner is expected to
conform to the pacing, sequence, and content presented in the program. In this
situation, a learner may be required to engage in practice sessions or pass a
post-test before finishing the training. To some extent, the treatment will
compensate for individual differences in aptitude. Learners will process
information similarly if it is presented in a very structured fashion.

In summary, aptitude theories prescribe looking at the interaction between
aptitudes and treatments. One way to conceptualize different treatments is

variation in the amount of learner control in a training situation. Previous studies

12

of learner control have all but ignored individual differences of learners. l next

consider the relationships between individual differences in learning outcomes.

Individual Differences in Training

Although the learner has long been recognized as an integral part of the
learning process, training research has only recently begun to utilize individual
differences to explain learning outcomes. Recent studies support the relationship
between individual’s characteristics and learning beyond training design (e.g.,
Colquitt & Simmering, 1998; Fisher & Ford, 1998; Martocchio & Webster, 1992;
Noe, 1986; Noe and Schmitt, 1986; Phillips & Gully, 1997; Quinones, 1995;
Webster & Martocchio, 1995). Additionally, many instructional design theories
account for individual differences, even though their emphasis is generally on
manipulating attitudes rather than measuring more stable characteristics (e.g.,
Gagne, 1985; Gagne & Medsker, 1996; Keller, 1983).

Models of instructional design and organizational training provide
information about individual differences that are likely to be related to learning.
Keller (1983) described four conditions that would lead to increased motivation in
learning. Interest, or curiosity, was the first. This is similar to Gagne’s prescription
for instruction to spark the attention of the learner (Gagne, 1985; Gagne &
Medsker, 1996). The training situation must be constructed to maximize interest
and interaction. However, individuals may have varying levels of innate curiosity
also. The second condition in Keller's model is relevance, which is the extent to

which a goal is related to the instruction. Therefore, the learner's goal is an

13

important consideration in the model. The third condition is expectancy, which is
a subset of the motivation to learn construct. This will be described in more
detail. Finally, satisfaction with the training may lead to higher motivation. Keller
notes that satisfaction may stem from extrinsic rewards and intrinsic motivation.

An influential model in organizational training literature has been Noe’s
(1986) model of training effectiveness. Noe included motivational and
environmental influences in his model, and predicted that motivation to learn
would be an integral part of training effectiveness. Motivation to learn is “a
specific desire of the trainee to learn the content of the training program” (Noe,
1986; p. 743). Motivation to learn is malleable and likely to vary depending on
characteristics of both the situation and the individual.

Recent studies of learning have supported a significant, positive
relationship between motivation to learn and learning outcomes. For example,
Quinones (1995) in his study of pre-training effects on training outcomes, found
that motivation to learn in a computerized decision-making task was significantly
related to reactions to training, quality and quantity of decisions, and behavioral
measures of training effectiveness. Another recent study, conducted by Colquitt
and Simmering (1998), found that measures of motivation to learn at the midpoint
of a six-week training course were significantly related to a measure of
declarative knowledge at the end of training.

Baldwin, Magjuka, and Loher (1991) found that motivation at different
points during training was related to a test of declarative knowledge. Motivation

to learn the material was assessed before training began. After training, a similar

14

scale was used to assess the degree to which the trainee perceived himself or
herself to have been motivated in training. Both measures had significant,
positive relationships with trainees’ ability to answer open-ended questions of
knowledge of training materials.

The relationship between motivation to learn and learning outcomes has
been supported in an open learning setting (Wart & Bunce, 1995). Open learning
occurs when individual learners work on their own with a high level of autonomy
and often involves multimedia applications (Warr & Bunce, 1995). A condition of
high learner control can be characterized as open learning because learners are
given some degree of autonomy over the content, pacing, and sequence of
material. Watt and Bunce (1995) used a sample of junior managers to
investigate the impact of trainee characteristics on learning outcomes in an open
learning setting. They found that motivation to learn was significantly related to
each learner’s tutors' assessment of their performance in the training program.
Additionally, Warr and Bunce recommend that motivation to learn should be
measured at multiple time periods in training because motivation may change as
exposure to the training content and method increase. Some limitations of this
study include a subjective measure of learning and a small sample size.
Therefore, further study using a larger sample and an objective measure of

learning (declarative knowledge) is warranted.

Hypothesis 1: Motivation to learn will be positively related to Ieaming.

15

Personality

Three motivational constructs that have been examined in the context of
training motivation and learning outcomes are conscientiousness, openness to
experience, and goal orientation. Conscientiousness and openness to
experience are two of the “Big Five” personality factors (Costa & McCrae, 1992).
Conscientiousness subsumes characteristics such as being hardworking,
dependable, and organized. Highly conscientious people are more likely to finish
things they have started and tend to be more responsible than individuals with
low conscientiousness (Costa & McCrae, 1992).

A meta-analysis of the “Big Five” personality factors and organizational
outcomes indicated that conscientiousness was the most robust predictor of job
performance (Barrick & Mount, 1991). The estimated true correlation between
conscientiousness and job performance criteria was determined to be P=.22
(Barrick & Mount, 1991). Although training outcomes and job performance differ,
these results indicate the importance of conscientiousness to the work 1
environment.

Conscientiousness can have an impact on learning as well as job
performance. Learners who are conscientious are more likely to pay attention to
details and to work through content in an orderly fashion. Colquitt and Simmering
(1998) found that conscientiousness was significantly related to motivation to
Ieam and cognitive outcomes in a longitudinal training study. Furthermore, they
found that conscientious individuals pemisted more in the face of initial failure in

training. Although failure is not a part of every training situation, this finding points

16

to the tendency of conscientious individuals to overcome obstacles in training.
Another study that supports the relationships among conscientiousness,
motivation to learn, and learning was conducted by Gellatly (1996). In this study,
conscientiousness was significantly related to expectancy, an element of
motivation to learn. Furthermore, expectancy was significantly related to
performance on an arithmetic task, which measured a cognitive learning

outcome.

Hypothesis 2: Conscientiousness will be positively related to motivation to

Ieam.

Hypothesis 3: Conscientiousness will be positively related to Ieaming.

There is increasing evidence that the relationship between
conscientiousness and behavior may be moderated by situational factors. For
example, Barrick & Mount (1993) conceptualized the degree of autonomy in a job
as representative of the strength of the situation. They found that when a
situation was weak, conscientiousness was more predictive of job performance.
This can be expected in a training situation as well. In a weak situation, a more
conscientious person should perform better than a less conscientious person
because of internalized perseverance and diligence.

In my dissertation, the weak situation of high learner control should allow
an individual’s conscientiousness to play a greater role in learning. In a high

learner control treatment, a learner’s conscientiousness should have a much

17

greater impact. An individual who is more conscientious will be more likely to
persevere, even when his or her interest in the training has diminished.
Furthermore, these individuals are likely to be more engaged in the content,
participating in optional activities even when they are not required. The
conscientious individual’s need to complete work will drive this behavior.

In the low learner control situation, the conscientiousness and motivation
to learn will not be as strongly related. If learners are given few opportunities to
affect their learning, then their characteristics won’t be as influential in training. In
the low Ieamer control condition, the level of motivation to learn will likely be

average (see Figure 2).

Figure 2

Expected Interaction between Individual Differences and Treatment

 

 

 

High __ / High Learner Control
///
l/
l//
Motivation // Low Learner Control
to Learn //
/
///
Low I I
l l
Low High

Individual Difference
(e.g., Conscientiousness)

18

Hypothesis 4: Leamer control will moderate the relationship between
conscientiousness and motivation to Ieam such that when there is high
control, conscientious learners will have higher motivation to Ieam than
low conscientious learners. In the low Ieamer control treatment, motivation

to Ieam will be similar for all individuals.

Openness to experience is another Big Five personality characteristic
(Costa & McCrae, 1992). An individual’s level of openness is related to his or her
intellectual curiosity, sensitivity, and broad-mlndedness. An individual who is
open is more likely to be interested in engaging in new tasks. Furthermore,
openness may reduce feelings of apprehension often associated with new
situations. Barrick and Mount (1991), in their meta-analysis of the “Big Five”
personality traits, suggested that openness may be related to training proficiency
because trainees who are more open are more likely to have positive attitudes
towards Ieaming experiences in general. I

There has been little direct research involving the relationship between
openness to experience and learning. Despite the lack of research in this area,
studies using a construct similar to openness to experience are relevant.
Specifically, research on cognitive playfulness provides information related to the
effects of openness to experience. Cognitive playfulness is an individual
difference that describes a person’s tendency to interact spontaneously,

inventively, and imaginatively with microcomputers (Webster & Martocchio,

19

1992). This definition is similar to the facets of intellectual curiosity and open-
mindedness that are included in the construct of openness to experience.

Webster and Martocchio (1992) found that, in the context of a software
training course, cognitive playfulness was positively related to learning and
mood, an affective outcome. In this study, Ieaming was measured using a
declarative knowledge test and participants indicated their mood on a self-report
measure. Webster & Martocchio (1995) also found support for the relationship
between cognitive playfulness and positive training outcomes. Within a computer
training setting, learners with high cognitive playfulness exhibited more focus and
attention to the task and subsequently had higher scores on a performance
measure of cognitive outcomes of training. Furthermore, learners high in
cognitive playfulness tended to use the computer training software more
frequently and had more positive attitudes towards the training.

Although cognitive playfulness is specifically referenced to
microcomputers, a more general measure may provide insight into learning using
media other than microcomputers. The central mechanism of cognitive
playfulness is captured by the intellectual curiosity in the openness to experience
construct. Therefore, these results should generalize to links between openness

to experience and training outcomes.

Hypothesis 5: Openness to experience will be positively related to

motivation to Ieam.

20

Pg.

 

Hypothesis 6: Openness to experience will be positively related to

Ieaming.

In the different levels of learner control, openness to experience should
play a large part in learners’ motivation and interest. In a high learner control
situation, an individual’s openness to experience should dictate the degree to
which he or she explores content and sequencing and optional elements of
training, as openness involves intellectual curiosity (Costa & McCrae, 1992).
Learners who are more curious and more tolerant of new situations should find
training more motivating. In contrast, learners who experience low learner control
have prescribed pacing, content, and sequence, and are less likely to differ in

their Ieaming in relation to their openness to experience.

Hypothesis 7: Learner control will moderate the relationship between
openness to experience and motivation to Ieam such that when there is
high control, highly open learners will have higher motivation to Ieam than
leamers with low openness. In the low learner control treatment,

motivation to Ieam will be similar for all individuals.

As stated previously, in an aptitude theory, aptitudes and treatments will
interact to affect outcomes. Snow (1978) notes that aptitudes related to
information processing in different treatments can differ in reference to whether a

person is goal—directed or not. In a training situation where there is no explicit

21

performance goal, learners’ goal orientations will impact their behaviors. Keller
(1983), in his instructional theory, emphasized the importance of relevance, or
the extent to which a goal is related to instruction.

In training situations, learners often may choose different types of goals
that can impact their subsequent actions (Button, Mathieu, & Zajac, 1996). Goal
orientation consists of mastery orientation (or learning orientation) in which

individuals are focused on gaining competence and of performance orientation in

V' " . .-_-1.'!_

which individuals are focused on demonstrating competence. Individuals with

mastery orientations are more likely to engage in difficult tasks in which they can

 

learn even if there is a high chance of failure. Conversely, a performance
orientation leads individuals to avoid difficult tasks so they will not risk looking
bad in front of others.

Recent training research has supported the importance of goal orientation
in understanding training outcomes. Fisher & Ford (1998) found that mastery
orientation was significantly related to the mental workload that individuals
perceived to have during training. In other words, the degree to which individuals
felt that they spent their time on tasks and the amount of effort they gave was
related to their emphasis on mastery goals. This study also indicated that
performance orientation may detract from learning, as individuals with high
performance orientations reported that they felt they exerted less on-task effort.
Therefore, it is likely to expect that individuals with strong mastery orientations

will be more engaged in the training and more likely to be motivated to learn and

22

that individuals with strong performance orientations will be less motivated to
Ieam.

Two recent studies have investigated the specific effects of goal
orientation on learning outcomes. Colquitt and Simmering ( 1998) found that
mastery orientation was significantly related to motivation to learn in a training
situation. Furthermore, performance orientation was negatively related to
motivation to Ieam and learning. Phillips and Gully (1997) found that mastery
orientation was positively related to self-efficacy, which was related to
performance on a declarative knowledge test. Additionally, performance
orientation was negatively related to self-efficacy. Therefore, this pattern of
relationships, in which mastery orientation facilitates training outcomes and
performance orientation detracts from them, should be found in similar
investigations of learning.

In a training situation, individuals with mastery goals are less concerned
with scoring well on tests, but are focused on Ieaming the material (Button et al.,
1996). Performance oriented individuals are more intent on demonstrating their
competence by performing well on tests or completing training quickly,
sometimes at the expense of not really learning the material. Therefore, I expect
that individuals with mastery orientations will have a stronger motivation to Ieam
and more positive learning outcomes than individuals with performance
orientations.

Early research of goal orientation conceptualized the construct as a single

dimension with performance orientation and Ieaming orientation on opposite

23

 

ends of a continuum using studies of mostly schoolchildren (Dweck, 1986; Elliott
& Dweck, 1988). However, recent research has conceptualized goal orientation
differently than Dweck and her colleagues. Button, et al. (1996) presented factor-
analytic evidence that goal orientation is best conceptualized as two independent
dimensions rather than a unidimensional continuum. That is, individuals can be
simultaneously high or low on both learning and performance orientations.

Button et al. (1996) also argued that individuals have dispositional goal
orientations which predispose them to react to situations in specific ways,
although situational cues can impact those predispositions (see also Dweck,
1975). Such situational cues could be the training treatment.

Linking these constructs to aptitude theory, Snow (1978) notes that
individuals process information differently depending on how they approach
goals and subgoals and their strategies for learning. The degree to which
learners pay attention to material and concentrate in order to reach a goal may
be influenced by the treatment. When a treatment allows learners morecontrol,
their goal orientation should have a stronger effect than when there is less

learner control.
Hypothesis 8: Mastery orientation will be positively related to motivation to

Ieam.

Hypothesis 9: Mastery orientation will be positively related to Ieaming.

24

Hypothesis 10: Learner control will moderate the relationship between
goal orientation and motivation to Ieam such that when there is high
control, learners with a strong mastery orientation will have higher

motivation to Ieam than those with a weak mastery orientation.

Hypothesis 1 1: Performance orientation will be negatively related to
motivation to Ieam.

Hypothesis 12: Performance orientation will be negatively related to
Ieaming.

Hypothesis 13: Learner control will moderate the relationship between
goal orientation and motivation to Ieam such that when there is high
control, learners with a strong performance orientation will be less

motivated to Ieam than those with a weak performance orientation.

Attitudes

Aptitude theory does not specifically address attitudes. However, they are
an integral part of instructional design theories (e.g., Gagne, 1985; Gagne &
Medsker, 1996; Keller, 1983). These theories commonly give prescriptions for

training treatments that are intended to alter attitudes of learners. Therefore, they
are important to consider in investigations of Ieaming.
Legitimacy is one attitude that can be referenced to a training treatment.
Legitimacy has been used in the organization theory literature to explain the

degree to which organizational structures and functions are “taken for granted”

25

.‘H’

 

(Scott, 1991; Hannan & Carroll, 1992). Legitimacy is an integral part of both
Institutional Theory and Population Ecology theory. lnstitutionalists contend that
the more legitimate a practice, the more likely individuals will engage in it. For
instance, the more common training is to an organization, the more likely
individuals will find that it is legitimate and continue to be involved in it. According
to the theory, legitimacy can stem from three sources: rules and regulations,
norms, and individuals’ cognitive structures.

Population ecologists use the term legitimacy in the same sense as
lnstitutionalists, but argue that legitimacy stems solely from density. The basic
notion of Population Ecology is that density, which is simply the frequency of a
practice, leads to legitimacy. Therefore, the more prevalent a practice, the more
legitimate it becomes. In the context of organizational training, it can be inferred
that the more prevalent training programs or specific attributes of those training
programs are, the more likely individuals will view them as legitimate. For
example, the prevalence of a particular training medium will lead to its increased
legitimacy.

The legitimacy of a training experience should influence learning
outcomes. A learner who believes the training method to be legitimate will be
more likely to become engaged in it. Conversely, if a learner feels that the

training is not a suitable way to learn the material, he or she will be less engaged.

Hypothesis 14: Perceptions of legitimacy will be positively related to

motivation to Ieam.

26

Hypothesis 15: Perceptions of legitimacy will be positively related to
Ieaming.

Hypothesis 1 6: Learner control will moderate the relationship between
perceptions of legitimacy and motivation to Ieam such that when there is
high control, Ieamers who believe the training is highly legitimate will have
higher motivation to Ieam than learners with lower perceptions of
legitimacy. In the low learner control treatment, motivation to Ieam will be

similar for all individuals.

A New Training Medium: Web-Based Training

Aptitude-treatment interaction can be studied with any number of training
media. However, in an investigation of learner control, one newly developed
training medium that is especially relevant is web-based training (WBT). Web-
based training, or web-based instruction (WBI) has become an increasingly
popular topic of business periodicals (e.g., Cohen, 1997; Craiger & Weiss, 1997;
Curtin, 1997; Filipczak, 1996; Pollack & Masters, 1997; Stevens, 1996; Wulf,
1996). Web-based training has been defined as “a hypermedia-based
instructional program which utilizes the attributes and resources of the World
Wide Web to create a meaningful Ieaming environment where Ieaming is
fostered and supported” (Khan, 1997).

Two components that may be built into web-based training that are
important within the context of learner control are hyperlinks and practice

exercises. Hyperlinks are a unique component of web-based training that may

27

prove useful to learning. Hyperlinks are text or graphics that, when selected by a
trainee with a mouse-click, move to another “page” within the training content.
Hyperlinks can be linked to anything within a training program or to the World
Wide Web for access to outside sources. This increased ability to find and gather
information is a major benefit of WBT. Additionally, hyperlinks within a WBT
program are often discretionary, so a trainee’s motivation to Ieam may dictate the
frequency of hyperlink use.

Practice has long been recognized as beneficial to learning within training
(Gagne & Medsker, 1996). WBT allows for various types of interactive or non-
interactive practice. Practice exercises can be created In the form of games in
order to attract and maintain trainee attention. Altemately, they can be fashioned
to simulate actual job tasks to create training fidelity and foster transfer of
training. Regardless of how it is designed, practice is an important component of
WBT.

Web-based training is also an interesting medium to use because of
technological advances in organizations. The importance of technology in human
resource management systems and training, in particular, has been emphasized
recently (e.g., Goldstein & Gilliam, 1990; Howard, 1995; Tumage, 1990). The
potential usefulness of web-based training for instruction can be evaluated by
examining the ways in which media are chosen for instruction. If WBT is matched
to the characteristics of instruction, it is more likely to be effective. In choosing

media for instruction, three factors are important to consider: (1) the physical

28

attributes of the medium, (2) learner, setting, and task characteristics, and (3)
practical factors such as cost (Reiser & Gagne, 1983).

The physical attributes of the medium are its capabilities to provide text,
pictures, sound, or motion (Reiser & Gagne, 1983). Web-based training has
many of these capabilities, and a benefit of WBT is that they can be used
simultaneously. Learner characteristics may refer to learner abilities, such as
reading abilities or other skill levels. These abilities can be construed to be the
same as the aptitudes. Because WBT has many physical attributes available, it
can be modified to meet different learner characteristics. The amount of learner
control or level of readability is easily changed (Khan, 1997). Setting
characteristics can refer to the physical location of training or in what type of
setting (e.g., group, individual, etc.) training occurs. Web-based training may be
used anywhere a computer is linked to a program, and may be used in pairs or
small groups, although often WBT is fashioned for individual use. Task
characteristics are the elements of the tasks that can only be supported by
certain media. For example, in order to understand a spoken foreign language,
you must have an audio component to the training medium. Many different task
characteristics can be supported by WBT, excluding situations in which the
learner must work in direct physical contact with a trainer or actual equipment
must be used (Khan, 1997). These exceptions are few, however—even
simulations can be presented via video in WBT. Therefore, many types of tasks

can be taught using WBT.

29

Tedia

CCIICE’I

The final factors to consider in choosing instructional media are practical
factors such as cost, time, and availability of trainers or equipment. Reiser and
Gagne (1983) caution that practical factors should not be the primary concern
when selecting media, but should be considered only after the other criteria for
media have been met, therefore maximizing instructional value. Cost is a main
concern for WBT because of the use of microcomputers, which are a relatively
expensive instructional medium. However, with the flexibility of time and location
for use of WBT, several learners can share computers. Furthermore, most
organizations today have computers and related equipment that are capable of
handling WBT programs. Beyond the equipment, another cost of WBT is in
developing the training programs. Few organizations have the internal expertise
to create effective WBT programs, so consultants are often used. Although both
of these are major concerns for WBT, the cost of these in the long run is
diminished, especially when compared to costs of other non-web computer-
based training (Craiger & Weiss, 1997; Khan, 1997). For example, if a CD-ROM
is used for a training program, and changes to content will involve a change to
the program and reissue of the CD-ROM, it can be very costly, especially if there
are many learners. Web-based training content, however, is generated from a
single source (the Internet or a company intranet), and changes to the content
are immediate and comprehensive (Filipczak, 1996‘). All users will see changes
without needing extra or different equipment.

Web-based training enjoys advantages over other types of more

commonly used media when these three factors are considered. Table 1

3O

compares the factors identified by Reiser & Gagne (1983) of two other common
types of training media. WBT offers advantages of increased flexibility and
capabilities.

Web-based training has disadvantages in relation to other instructional
media. Because the program must be run off of the World Vlfide Web or a
company’s intranet, the capabilities for video and audio richness are not as
extensive as for CD-ROM based programs. Additionally, intranets may be
susceptible to theft, and security measures must be built into the program to
protect information (Canterucci, 1997). This may increase costs of development
and maintenance. Also, web-based programs are often designed for a specific
type of web browser and are limited to the capabilities of certain personal
computers (Curtin, 1997). This can result in programs being built for the “lowest

common denominator" of equipment that may preclude more advanced features.

31

 

Table 1

Comparison of Common Training Media

 

 

 

Medium! Instructor Printed Text Web-based
Characteristic Training
Physical Limited to oral Limited to written Can include
Attributes language and language and audio, video, and
demonstration pictures interactive
components
(a) Learner (a) Can (a) Can (a) Can
individualize individualize individualize
material in learning material if
small groups, strategy only built in
but not in
larger ones
(b) Setting (b) Limited to (b) Can use in (b) Limited to
where any setting where
instructor can computer can
physically be physically be
(c) Task (c) Suitable for (c) Suitable for (c) Suitable for
characteristics tasks that can tasks that can tasks that can

 

 

 

ranges from high
to low; can be
difficult to find
qualified
instructors

materials need to
be updated
frequently

 

 

be taught be learned by be taught
orally or reading only verbally, with
demonstrated text, with
pictures, with
simulation, or
with video
Practical factors Cost of trainer Low cost unless High initial cost,

but low cost to
maintain and
update

 

 

Note. Based on Reiser & Gagne (1983).

32

 

Additional Outcomes

One of the benefits that proponents of web-based training have identified
is that it can be a more efficient means of training. In other words, learners are
expected to spend less time in training. Although I don't compare the efficiency of
WBT to other training media, I will examine the differences in efficiency between
two levels of learner control. In general, I expect that when the learner has high
control, he or she will spend less time on training. It is a fundamental assumption
of learner control research that learners know what is most effective and efficient

for them.

Hypothesis 1 7: Learners in the high leamer control condition will spend
less time on training overall than Ieamers in the low Ieamer control

condition.

Kraiger, et al (1993) identify affective based Ieaming outcomes as one of
three important training outcomes. This includes issues such as attitudes,
motivation, and goals that are relevant to the objective of the training program.
These learner attitudes can determine behavior or performance, and are
therefore important indicators of training effectiveness (Gagne, 1984; Kraiger et
al., 1993).

For any training course, an affective outcome of interest may be the

learner’s level of desire to participate in another similar training program. In the

33

to

RR

case of WBT, this may be especially important, because the investment in a
WBT program is expected to pay off over time. Unlike a lecture, which may have
a single administration and a one-time cost, WBT is expensive to start but
becomes more cost-effective the more that it is used. If learners find WBT to be
something to which they have an unfavorable reaction, their disinterest in

pursuing further WBT can make the program a losing investment for an

 

r
organization. Furthermore, dissatisfied learners may communicate negative .‘_
reactions to training to other learners and influence their views.

The amount of learner control should have a direct effect on the level of :-
satisfaction with training. In general, individuals who are allowed to control the E

content, pacing, and sequence of the training should feel trusted and
empowered. Therefore, they will likely be satisfied and perhaps more inclined to

participate in future web-based training.

Hypothesis 18: Individuals in the high learner control treatment will have
more positive affective outcomes than individuals in the low leamer control

treatment.

Transfer of training is the degree to which trainees effectively apply the
knowledge, skills, and attitudes gained in a training context to the job context
(Ford & Kraiger, 1995). One expected outcome of transfer of training is whether
or not the skills and behaviors learned in training generalize to situations in a job

context that are similar to, but not the same as, the training (Baldwin & Ford,

34

1988). Baldwin 8 Ford (1988) proposed a model of transfer of training, in which
trainee characteristics, training design, and the work environment were related to
Ieaming and retention. Transfer of training (generalization and maintenance of
trained skills and behaviors) was modeled to be influenced by trainee
characteristics, the work environment, and the learning and retention in training.
Learning is necessary for transfer of training to occur, but it is not sufficient to
ensure transfer. I.
In this dissertation, I am interested in the link between learning in training

and transfer of training. Transfer of training is always a concern when studying

.' 1.1“ it" -_

the effectiveness of training, and although it is not my main research question, I

 

lull."

will address it. Additionally, transfer of training is affected by a number of different
factors (e.g., work environment, training design), but the context of my study
does not allow for investigation of these. Therefore, I will only test the
significance of the relationship between learning and transfer of training and not

focus on additional antecedents of transfer of training.
Hypothesis 19: Learning will be positively related to transfer of training.

Other Factors

In the context of web-based training, some individual characteristics may
covary with training outcomes. In particular, computer experience, computer
anxiety, gender, primary language, and cognitive ability of learners may be

important factors to consider. These variables are discussed in more detail.

35

i "
ECi'i :

ii

3:”
(1'

WW

32K.

Computer anxiety, or “computerphobia,” consists of negative feelings
towards computers that affect one’s ability to use and understand them (Murrell &
Sprinkle, 1993; Ray & Minch, 1990). These feelings are likely to be related to
learning outcomes in web-based training (Cornell & Martin, 1997). Learners who
feel anxious about situations tend to lose focus on the task and Instead focus on
worry (Wine, 1977). Furthermore, computer anxiety has been found to be related
to declarative knowledge gained in a computer training program (Martocchio, E
1994). Therefore, in a WBT context, computer anxiety should be a covariate of

outcomes.

 

Computer experience should be negatively related to computer anxiety.
Computer experience, which can be conceptualized as the amount of time
individuals spend with computers and their degree of comfort with computers, is
likely to decrease feelings of anxiety.

There has been little research regarding the degree to which gender is
related to performance on computer-related tasks. One study of gender
differences in computer-related attitudes found that first-year college students
had no significant differences in computer experience, but did have differing
attitudes; specifically, males felt more comfortable using computers and rated
their skill levels as higher (Wilder, Mackie, & Cooper, 1985). Although this study
would support a prediction for significant gender differences in computer-related
attitudes, there are several reasons why this may not be the case. First,
explanations for differences between the males and females of the sample were

not fully explored. A measure of generalized self-efficacy may have revealed

36

differences that carried over to computer-related attitudes. Second, the data for
this study was collected nearly fifteen years ago. It is likely that with the
increased use of computers in workplaces and schools that the level of comfort
with them may have increased for females, particularly since many traditionally
female—dominated jobs, such as secretarial positions now frequently use
computers.

Current research involving computer training has revealed no significant
differences between males and females in computer experience and training
outcomes (e.g., Martocchio, 1994; Webster & Martocchio, 1995). However, most
studies of learner control in computer training have not included assessments of
the relationship between gender and training outcomes (e.g., Lee & Lee, 1991;
Murphy & Davidson, 1991; Pridemore 8. Klein, 1991). Therefore, I will investigate
gender as a possible covariate of training outcomes.

Most WBT programs require some level of reading ability, especially when
the objective of training is to increase Ieaming in the area of cognitive outcomes
such as declarative knowledge. Therefore, a learner’s primary language is likely
to be a covariate of performance in a WBT program that has a reading element. I
will assess learners’ primary language as a control variable.

Performance in training, and specifically when cognitive outcomes are
assessed, can be a function of cognitive ability. Therefore, I will collect two

measures of cognitive ability as control variables in this study.

37

METHOD

Participants

One hundred eighty-nine subjects participated in this study. Participants
were undergraduate college students recruited from two classes: an upper-level
human resource management course and an upper-level communications
course. The university has categorized both courses as Tier II writing courses,
which means that there must be a significant focus on writing in the assignments.
The grammar skills of these students tend to be low. Therefore, a task in which
students Ieam grammar may contribute to performance in the class, but does not
guarantee it.

Upper-level undergraduate student samples have been used in recent
studies of individual differences in learning, including Colquitt & Simmering
(1998) and Phillips & Gully (1997). In my dissertation, a student sample is
appropriate for two reasons. First, the training material is related to class
performance, and therefore, the participants have a vested Interest in engaging
in the task. Second, the ability for strict control of treatments improves the
internal validity of the research (Cook & Campbell, 1979).

The trainees had incentive to perform in this situation because of possible
rewards in addition to course credit. First, the content of training was directly
related to class assignments. Second, the grammar skills that participants

learned could be used in assignments for other classes. Finally, I offered three

38

$20 gift certificates to a local store to be randomly awarded to three study
participants.

Eighty-four percent of the participants were enrolled in the human
resources management class, and 16% were enrolled in the communications
class. The study participants were 66% female; 44% were business majors, 20%
were dietetics majors, 15% were communications majors, 7% were
interdisciplinary social studies majors, 7% were majoring in other areas of
management (e.g., food industry management), and the remaining 15% had
other majors (primarily liberal arts). Participants’ ages were estimated to be, on
average, in the early 203, as is typical for the course.

One concern with using two separate classes is differences between the
two that may be related to variables of interest in the study. I conducted t-tests to
determine any significant differences between the participants in the two classes.
The classes differed significantly on three variables: learner control (t=2.41,
df=40.90, p<.05), major (t=6.49, df=63.02, p<.05), and time spent in training
(t=2.22, df=60.96, p<.05). Because of these significant differences, I created a
dummy variable to account for which class each participant was enrolled in and
entered that variable as a control in some analyses. I also created a variable to
account for any interaction between class and learner control. I used this control

variable in some analyses.

39

Power Analysis

A power analysis is an indication of the degree to which the sample size of
a study is large enough to determine significance of effects. Cohen (1992)
categorizes effect sizes for multiple regression as follows: small = .02, medium =
.15, and large = .35. In my sample, one hundred eighty-nine people provided

useful data for the study. With a sample size of 189 and a significance level of

p<.05, I will be able to find medium effect sizes for multiple regression using up to I:
8 independent variables.
Web-Based Training L

 

The web-based training in this study taught basic grammar related to
business writing (see Appendix A). In order to design WBT and offer content
appropriate to the needs of the participants, l consulted writing assignments from
previous students of the human resource management course. Using these and
other references, I wrote the web-based training to target the most frequent and
basic errors in student writing.

The WBT was divided into three modules: possessives and contractions,
sentence structure, and word choice. Within each module were two to three
topics. Each module included a description of the grammar rules and examples
of correct and incorrect grammar usage. Each module also included three
optional interactive practice exercises and one hyperlink labeled “click here for a
helpful hint." Therefore, learners could access a total of nine practice exercises

and three hyperlinks.

4O

The web-based training was available on a computer server to which all
participants had access from either their homes or from university computer
laboratories. Participants were required to have the uniform resource Iocator
(URL) for the site in order to gain access to it. The participants were asked to
provide their student identification number and class section number, which
allowed for verification that participants were a part of the relevant participant

pool.

Procedure

 

The time line for this study is presented in Figure 3. In the human resource
management course, the web-based training was available to the participants at
any time on any computer connected to the World VIfide Web for three weeks
before the first writing assignment was due. The class instructor emphasized the
usefulness of the web-based training for this assignment, which was an individual
research paper. I

The communications course did not require an individual writing
assignment similar to that in the human resource management class. Therefore,
the measure of transfer of training (the writing assignment) was not available for
some of the participants. The instructors for these classes also emphasized the
usefulness of the web-based training for writing assignments. Participants in the
communications course had access to the web-based training for a period of

approximately one month.

41

Participants in both courses were allowed to access the web-based
training as often as they liked, since there was an option to view the training
without participating in this dissertation research. However, participants were not
allowed to take part in the study more than once or if they had previously viewed
the web-based training.

The structure of the web-based training is pictured in Figure 4. The web-
based training was designed to be approximately 1-2 hours of instruction in
grammar. The length of the training session varied due to the speed at which
individual learners progressed through the training and whether or not they
completed the optional practice exercises and accessed optional hyperlinks.

The study was conducted under one of two conditions: high learner control
and low learner control. In the high learner control condition, the trainees were
given the opportunity to review the three training modules in any order and. the
topics within the modules in any order. Participants were informed that all
practice exercises were optional, but that they should do at least one. In the low
learner control condition, the trainees were required to progress through the
modules in a structured sequence. Participants in the low learner control
condition were told that all practice exercises were optional, but that they should
try to do all of them. In both conditions, the three hyperlinks were optional. This
afforded some learner control, but considerably less control in the second

experimental condition than the first.

42

Figure 3

Study Time Line

Individual writing
assignment given
(HRM class only)

WBT announced in class;
Available to participants
approximately 3-4 weeks

Participants complete WBT

Individual writing
assignments collected
(HRM class only)

Wonderlic Personnel Test
scores collected

Participants debriefed

January
1999

 

Eafly
—_ February
1999

__ February
1999

Late
—— February
1999

—— March 1999

 

—_ March 1999

43

In both conditions, participants were informed that they must complete all
modules. In order to ensure that trainees in the high learner control condition
reviewed all modules before moving to the post-test, the web-based training was
designed to hide the hyperlink to the post-test until the program verified that all
three modules had been accessed. This prevented any problems resulting from a
confound between the amount of content reviewed and the learner control
treatment.

Participants logged onto the training program by entering their university
assigned student number and section number. They were then asked to print out
the consent form, sign it, and return it to the researcher after they completed the
training. The consent form is shown in Appendix B. The training began with
information about the study and offered participants the option of continuing the
study or only viewing the lesson without participating in the study. For those who
decided to participate in the study, directions were given.

Participants were asked to complete several individual differences
measures within the WBT before moving on to training content. Participants
completed conscientiousness, openness to experience, goal orientation,
computer anxiety, and other (e.g., demographic) measures using radio buttons
(see Appendix C). After completing the individual differences and demographic
measures, the trainees completed a 30-item dichotomous choice pre-test to
measure grammar ability. They were then asked to read a paragraph with
grammatical errors and indicate the number of errors. Participants were given no

indication of their performance on this pre-test. Both tests are in Appendix D.

44

Figure 4

Web-Based Training Structure

 

Log-in using

 

student number

 

 

l

 

6

 

 

Consent form

 

 

l

 

 

Random
assignment to
condition

 

 

 

l

 

Individual

 

differences
measures

 

l

 

 

 

Training
content
(3 modules)

 

 

 

 

l

 

 

Pre-test and
pre-test
paragraph

 

 

 

 

Motivation to
Learn and
Legitimacy

measures after
each module

l

 

 

l

 

45

 

Post-test and
post-test
paragraph

 

L

 

 

Affect measure

 

 

 

Trainees were then randomly assigned (by the program) to the learner
control conditions and given directions to continue with the web-based training.
Trainees were not informed as to which condition they were assigned. Within the
web-based training itself, the trainees in the high learner control condition were
given a choice of sequence of the three modules and the sequence of the topics
within those modules. Trainees could not continue on to the post-test until they
had completed all modules. In the low learner control program, the modules were
offered in the same order, and no choice of sequence was given. The trainees
then progressed through the program. During the training, at the end of each
module, trainees’ motivation to learn and perceptions of legitimacy of learning
grammar and of web-based training were measured.

At the completion of all three modules, trainees were given a 30-item
dichotomous choice post-test and a paragraph to read. Both the post-test and
post-test paragraph were identical to the pre-test and pre-test paragraph. After
submitting their post-test scores, but before logging off the training program, the
trainees answered seven affective outcome items. Finally, the trainees were
thanked for their participation and given the researcher's e-mail address in order
to accommodate any further questions.

After the final date of the availability of the web-based training, I
administered a short test of cognitive ability to the participants in separate
classroom periods. I also collected scores from teaching assistants that indicated
the frequency of grammar errors on in writing assignments for each participant

who was enrolled in the human resource management course. The participants

46

in the communications course were not required to complete a similar writing
assignment to that in the other course, so no scores for this outcome were
collected from that group of participants.

After the participants completed the training, I informed them of the
purpose of the study and the two learning conditions in a separate classroom
presentation. Preliminary aggregate results were presented to the participants in
the debriefing session and the three gift certificates were awarded. At that time,

participants were able to ask questions about the study.

Pilot Testing

In the fall semester of the human resource management class, I
conducted pilot testing of the web-based training. This pilot test was used in this
dissertation in order to verify the reliability and validity of several of the
measures. In instances in which new measures were used, I examined the
psychometrics of the scales using the pilot data. However, all information
reported in the Measures section is based only on the psychometric evaluation of

the measures based on data collected during the actual study.

Measures
These measures, excluding the Wonderlic Personnel Test, the transfer of
training measure, and the pre-test and post-test are shown in Appendix C. The

pre-test and post-test are in Appendix D.

47

I used two types of measures in this study: those that have demonstrated
reliability and validity and those that were newly developed for this study or that
lacked consensus regarding the number of items. As a result, factor analysis was
used to assess the dimensionality of these constructs. In these analyses, I used
Principal Axis factor analysis with a varimax rotation. If there were cross-loadings
of items on factors, I evaluated this in two ways. First, I accepted the factor

loading only if it was larger than :30. Second, I would only accept a factor
loading it it was 1.10 larger than the loading of that item on another factor (Dillon

& Goldstein, 1984).

Independent Variables

Conscientiousness. Conscientiousness was measured using Costa &
McCrae's (1992) short form scale (12 items) from the NEO-Pl instrument. The
reported test-retest reliability of this scale is .83. In this sample, the coefficient
alpha reliability was .75.

Openness to Experience. Costa & McCrae’s (1992) short form NEO—Pl
scale of Openness to Experience was used. The test-retest reliability for this 12-
item scale is reported as .80. I performed a factor analysis to determine the
discriminant validity of the conscientiousness and openness to experience
scales. Using principal axis factoring and a varimax rotation, I found that the 12
conscientiousness items and one openness item loaded onto the first factor. The
remaining 11 openness items loaded onto the second factor. Item 5, which is

“Poetry has little or no effect on me,” did not conform to the rest of the scale, and

48

therefore it was dropped. The revised scale, with 11 items, had a coefficient
alpha reliability of .70.

Goal Orientation. The eight-item scale from Button et al. (1996) was used
to assess mastery orientation. Coefficient alpha reliability with this sample was
.76. The eight-item scale from Button et al. (1996) was used to assess
performance orientation. Coefficient alpha reliability with this sample was .78.
Although mastery and performance orientation were previously conceptualized
as ends of a continuum, recent evidence supports the notion that they are two
separate scales (Button et al., 1996). A factor analysis indicates that the items
from these scales load onto discrete factors.

Legitimacy. The legitimacy of training was measured with six items. The
items were written to measure the perceptions of legitimacy of two things:
Ieaming grammar at this point in the participants' education (at the college
undergraduate level) and the use of web-based training. I measured legitimacy at
three time periods, once after each module of the web-based training was
completed. I felt that because this was a malleable attitude, participants’
perceptions may have changed as they progressed through the WBT.

This was a new scale specifically written for this dissertation, and I
performed an exploratory factor analysis on the items (at all three time periods)
to determine the dimensionality of the scale. Using Principal Axis factor analysis
and a varimax rotation, I found three factors. Items 3, 4, and 6 loaded into factor
one with 49.33% variance explained. This item represents the legitimacy of web-

based training. Items 1 and 5 loaded into factor two with 25.76% of the variance

49

explained, which represents the legitimacy of learning grammar. Item 2 loaded
into a third factor. This item is a measure of the legitimacy of using web-based
training for Ieaming grammar. Because this is a one-item measure, I decided not
to use it in further analyses. In analyzing the hypotheses, I used only factors one
and two, which I named “legitimacy of web-based training” and “legitimacy of
grammar,” respectively.

I measured legitimacy at three time periods, once after each module of the
WBT was completed in order to capture any changes in this attitude throughout
training. Unfortunately, administration of this measure at three time periods may
have contributed to participants’ fatigue, as there was considerable missing data
from the time 3 measure. The number of participants who completed the time 3
measure was 83, which was 44% of the sample.

I performed a t-test between the group that completed the time 3 measure
and the group that did not. The t-test revealed that the random assignment to
learner control was related to the completion of the time 3 measure. The
correlation (r=.87, significant at p<.01) between those who had missing data and
the learner control condition indicates that individuals in the high learner control
condition were more likely to skip the time 3 measure. Perhaps these
participants, in being able to see the amount of material left necessary to
complete the training, were more impatient, and therefore skipped answering
these questions. Another variable that was related to the completion of the time 3
measure was the number of practice exercises completed. These two variables

are positively correlated (r=.56, significant at p<.01), which means that

50

participants who completed practice exercises were also more likely to complete
the time 3 measures.

This missing data is problematic, as there are differences between the
groups of participants who completed the time 3 measure and those who did not.
The correlations among the three administrations of each of the legitimacy scales
are above .79, and therefore, I chose to use only the time 1 data from each. This
scale has the least missing data, and therefore will result in the best statistical
power for analyses.

Computer Anxiety. A computer anxiety scale was adapted from Howell,
Mullins, Fisher, Schmitt, and Kozlowski (1998). The full scale includes 17 items
that assess both computer anxiety and Internet anxiety. I chose nine questions
from the scale that I believed were relevant to measuring computer anxiety. The
items related to Internet anxiety were, in general, a measure of a person’s
anxiety for finding information on the Internet. Because my study (and WBT) is
concerned with use of computers in general, and not finding information on the
lntemet, I did not use those eight items.

This measure is relatively new, so I conducted a factor analysis to
determine its factor structure. Using Principal Axis factor analysis and a varimax
rotation, I found that the nine items loaded into one factor. The coefficient alpha
for this scale was .88.

Computer Experience. Participants were asked to estimate the number
of hours they spent per week using (1) computers in general and (2) the World

Wide Web (WWW) for any reason. Additionally, participants were asked to

51

indicate whether or not they had ever taken web-based training. Finally,
participants were asked to indicate their perceptions of their own skills using ( 1)
computers in general and (2) Internet browsers (such as Netscape or lntemet
Explorer).

Gender. Gender was collected as a self-report item. Males were coded as
zero, and females were coded as one.

Primary Language. Participants were asked to indicate the language that
they considered their primary (native) language. Participants were able to
indicate a secondary language to account for bilinguality. English was coded as
zero; others were coded as one to represent English as a second language.

Major. Participants were asked to indicate their major area of study.
Because the web-based training emphasized business style writing and
referenced business situations in examples, whether or not a participant is a
business major could be a covariate. Major was coded as 1 for business majors
and as 0 for non-business majors. .

Cognitive Ability. Participants completed the Wonderlic Personnel Test,
a 50-item, 12-minute test in a separate classroom session. This measure is
frequently used in employment testing to assess cognitive ability.

Academic Achievement. Participants reported their approximate college

grade point average.

52

Dependent Variables

Motivation to Learn. There are several approaches to measuring
motivation to Ieam. One method is to use questions that directly assess a
person's motivation to learn. I adapted four items from Noe &_ Schmitt (1986) to
directly measure these attitudes. While this measure is straightfonrvard and
provides high face validity, it is fairly transparent, and therefore, susceptible to
social desirability bias (Mathieu & Martineau, 1997).

A second method for assessing motivation to learn is based on
Expectancy Theory (Vroom, 1964). This method employs three scales: valence,
instrumentality, and expectancy, which are either used separately or multiplied to
form a composite measure (Mathieu & Martineau, 1997). For these scales, l
adapted items from Lawler & Suttle (1973) to assess valence (7 items),
instrumentality (3 items), and expectancy (2 items). The expectancy approach is
less transparent and allows for more invdepth investigation, as the three scales
account for each element of motivation (Mathieu & Martineau, 1997). Because of
their level of detail and transparency, I expect that since the direct motivation to
learn measure and the VIE composite measure will be positively correlated, they
can provide unique information.

I factor analyzed the direct motivation to Ieam measures and found that
the four items loaded onto one factor. I also factor analyzed the valence,
instrumentality, and expectancy scales. lnstrumentality and expectancy each
loaded onto separate scales. However, valence fell into two factors. Items 3-7

loaded onto factor one (56.09% variance explained), and items 1 and 2 loaded

53

onto factor two (19.40% variance explained). Upon examination of these items, I
determined that items 1 and 2 were different from the others in that they
assessed the valence of classroom performance and not performing well in
training. The instrumentality and expectancy items only referenced training
performance, so assessing the valence of classroom performance was not
relevant. Therefore, I dropped items 1 and 2, making the revised scale five items.
In order to calculate a composite VIE measure, I multiplied the score for
the revised valence scale, the instrumentality scale, and the expectancy scale.

This multiplicative approach follows Vroom’s (1964) conceptualization of

 

motivation, and has been used in previous motivation to learn research (e.g., ;
Mathieu, Tannenbaum, & Salas, 1992).

I measured motivation to learn and VIE at three time periods in order to
capture any changes in motivation over the course of training. However, these
measures were administered with the legitimacy measures, and I encountered
the same missing data problem as with that scale. The correlations among each
of the scales at the three time periods ranged from .69 to .85.

Therefore, I used only the time 1 measure to avoid problems associated
with missing data in the second and third administrations of the measure.

Pre-testhost-test. The pre-test and post-test were the same 30-item
dichotomous choice test (see Appendix D). It included 10 questions from each
module of training, and participants were asked to indicate whether or not the
use of a word in a sentence or a complete sentence was correct or incorrect.

Pilot testing indicated that this format increased the variance of scores over a

54

format in which participants were asked to choose the correct word to put in a
blank in a sentence. The dichotomous 30-item post-test was scored by adding
together the number of correct answers out of 30. The pre-test was administered
before training began, but participants were given no indication of their score.

The mean difference between the pre-test and post-test (using all 30
items) was 2.13. This difference was surprisingly low, and was possibly caused
by ceiling effects (i.e., high scores on the pre-test). In order to understand the
small difference between the pre-test and post-test, I created another pre-test
and post-test score, one that eliminated items in which 90% or more of the
participants answered correctly. However, after removing these items, the mean
difference between the pre-test and post-test did not increase. Therefore, I
reverted to the original score, which used all 30 items.

There are several ways to statistically determine the degree to which
participants in an experiment have learned. Arvey and Cole (1989) discuss four
methods for determining whether or not a training intervention is statistically
significant. Two of these, gain-score design and analysis of covariance, are
relevant to my study. The gain-score design makes use of difference scores,
specifically, the difference between each individual's pretest and posttest score
(Arvey and Cole, 1989). However, this method is based on the assumption that
the pretest and posttest scores are perfectly correlated. The pre-test and post-
test scores are correlated .42 (significant at p<.05) and the pre-test paragraph
and post-test paragraph are correlated .71 (significant at p<.10). Because these

scores are not perfectly correlated, the gain-score approach is not the best

55

method. In contrast, the analysis of covariance approach uses the pretest score
as a covariate when predicting the posttest score, therefore accounting for the
actual correlation between the two tests. This analysis of covariance is equivalent
to multiple regression procedures in which the pretest is used as a control.

Pro-test ParagraphIPost-test paragraph. | used a second measure for a
cognitive outcome that required participants to identify the number of errors in a
paragraph. This paragraph was written to read like a typical business memo. It
included common errors in possessives and contractions, sentence structure,
and word choice. The participants were asked to read the paragraph on the
screen and then indicate, in text boxes, the number of each type of error (e.g.,
number of apostrophes incorrectly used). The scores for the pre-test paragraph
and post-test paragraph were derived by computing the absolute value of the
difference of each answer from the correct answer. These absolute values were
then added together for all questions. Therefore, a higher score indicates that
either the participant did not identify enough errors in the paragraph or identified
correct grammar as faulty, or both.

Affective Outcomes. Affective outcomes such as satisfaction with
training and willingness to participate in future WBT were assessed with eight
items. The coefficient alpha for this scale was .93.

Training Time. The amount of time each participant spent with the
training program was monitored via the computer. The web-based training
recorded the time that the participant first began. The computer also noted the

time that the trainee spent on each module and the time that the trainee ended

56

 

 

the program. One concern with a measure of training time is that participants
may spend time during training doing other activities such as talking with others,
using the restroom, or doing other work. This was not accounted for in the
training time data. However, in the program directions, participants were asked to
complete the training in one sitting.

Transfer of Training. Participants were required to complete an individual
paper assignment for the human resource management course (see Appendix E
for the assignment). The use of grammar in this paper is a measure of the
transfer of training. I reviewed each paper and counted the number of grammar
errors made. I did not count any errors related to content that were not taught in
the web-based training (e.g., misuse of commas). A high score on this measure
indicates poorer performance.

Because the length of these papers varied by participant, I also estimated
the number of words used in each paper. I counted the words used on one full
page of each participant’s paper and then multiplied this by the number of pages
(and partial pages) written. Participants who write more words have more of an
opportunity for grammatical errors. Therefore, by estimating the length of each
paper, this allows for a more fair comparison among participants.

In analyses, I did not create a ratio variable of number of errors divided by
number of words. Instead, I controlled for the length of the paper. Using a ratio
variable reduces the ability to explain relationships. For instance, if the post-test
score is positively related to the ratio of errors to length, this could mean that the

post-test score is capturing variance in the length of the paper and not the

57

number of errors. This is likely, since length of paper and post-test score may be
spuriously correlated. Therefore, in the analysis of transfer of training, I controlled
for the length of each paper.

Use of Practice Exercises. Nine practice exercises (three within each
module) were available for trainees to use for review. All practice exercises were
optional. The computer program unobtrusively measured the number of
exercises to which each participant linked. However, this measure did not
indicate the number of questions within the exercises that participants answered.
Therefore, it is possible that participants simply went to the exercise and returned
to the WBT without actually completing the exercise.

Use of Hyperlinks. Each module contained one hyperlink marked “click
here for a helpful hint.” These hyperlinks offered advice beyond the training that
related to performance on the individual writing assignment. The computer
program unobtrusively measured which hyperlinks each participant accessed.

The three hyperlinks are listed in Appendix A.

58

RESULTS

Data Analysis Strategy

Based on my conceptual model and on my hypotheses, there are two
major outcomes of interest in this dissertation: motivation to Ieam and Ieaming. I
used two measures of each outcome. I used a direct measure of motivation to
learn and the composite variable of valence, instrumentality, and expectancy as
two dependent variables. Additionally, I used the dichotomous choice post-test
score and the post-test paragraph score as separate variables representing
Ieaming. Therefore, in each analysis of direct effects of individual differences, I

used four dependent variables.

Descriptive Statistics

The means, standard deviations, and zero-order correlations are
presented in Table 2. There are several notable descriptive statistics; First, there
are some differences between the two classes that were used as participants.
The communications class participants did not score as well on the post-test
paragraph as the business class participants. This is most likely due to the
content of the paragraph, which is in a business context. It is possible that these
students’ lack of familiarity with this content made the task of identifying errors
more difficult. Another difference between the classes is the affective response
after training. The participants from the business class had more positive

reactions to the training. Again, this may have been due to the business focus of

59

EoEfimE. n * .movq. ._mcomm_u 9: :0 2525555.“ 8%: 2.... mafia E20580 .mmrnz .202

G
e
d
e
u
D.
a
u
‘9
A
e
u.
e
M
a
s

salqeueA roriuoa

 

EOBEELOESE ..eEO.BeN En ices—«Eco Enocflw .2862

N min...

60

flamenco; u ._. .mo.vn. ._mcomm_u 9.28 2323555.; 63m: Em mafia £20580 .omwuz .902

a
a
d
e
u
D.
e
u
19
A
e
u.
a
N
a
s

sionuoo

 

61

Of the content. Neither of these differences is important enough to warrant
exclusion of the communications class sample, but they do present the need for
explanation.

A second set of descriptive statistics that should be explained is the
correlations with learner control. The learner control conditions were scored
Iow=0 and high=1. Therefore, the high negative correlation between learner
control and frequency of practice indicates that participants in the high learner

control condition were less likely to use the practice exercises.

Handling of Missing Data

As in any research study, this experiment lacks some data from
respondents. In this study, there are two types of missing data. First, there are
instances in which individuals did not complete any items of a scale. In this case,
I used the strategy of painrvise deletion to account for this missing data. Vifith this
method, only those participants who have data for the variables being analyzed
are included in the analysis (Roth, 1994). Therefore, if a participant completed all
scales except the conscientiousness scale, his or her data would not be included
in an analysis involving conscientiousness, but would be used in all other
analyses. One exception to this is in the case of missing time 3 data, which I
discussed previously.

Second, there is random missing data from several of the measures.
These missing data were generally one or more items that were not completed

within a scale. None of the missing data appears to follow a pattern. In other

62

words, it is not any particular item that seems to be commonly skipped. The
strategy I chose to accommodate for this missing data was mean substitution. In
mean substitution, the mean value of the variable is entered in place of the
missing data. The mean value that I used in was the mean of that missing item
across all participants. In other words, if the first item of the conscientiousness
scale was missing, I found the mean for that item across all other participants to
use as the replacement value. The mean replacement strategy improves the
statistical power in this study, but it also provides a conservative estimate of
these scores (Roth, 1994). With mean substitution, the mean of the scale
remains the same, while the variance is reduced (Roth, 1994).

I used mean substitution for missing data in the following 11 scales:
conscientiousness, openness to experience, mastery orientation, performance
orientation, computer anxiety, legitimacy of learning grammar, legitimacy of web-
based training, motivation to learn, valence, instrumentality, and expectancy. In
these scales, there are a total of 67 items. For the 189 participants on the 67
items, I used mean replacement for 566 data points, which is 4.47% of this data
from these scales. Mean replacement is typically considered a useful alternative

when less than 10% of the data are missing (Roth, 1994).

Test of Overall Significance of Model
Before testing the hypotheses, I tested the overall significance of the
model using multivariate multiple regression. Multivariate multiple regression

tests the significance of a set of independent variables predicting several

63

dependent variables. This is an appropriate method to use when a model has
correlated multiple dependent variables.

For this analysis, I entered motivation to Ieam, the composite VIE
measure, the post-test score, and the post-test paragraph score as my four
dependent variables. These are all moderately correlated with one another (see
Table 2). I then entered learner control (the experimental treatment) as the
dichotomous factor in this analysis. Finally, I entered the following variables as
covariates (predictors): computer anxiety, class, conscientiousness, gender,
grade point average, hours spent using World Wide Web, legitimacy perceptions,
mastery orientation, openness to experience, performance orientation, pre-test
scores, pre-test paragraph scores, skill using computers, and previous
experience with web-based training.

A test of between-subjects effects for each of the four dependent
variables indicated that motivation to Ieam (F=3.87, p<.05), V x l x E (F=3.50,
p<.05), post-test score (F=8.15, p<.05), and post-test paragraph score (F=13.47,
p<.05) were significantly related to the set of predictors. Therefore, there are

significant effects for the model that should be investigated.

Analysis of Hypotheses
I analyzed the direct effects hypotheses using hierarchical multiple
regression. I entered the relevant control variables in the first step and the

independent variables of interest in the second step.

64

In order to identify possible control variables to be used in the analyses of
each hypothesis, I entered all of the potential control variables first. Those
variables that had significant beta weights were retained and the others were
dropped. Therefore, there were different control variables used for predicting
different dependent variables. When post-test scores were the dependent
variable, I controlled for pre-test scores, pre-test paragraph scores, hours spent
using the World Vlflde Web, and self-reported skill using a computer. When post-
test paragraph scores were the dependent variable, I used computer anxiety,
class, class X learner control, gender, grade point average, and pre-test
paragraph scores as control variables. When predicting motivation to learn (direct
measure), I controlled for gender. Finally, when predicting the VIE composite, I
used gender and hours spent using the World Vlfide Web as control variables.

Hypothesis 1 predicted that motivation to learn (direct measure of
motivation and VIE) would be positively related to learning outcomes (post-test
and post-test paragraph). These results are presented in Tables 3, 4, 5, and 6. I
expected that individuals who had higher levels of motivation would also Ieam
more in training; however, this hypothesis was not supported. The beta weights
indicate that the control variables had significant effects on the dependent

variables, but that motivation and VIE did not.

65

Table 3
Regression Analysis of Motivation to Learn Predicting Post-test Scores
(Hypothesis 1)

 

 

 

 

 

 

 

 

 

 

 

 

B R2 df ARZ F F
Change Overall

Step 1 ‘

Hrs (vwvw) -.16* .24* 4, 162 -- 12.43* 12.43*

Pre-test 35:

Pie-test Para ”13*

Skill -04
Step 2

Mot to Learn -.01 .24* 1, 161 .00 .03 9.89*
Note. * p< .05.
Table 4

Regression Analysis of Motivation to Learn Predicting Post-test Paragraph
Scores (Hypothesis 1)

 

 

 

 

 

 

 

 

 

 

 

 

B R2 df AR2 F . r:
Change Overall

Step 1

Anxiety -.13* .59* 6, 158 -- 37.63” 37.63*

Class 3”:

Class x LC 42*

Gender '-1 2*

G.P.A. --21*

Pre-test Para 50
Step 2
Motto Learn .04 .59*. 1, 157 .00 .66 3228*
Note. * p< .05.

66

Table 5

Regression Analysis of VIE Predicting Post-test Scores (Hypothesis 1)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l3 R2 df AR2 F F
Change Overall

Step 1 _

Hrs (WWW) -.16* 24* 4, 162 --- 12.43* 12.43*

Pre-test .36*

Pre-test Para -.18*

Skill .15*
Step 2

V x l x E -.00 .24* 1, 161 .00 .00 9.89*
Note. * p< .05.
Table 6
Regression Analysis of VIE Predicting Post-test Paragraph Scores
(Hypothesis 1)

p R2 df AR2 F F
Change Overall

Step 1 *

Anxiety «12* .59* 6, 153 3753* 3163*

Class 31*

Class X LC ~12

Gender '-12*

G.P.A. --22*

Pre-test Para 3“
Step 2

V x I x E .04 .59* 1, 157 .00 .49 32.22*
Note. * p< .05.

67

 

 

Next, I predicted that conscientiousness would be positively related to
motivation to learn (H2) and learning outcomes (H3). These results are
presented in Tables 7-10. Hypothesis 2 was not supported. Conscientiousness
was not significantly related to motivation to learn. Hypothesis 3 was also not
supported. Conscientiousness did not significantly predict learning above and

beyond the effects of the control variables.

Table 7
Regression Analysis of Conscientiousness Predicting Motivation to Learn
(Hypothesis 2)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

B R2 df AR2 F F
Change Overall
Step 1
Gender .12 .02 1. 173 --- 3.38* 3.38*
Step 2
Note. * p< .05.
Table 8
Regression Analysis of Conscientiousness Predicting VIE (Hypothesis 2)
is R2 df AR2 F F
Change Overall
Step 1
Gender .19* .06* 2, 170 --- 5.81* 5.81*
Hrs (WWW) --09
Step 2
Consc .13 .08* 1, 169 .02 3.04 4.93*
Note. * p< .05.

68

Table 9
Regression Analysis of Conscientiousness Predicting Post-test Scores
(Hypothesis 3)

 

 

 

 

 

 

 

 

 

 

 

 

is R2 df ARZ F F
Change Overall

Step 1 1

Hrs (WVWV) -.15* .24* 4. 163 12.56* 12.56*

Pre-test .36*

Pre-test Para -.18*

Skill .14
Step 2

Consc .04 .24* 1, 162 .00 .38 10.08*
Note. * p< .05.
Table 10

Regression Analysis of Conscientiousness Predicting Post-test Paragraph
Scores (Hypothesis 3)

 

 

 

 

 

 

 

 

 

 

 

 

I3 R2 df ARZ F F
Change Overall

Step 1

Anxiety . -.14* .59* 6, 159 ~-- 37.88* _ 37.88*

Class 3'1:

Class x LC -11

Gender “-09,.

G.P.A. --19*

Pre-test Para 52
Step 2

Consc -.07 .59* 1, 158 .00 1.59 32.81*
Note. * p< .05.

Hypothesis 4 predicted that conscientiousness and learner control would
interact to predict motivation to learn. Specifically, I expected that when learner

control was high, there would be a stronger relationship between

69

conscientiousness and motivation to learn. I analyzed this hypothesis using
multiple moderated regression. I entered the variables hierarchically in the
following manner. I entered the control variables in the first step, the individual
difference variable (i.e., conscientiousness) in the second step, the direct effect
of the moderator (learner control) in the third step, and the interaction term
(conscientiousness X learner control) in the fourth step. In order to have a
significant interaction, the change in the F statistic for the fourth step must be
significant. The results of the analyses for Hypothesis 4 are presented in Tables
11 and 12. The regression results indicate that the interaction between
conscientiousness and learner control was not significant in predicting motivation

to Ieam.

Table 1 1
Regression Analysis of Conscientiousness X Learner Control Predicting
Motivation to Learn (Hypothesis 4)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l3 R2 df ARZ F F
Change Overall
Step 1
Gender .12 .02 1, 173 --- 3.38 3.38
Step 2
Consc -.02 .02 1, 172 .01 .92 2.15
Step 3
Learner Control -.84 .03 1, 171 .00 .39 1.56
Step 4
Consc x LC .80 .03 1, 170 .01 1.33 1.50
Note. * p< .05.

7O

Table 12

Regression Analysis of Conscientiousness X Learner Control Predicting

 

 

 

 

 

 

 

 

 

 

 

 

 

VIE (Hypothesis 4)
l3 R2 df AR2 F F
Change Overall

Step 1

Gender .19* .06* 2, 170 --- 5.81* 5.81*

Hrs (WWW) --09
Step 2

Consc .12 .08* 1, 169 .02* 3.76* 5.19*
Step 3

Learner Control -.27 .09* 1, 168 .00 .11 3.90*
Step 4

Consc x LC .25 .09* 1, 167 .00 .13 3.13*
Note. * p< .05

Hypothesis 5 posited that individuals who were more open to experience

would also have higher motivation to learn. Table 13 presents the reSults for this

hypothesis using the direct measure of motivation to learn, and Table 14

presents the results using the VIE composite. This hypothesis was not

supported; openness to experience was not significantly related to motivation to

Ieam.

71

 

Table 13
Regression Analysis of Openness to Experience Predicting Motivation to
Learn (Hypothesis 5)

 

 

 

 

 

 

 

 

 

 

 

 

a R2 df AR2 F F
Change Overall

Step 1 '

Gender .15* .02 1, 173 -- 3.38 3.38
Step 2

Openness -.06 .02 1, 172 .00 .67 2.021
Note. * p< .05.
Table 14

Regression Analysis of Openness to Experience Predicting VIE
(Hypothesis 5)

 

 

 

 

 

 

 

 

 

 

 

 

B R2 df AR2 F i:
Change Overall

Step 1

Gender 22* .06* 1, 170 --- 5.81* 5.81*

Hrs (WWW) --09
Step 2

Openness .04 .07* 1, 169 .00 .22 3.93“V
Note. * p< .05.

I also predicted that openness to experience would be positively related
to learning. The results for Hypothesis 6 are in Tables 15 and 16. This hypothesis
was not supported. Openness to experience did not contribute significantly to
post-test scores nor to post-test paragraph scores beyond the effects of the

control variables.

72

Table 15

Regression Analysis of Openness to Experience Predicting Post-test

 

 

 

 

 

 

 

 

 

 

 

Scores (Hypothesis 6)
p R2 df AR2 F F
Change Overall

Step 1 ’

Hrs (vwvw) -.16* .24* 4, 163 --- 12.56* 12.56*

Pre-test 34:

Pre-test Para '-13*

Skill -14
Step 2

Openness .10 .24* 1, 162 .01 1.86 10.47*
Note. * p< .05.
Table 16

Regression Analysis of Openness to Experience Predicting Post-test
Paragraph Scores (Hypothesis 6)

 

 

 

 

 

 

 

 

 

 

 

B R2 df ARZ F F
Change Overall

Step 1

Anxiety -.12* .59* 6, 159 -- 37.88* 37.88*

Class 10*

Class X LC -11 *

Gender --12*

G.P.A. ~22"

Pre-test Para -61
Step 2

Openness .05 .59* 1, 158 .00 .79 32.53*
Note. * p< .05.

73

 

 

In addition to the direct effects hypotheses relating openness to
experience to the outcome variables, I expected an interaction between
openness to experience and learner control. In Hypothesis 7, I predicted that
when learner control was high, openness to experience would be more positively
related to motivation to learn. In Table 17, I present the results for the interaction
predicting the direct measure of motivation to Ieam. The results for the interaction
predicting the VIE composite is in Table 18. Hypothesis 7 was not supported; the

interaction terms were not significant in either analysis.

Table 17
Regression Analysis of Openness to Experience X Learner Control
Predicting Motivation to Learn (Hypothesis 7)

 

 

 

 

 

 

 

 

 

 

 

 

 

is R2 df AR2 F F
Change Overall

Step 1

Gender .15 .02 1. 173 --- 3.38 3.38
Step 2

Openness -.07 .02 1, 172 .00 .67 2.02
Step 3

Learner Control -.09 .03 1, 171 .00 .37 1.47
Step 4

Open x LC .04 .03 1, 170 .00 .01 1.09
Note. * p< .05.

74

 

Table 18

Regression Analysis of Openness to Experience X Learner Control

Predicting VIE (Hypothesis 7)

 

 

 

 

 

 

 

 

 

 

 

 

 

B R2 df AR2 F F
Change Overall

Step 1

Gender .22* .06* 2, 170 --- --- ---

Hrs (WWW) -.09
Step 2

Openness .03 .07* 1, 169 .00 .22 3.93*
Step 3

Learner Control -.07 .07* 1, 168 .00 .21 2.98*
Step 4

Open x LC .03 .07* 1. 167 .00 .00 2.37*
Note. * p< .05.

Hypotheses 8, 9, and 10 examine the relationship between mastery

orientation and training outcomes. In Hypothesis 8, I expected that mastery

orientation would be positively related to motivation to learn. These results are

presented in Tables 19 and 20. This hypothesis was supported. Mastery

orientation was significantly related to motivation to learn (direct measure) and

the VIE composite beyond the effects of the control variables.

I predicted that mastery orientation would be positively related to post-test

and post-test paragraph scores. These results are presented in Tables 21 and

22, respectively. Mastery orientation did not contribute significantly to learning

beyond the effects of the control variables. Therefore, mastery orientation is not

predictive of post-test and post-test paragraph scores.

75

 

Table 19

Regression Analysis of Mastery Orientatio

n Predicting Motivation to Learn

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(Hypothesis 8)
a R2 df ARZ F F
Change Overall
Step 1 '
Gender .14* .02 1, 173 --- 3.38* 3.38*
Step 2
Mastery Orient .24* .08* 1, 172 .06* 10.52* 7.04*
Note. * p< .05.
Table 20
Regression Analysis of Mastery Orientation Predicting VIE (Hypothesis 8)
p R2 df ARZ F F
Change Overall
Step 1
Gender .23* .06* 2, 170 --- 5.81* 5.81*
Hrs (WWW) .43
Step 2
Mastery Orient .29* .15* 1, 169 .09* 16.62* 9.77*
Note. * p< .05.
Table 21
Regression Analysis of Mastery Orientation Predicting Post-test Scores
(Hypothesis 9)
p R2 df AR2 F F
Change Overall
Step 1
Hrs (www) -.15* .24* 4, 163 --- 12.56* 12.56*
Pre-test ~35:
Pre-test Para '-13*
Skill -15
Step 2
Mastery -.01 .24* 1, 162 .00 .01 9.98*
Note. * p< .05.

76

 

 

 

Table 22
Regression Analysis of Mastery Orientation Predicting Post-test Paragraph

Scores (Hypothesis 9)

 

 

 

 

 

 

 

 

 

 

 

 

p R2 df AR2 F F
Change Overall

Step 1

Anxiety -.12* .59* 6, 159 --- 37.88* 37.88*

Class -1 1:

Class x LC -12 *

Gender '-1 1*

em. 22*

Pre-test Para -51
Step 2

Mastery Orient .03 .59* 1, 158 .00 .30 32.37*
Note. * p< .05.

In Hypothesis 10, I expected that learner control would moderate the
relationship between mastery orientation and motivation to Ieam. Specifically, I
expected that when learner control was high, mastery orientation would have a
stronger relationship with motivation to learn. In the low learner control condition,
I expected that mastery orientation would not be as strongly related to motivation
to learn. The results of the moderated multiple regression, presented in Tables
23 and 24, indicate that this hypothesis was not supported. The Interaction was
not significant for either the motivation to learn direct measure or the VIE

composite.

77

Table 23

Regression Analysis of Mastery Orientation X Learner Control Predicting

Motivation to Learn (Hypothesis 10)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I3 R2 df ARZ F F
Change Overall

Step 1

Gender .14* .02* 1, 173 --- 3.38 3.38
Step 2

Mastery Orient .24* .08* 1, 172 .06* 10.52* 7.04*
Step 3

Learner Control -.09 .08* 1, 171 .00 .58 4.88*
Step 4

Mastery x LC .03 .08* 1, 170 .00 .00 3.64*
Note. * p< .05.
Table 24
Regression Analysis of Mastery Orientation X Learner Control Predicting
VIE (Hypothesis 10)

B R’- df AR2 F F
Change Overall

Step 1

Gender .23* .06* 2, 170 --- 5.81* 5.81*

Hrs (WWW) -.11
Step 2

Mastery Orient .31* .15* 1, 169 .08* 16.62* 9.77*
Step 3

Learner Control .15 .15* 1, 168 .00 .23 7.35*
Step 4

Mastery x LC -.18 .15* 1, 167 .00 .05 5.86*
Note. * p< .05.

78

 

 

Hypotheses 11, 12, and 13 predict the relationship of performance
orientation with the training outcomes. These predictions are opposite of the
mastery orientation hypotheses; I expected performance orientation to be
detrimental to motivation and learning. In Hypothesis 11, I predicted that
individuals with higher levels of performance orientation would have lower levels
of motivation to learn. The results of these analyses, as presented in Tables 25
and 26, indicate a significant relationship that is opposite to predictions.
Performance orientation was positively related to the direct measure of
motivation to learn. However, this effect was not found using the composite VIE

measure as the dependent variable.

Table 25
Regression Analysis of Performance Orientation Predicting Motivation to
Learn (Hypothesis 1 1)

 

l3 R2 df AR2 F F

 

 

 

 

 

 

 

 

 

 

 

Change Overall
Step 1
Gender .14 .02 1, 173 --— 3.38 3.38
Step 2
pens Orient .19* .06* 1, 172 .04* 6.73* 5.11*
Note. * p< .05.

79

Table 26
Regression Analysis of Performance Orientation Predicting VIE
(Hypothesis 1 1)

 

 

 

 

 

 

 

 

 

 

 

 

is R2 df ARZ F F
. Change Overall

Step 1

Gender .23* .06* 2, 170 --- 5.81* 5.81*

Hrs (VWVVV) “-03
Step 2

perf Orient .10 .07* 1, 169 .01 1.84 4.50*
Note. * p< .05.

I expected that performance orientation would be negatively related to
performance on the post-test and post-test paragraph (Hypothesis 12). The
results for this regression analysis are presented in Tables 27 and 28. This
hypothesis was not supported. Performance orientation was not significantly
related to‘either measure of learning. In fact, the beta weights for performance
orientation in these analyses were positive, which is the opposite of the
hypothesized direction.

Hypothesis 13 predicted that the interaction between learner control and
performance orientation would be significantly related to motivation to learn. The
results of the moderated regressions predicting the direct measure of motivation
to learn and the VIE composite are presented in Tables 29 and 30, respectively.
This hypothesis was not supported; the interaction term did not contribute
significantly beyond the direct effects of learner control and performance

orientation.

80

Table 27
Regression Analysis of Performance Orientation Predicting Post-test

Scores (Hypothesis 12)

 

 

 

 

 

 

 

 

 

 

 

B R2 df AR2 F F
Change Overall

Step 1

Hrs - W -.15* .24* 4, 163 --- 12.56* 12.56*

Pre-test 36:

Pre-test Para "13*

Skill -15
Step 2

perf Orient .03 .24* 1, 162 .00 .13 10.02*
Note. * p< .05.
Table 28

Regression Analysis of Performance Orientation Predicting Post-test
Paragraph Scores (Hypothesis 12)

 

 

 

 

 

 

 

 

 

 

 

is R2 df AR2 F _ F
Change Overall

Step 1

Anxiety -.13* .59* 6, 159 --- 37.88* 37.88*

Class -1 1:

Class X LC -12 *

Gender '-1 1*

G.P.A. --21*

Pre-test Para 51
Step 2

Perf Orient '.04 .59* 1, 158 .00 .71 32.51*
Note. * p< .05.

81

 

 

Table 29

Regression Analysis of Performance Orientation X Learner Control

Predicting Motivation to Learn (Hypothesis 13)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

B R2 df AR2 , F F
Change Overall

Step 1

Gender .14 .02 1, 173 '" 3.38 3.38
Step 2

perf Orient .10 .06* 1, 172 .04* 6.73* 5.11*
Step 3

Learner Conn-0| -.59 .06* 1, 171.00 .31 3.50*
Step 4

'36er LC .56 .06* 1, 170 .01 1.16 2.91*
Note. * p< .05.
Table 30
Regression Analysis of Performance Orientation X Learner Control
Predicting VIE (Hypothesis 13)

l3 R2 df AR2 F F
Change Overall

Step 1

Gender .23* .06* 2, 170 --- 5.81* 5.81*

Hrs (VWWV) --08
Step 2

Perf Orient .16 .07* 1, 169 .01 1.84 4.50*
Step 3 *

Learner Control .37 .08* 1, 168 .00 .1 1 3.39
Step 4 *

perf X LC -.40 .08* 1, 167 .00 .59 2.82
Note. * p< .05.

82

 

 

Hypotheses 14, 15, and 16 addressed the relationships among the
perceptions of legitimacy and the training outcomes. As described previously, I
split the legitimacy items into two separate measures: ( 1) perceptions of
legitimacy of learning grammar and (2) perceptions of legitimacy of web-based
training.

Hypothesis 14 predicted that perceptions of legitimacy would be positively
related to motivation to learn. These results, which are presented in Tables 31—
34, indicate that this hypothesis was partially supported. Perceptions of
legitimacy of learning grammar were positively related to the direct measure of
motivation to learn and the VIE composite. However, perceptions of legitimacy of

web-based training were not significantly related to motivation to learn.

Table 31
Regression Analysis of Perceptions of Legitimacy of Learning Grammar
Predicting Motivation to Learn (Hypothesis 14)

 

 

 

 

I3 R2 Df AR2 F F
Change Overall
Step 1
Gender .15* .02* 1, 172 --- 4.07* 437..
Step 2

 

 

 

 

 

 

 

Legit Grammar .46* .23* 1, 171 .21* 45.87* 2550*

 

Note. * p< .05.

83

Table 32

Regression Analysis of Perceptions of Legitimacy of Learning Grammar
Predicting VIE (Hypothesis 14)

 

 

 

 

 

 

 

 

 

 

 

I3 R2 Df ARZ F F
Change Overall

Step 1 -

Gender .23* .07* 2, 169 --- 5.85* 5.85*

Hrs (WWW) --05
Step 2

Legit Grammar .30* .15* 1, 168 .09* 17.77* 10.21*
Note. * p< .05.
Table 33

Regression Analysis of Perceptions of Legitimacy of WBT Predicting

Motivation to Learn (Hypothesis 14)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

p R2 Df AR2 F F
Change Overall
Step 1
Gender .21* .04* 1, 172 --- 7.76* 7.76*
Step 2
Legit WBT .04 .05* 1, 171 .00 .29 . 4.01*
Note. * p< .05.
Table 34
Regression Analysis of Perceptions of Legitimacy of WBT Predicting VIE
(Hypothesis 14)
is R2 or AR2 F F
Change Overall
Step 1
Gender .23* .07* 2, 169 --- 5.85* 5.85*
Hrs (VWWV) --09
Step 2
Legit WBT .06 .07* 1, 168 .00 .66 4.11*
Note. * p< .05.

84

 

 

 

Hypothesis 15, which predicted that perceptions of legitimacy would be
positively related to learning, was not supported. The results, which are
presented in Tables 35-38, indicate that neither the perceptions of the legitimacy
of learning grammar nor web-based training were related to post—test and post-

test paragraph scores.

Table 35
Regression Analysis of Perceptions of Legitimacy of Learning Grammar
Predicting Post-test Scores (Hypothesis 15)

 

 

 

 

 

 

 

 

 

 

 

 

l5 R2 Df AR2 F i:
Change Overall

Step1
HIFS(WWW) --15* -24* 4.162 12.43* 12.43*
Pre-test 35: _
Pre-test Para ”18.
Skill .15
Step2
Legit Grammar .01 .24* 1,161 .00 .01 9.89*
Note. * p< .05.

85

Table 36
Regression Analysis of Perceptions of Legitimacy of Learning Grammar
Predicting Post-test Paragraph Scores (Hypothesis 15)

 

 

 

 

 

 

 

 

 

 

 

 

p R2 or ARZ F F
Change Overall

Step 1 ‘

Anxiety -.13* .59* 6, 158 --- 37.63* 37.63*

Class -1 1:

Class X LC -12 *

Gender '-1 1*

G.P.A. --21*

Pre-test Para 5"
Step 2

Legit Grammar -.02 .59* 1, 157 .00 .15 32.11"
Note. * p< .05.
Table 37

Regression Analysis of Perceptions of Legitimacy of WBT Predicting Post-
test Scores (Hypothesis 15)

 

 

 

 

 

 

 

 

 

 

 

 

l5 R2 Df AR2 F i:
Change Overall

Step 1

Hrs (VWWV) . -.16* .24* 4, 162 --- 12.43* 12.43*

Pre-test 35:

Pre-test Para '-13*

Skill -13
Step 2

Legit WBT -.10 .25* 1, 161 .01 2.16 10.45*
Note. * p< .05.

86

Table 38
Regression Analysis of Perceptions of Legitimacy of WBT Predicting Post-
test Paragraph Scores (Hypothesis 15)

 

 

 

 

 

 

 

 

 

 

 

 

is R2 Df AR2 F F
Change Overall

Step 1 '

Anxiety -.13* .59* 6, 158 --- 37.63* 37.63*

Class ~10:

Class X LC -12 *

Gender '-1 1 *

G.P.A. --21*

Pre-test Para -61
Step 2

Legit WBT -.04 .59* 1, 157 .00 .45 3221*
Note. * p< .05.

Hypothesis 16 predicted that perceptions of legitimacy and learner control
would interact to predict motivation to learn. The results of the moderated
regression analyses, as presented in Tables 39-42, indicate that this hypothesis
was partially supported. In this analysis, the interaction term for learner control X
perceptions of legitimacy of web-based training was significant beyond the
effects of the control variables and the direct effects.

A plot of this significant interaction appears in Figure 5. As predicted,
when learner control was high, perceptions of legitimacy were more strongly
related to motivation to learn. Specifically, in the high learner control condition,
more positive perceptions of legitimacy were related to higher motivation to learn.
Conversely, when learner control was low, the perceptions of legitimacy were not
very strongly related to motivation to learn. There was an average level of

motivation to learn regardless of the perceptions of legitimacy. This follows

87

Mischel’s (1977) contention that in weak situations, personal characteristics have
more of an impact. In other words, in the weak situation of high learner control, a
person’s perceptions of legitimacy and motivation to learn are more strongly

related.

Table 39
Regression Analysis of Perceptions of Legitimacy of Learning Grammar X
Learner Control Predicting Motivation to Learn (Hypothesis 16)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

B R2 df AR2 F F
Change Overall

Step 1

Gender .15* .02* 1, 172 --- 4.07* 4.07*
Step 2

Legit Grammar .42* .23* 1, 171 .21* 45.87* 25.50*
Step 3

Learner Control -.27 .23* 1, 170 .00 .06 16.93*
Step 4

Legit Grammar .26 .23* 1, 169 .00 .20 12.69*

X LC
Note. * p< .05.

88

Table 40

Regression Analysis of Perceptions of Legitimacy of Learning Grammar X
Learner Control Predicting VIE (Hypothesis 16)

 

 

 

 

 

 

B R2 AR2 F F
Change Overall

Step 1

Gender .23* .07* 2, 169 --— 5.85* 5.85*

Hrs (WWW) --04
Step 2

Legit Grammar .38* .15* 1, 168 .09* 17.77* 10.21*
Step 3

Learner Control .56 .15* 1, 167 .00 ‘ .03 7.62*
Step 4

Legit Grammar -.57 .16* 1, 166 .01 .91 6.28*

X LC

 

 

 

 

 

 

 

Note. * p< .05.

Table 41

Regression Analysis of Perceptions of Legitimacy of WBT X Learner

Control Predicting Motivation to Learn (Hypothesis 16)

 

 

 

 

 

 

 

 

 

 

 

 

 

B R2 df AR2 i: F
Change Overall
Step 1
Gender .17* .02* 1, 172 --- 4.07* 4.07*
Step 2 *
Legit WBT -.11 .06* 1, 171 .04* 6.60* 5.41
Step 3 *
Learner Control -1.81* .06* 1, 170 .00 .29 3.69
Step 4 * * *
Legit WBT x 1.82* .16* 1. 169 .09 18.84 7.76
LC
Note. * p< .05.

89

 

 

Table 42

Regression Analysis of Perceptions of Legitimacy of WBT X Learner
Control Predicting VIE (Hypothesis 16)

 

 

 

 

 

 

 

 

 

 

 

 

 

is R2 df AR2 F F
‘ Change Overall

Step 1

Gender .23* .07* 2, 169 --- 5.85* 5.85*

Hrs (VWWV) --09
Step 2

Legit WBT -.04 .07* 1, 168 .00 .66 4.11*
Step 3

Learner Contro| -.62 .07* 1, 167.00 .14 3310"
Step 4

Legit WBT x .61 .08* 1, 166 .01 1.93 2.88*

LC
Note. * p< .05.

90

 

Figure 5

Plot of Perceptions of Legitimacy of WBT X Learner Control Predicting
Motivation to Learn

 

Legitimacy of WBT X Learner Control
DV: Motivation to Learn

4.5 -—
4.4

//
21? /

 

 

 

l“'"""" ._ -L
39 / I OW

 

 

 

. —.— High
3.8 /

3.7
3.6
3.5
3.4 , ,
Low High
Perceptions of Legitimacy of WBT

 

Motivation to Learn

 

 

 

 

 

 

 

91

Additional Hypotheses

Hypothesis 17 predicted that trainees in the high learner control condition
would spend less time on training overall than those in the low learner control
condition. Learner control and time spent in training were correlated r= -.12 (not
significant). The direction of this correlation indicates that trainees in the high
learner control condition spent less time in training. A t-test of the mean
difference of time spent in training between learners in the high learner control
condition and learners in the low learner control condition indicated that there
was no significant difference between the means. Therefore, this hypothesis was
not supported.

Hypothesis 18 predicted that individuals in the high learner control
treatment would have more positive affective outcomes than individuals in the
low learner control treatment. Learner control was not significantly correlated with
affective outcomes. Therefore, this hypothesis was not supported.

Finally, in Hypothesis 19, I predicted that learning would be related to
transfer of training. Specifically, I predicted that the scores on the post-test and
on the post-test paragraph would be negatively related to the number of errors in
each paper, controlling for length of paper. I used length of paper as a control
variable rather than creating a ratio variable of errors divided by length, because
the use of a ratio variable for this type of analysis does not provide the
appropriate information (Cohen & Cohen, 1983).

The results of this analysis are in Tables 43 and 44. The sample size for

this analysis was 132, as only the business course required the assignment that

92

was used for the transfer of training measure. The results of these analyses
indicate that both post-test and post-test paragraph scores were significantly
related to transfer of training. In Table 43, the negative beta weights for post-test
scores predicting transfer of training indicate that higher learning scores were
related to a smaller number of grammar errors in the assignment. In Table 44,
the beta weight is positive, indicating that the number of errors on the post-test
paragraph was positively related to the number of errors in the subsequent

assignment. Therefore, this hypothesis was supported.

Table 43

Regression Analysis of Post-test Scores Predicting Transfer of Training
(Hypothesis 19)

 

 

 

 

 

 

 

 

 

 

 

 

[3 R2 Df AR2 F F
Change Overall

Step 1

Paper Length '.04 .01 1, 130 "" 1.25 1.25
Step 2 '

Post-test -.29* .09* 1, 129 .08* 11.53* 6.44*
Note. * p< .05.
Table 44

Regression Analysis of Post-test Paragraph Scores Predicting Transfer of
Training (Hypothesis 19)

 

 

 

 

 

 

 

 

 

 

 

 

[a R’- Df AR2 F F
Change Overall
Step 1 *
Paper Length -.04 .01 1 , 130 --- 1.25* 1.25
Step 2
Post-test Para .30 .10‘ 1, 129 .09* 12.81* 7.09*
Note. * p< .05.

93

A summary of these results is presented in Table 45. Of the 19
hypotheses, two were fully supported, and two were partially supported.
Specifically, mastery orientation was positively related to motivation to learn, and
learning was positively related to transfer of training. There was also partial
support for the effects of legitimacy. The perceptions of legitimacy of Ieaming
grammar were positively related to motivation to Ieam, and the perceptions of
legitimacy of web-based training interacted with learner control to predict

motivation to Ieam.

Exploratory Analyses

in this study, I collected data on several variables without making direct
hypotheses about them. These variables were tangential to the main research
question and conceptual model. I took an exploratory approach to these

analyses; they were performed post hoc.

Mediation Analysis

The conceptual model for this dissertation (see Figure 1) implies a
mediated model, with individual differences predicting learning through
motivation to learn. While this is conceptually logical, the data do not support this
relationship. Although several of the individual differences are significantly

related to motivation to learn, motivation to learn did not significantly predict

94

Table 45

Summary of Hypothesized Results

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Hypothesis Predicted Relationship Outcome

1 Motivation -+ Learning Not supported

2 Conscientiousness —> Motivation Not supported

3 Conscientiousness —9 Learning Not supported

4 Conscientiousness X Learner Control Not supported
-) Motivation

5 Openness —+ Motivation Not supported

6 Openness —> Learning Not supported

7 Openness X Learner Control —+ Not supported
Motivation

8 Mastery Orientation —> Motivation Supported

9 Mastery Orientation —> Learning Not supported

10 Mastery Orientation X Learner Not supported
Control -—> Motivation .

11 Performance Orientation —> Not supported
Motivation

12 Performance Orientation —> Learning Not supported

13 Performance Orientation X Learner Not supported
Control —> Motivation

14 Legitimacy —-> Motivation Partially supported

15 Legitimacy —-) Learning Not supported

16 Legitimacy X Learner Control -> Partially supported
Motivation

17 Learner control -—> Time Not supported

18 Learner control —> Affective outcomes Not supported

19 Learning —+ Transfer of training Supported

 

 

 

95

 

learning beyond the effects of the control variables. Therefore, a mediation
analysis was not appropriate for the direct motivation to Ieam and composite VIE
measures. I also examined the effects of the separate valence, instrumentality,
and expectancy scales on the post-test and post-test paragraph scores. None of

the effects were significant, and again, a mediation analysis was unwarranted.

Use of Practice Exercises and Hyperlinks

Two unique measures that I obtained from my dissertation were the
frequency of the use of practice exercises and hyperlinks within the web-based
training. These measures may provide useful information about trainees’
motivation to learn. While i already measured motivation to learn directly and with
valence, instrumentality, and expectancy components, those measures were
self-reported. The frequencies of the use of practice exercises and the use of
hyperlinks were unobtrusive measures of motivation to Ieam. Therefore, I
believed that it would be interesting to see to what degree they contributed to
learning outcomes above and beyond the control variables.

The results for these four exploratory analyses are presented in Tables
46, 47, 48, and 49. I found two significant direct effects for these unobtrusive
measures. The number of practice exercises that a trainee used was significant
in predicting performance on the post-test but was not related to performance on
the post-test paragraph. Whereas the direct measure of motivation to learn and
the VIE composite were not significantly predictive of learning, the use of

hyperlinks was significantly related to performance on both the post-test and the

96

post-test paragraph. This indicates that the unobtrusive measures of motivation
were predictive of Ieaming outcomes beyond the effects of the control variables.
Table 46

Regression Analysis of Frequency of Use of PracticeExercises Predicting
Post-test Scores

 

 

 

 

 

 

 

 

 

 

 

 

[3 R2 df AR2 F F
Change Overall

Step 1

Hrs (www) -.19* .24* 4, 154 -- 12.12* 12.12*

Pre-test 33:

Pre-test Para ”21*

Skill -16
Step 2

Practice .12 .26* 1, 153 .02 3.12 10.45*
Note. * p< .05.
Table 47

Regression Analysis of Frequency of Use of Practice Exercises Predicting
Post-test Paragraph Scores

 

 

 

 

 

 

 

 

 

 

 

 

p R2 df AR2 F F
Change Overall

Step 1

Anxiety -.13* .77* 5, 151 --- 43.11" 43.11*

Class 3'1:

Gender 512*

G.P.A. --20*

Pre-test Para -52
Step 2

practice -.03 .77* 1, 150 .00 .31 3531*
Note. * p< .05.

97

Table 48

Regression Analysis of Frequency of Use of Hyperlinks Predicting Post-

 

 

 

 

 

 

 

 

 

 

 

test Scores
[3 R2 df AR2 F F
Change Overall

Step 1

Hrs (VWWV) -.13* .24* 4, 163 --- 12.56* 12.56*

Pre-test ~33:

Pre-test Para '-1 5*

Skill -15
Step 2

Hyperlinks .16* .26* 1, 162 .02* 5.00* 11 .29*
Note. * p< .05.
Table 49

Regression Analysis of Frequency of Use of Hyperlinks Predicting Post-
test Paragraph Scores

 

 

 

 

 

 

 

 

 

 

 

p R2 df AR2 F F
Change Overall

Step 1 '

Anxiety -.11* .57* 5, 160 -- 43.17* 43.17*

Class -1 0

Gender 40*

G.P.A. ”20*

Pre-test Para ~50
Step 2

Hyperlinks -.13* -.13* 1, 159 .02* 6.55* 3832*
Note. * p< .05.

98

 

 

DISCUSSION

Overview of Discussion

This discussion is divided into four parts. First, I summarize the results of
the hypothesized and exploratory analyses and consider the implications of these
results for theory building. Second, I address the limitations of the study. Third, i
provide suggestions for future research in this area. Finally, I consider the

practical application of these results.

Summary of Results and Implications for Theory Building

This dissertation was aimed at understanding the extent to which
individual differences of learners contributed to Ieaming in different training
conditions. Within the context of web-based training, i found that some individual
differences influenced learning in both high learner control and low learner
control training conditions.

The first hypothesis predicted that motivation to learn would be
significantly related to Ieaming outcomes. Surprisingly, this hypothesis was not
supported. Noe's (1986) model of motivational influences in training targeted
motivation to Ieam as critical to training success. Many studies have found
support for the positive link between motivation and learning (e.g., Colquitt &
Simmering, 1998; Quinones, 1995, Warr & Bunce, 1995). Additionally, a recent

meta-analysis of motivation to learn found that motivation to learn was

99

moderately related to declarative knowledge and skill acquisition (Colquitt,
LePine, and Noe, under review).

In this study, neither measure of motivation to learn was significantly
related to learning. It appears that the majority of the variance explained in
learning outcomes was due to pre-test and pre-test paragraph scores. This
indicates that, in this case, ability was more predictive than motivation. However,
exploratory analyses revealed that two other measures of motivation to learn
were significantly related to learning outcomes beyond the effects of the pre-test
and pre-test paragraph.

The practice exercises and hyperlinks in this study can be construed as
unobtrusive measures of motivation to learn. At no time were participants led to
believe that the number of practice exercises used or hyperlinks visited would be
known to the researcher. Those participants who voluntarily chose to engage in
practice exercises and follow hyperlinks were likely to have been motivated to
Ieam the training material.

The measure of the frequency of the use of practice exercises was
positively related to performance on the post-test. This result is not surprising, as
the exercises were very similar in content and style to the post-test. The practice
exercises appeared to enhance learning beyond the effects of the control
variables. I would expect this result in future studies.

I was surprised, however, to find that the use of hyperlinks was
significantly related to both post-test and post-test paragraph scores. The content

of the hyperlinks was completely unrelated to any material on either post-test

100

measure (see Appendix A). Therefore, I believe that the use of hyperlinks is a
meaningful measure of motivation to Ieam. This measure is unobtrusive; the
trainees did not know that the computer could track their movement within the
training.

It is interesting to note that the use of hyperlinks and the number of
practice exercises accessed were not significantly correlated with either the
direct measures of motivation to learn or the VIE composite. Furthermore, the
two unobtrusive measures were not significantly related to the elements of VIE:
valence, instrumentality, and expectancy. This leads me to believe that perhaps
the self-report data on the direct motivation to learn measures may not have
been accurate. Mathieu 8. Martineau (1997) have noted that direct motivation to
Ieam measures tend to be transparent and susceptible to responses based on
impression management or social desirability.

Although VIE measures are generally believed to be less transparent, the
results of this study imply that this may not be the case. This could be because
the participants have previously been exposed to expectancy theory. The
students in the human resource management course are required to take a
course in organizational behavior either before or concurrent with the HRM class.
This organizational behavior course includes the topic of expectancy theory, in
which all three elements are fully explained. Therefore, with this group of
participants, it is possible that the VIE measures were also transparent.

An alternate explanation for the participants’ use of hyperlinks may be

curiosity. Openness to experience, which in part measures curiosity, was

101

significantly correlated with the use of hyperlinks (r=.17, p<.05). This correlation
may have been higher had the reliability of the openness to experience scale
been higher.

Although the links between the self-reported measures of motivation to
Ieam and learning were not supported, it is still interesting to examine the
relationship of the individual differences on motivation to learn and learning. I
conducted analyses to examine the extent to which the set of individual
differences in this study accounted for variance in the dependent variables. I
regressed each of the dependent variables on conscientiousness, openness to
experience, mastery orientation, performance orientation, legitimacy of grammar,
and legitimacy of WBT. Beyond the effects of the control variables, the
independent variables accounted for 25% (p<.05) of the variance in motivation to
learn, 14% (p<.05) of the variance in the VIE composite, 3% (non-significant) of
the variance in post-test scores, and .8% (non—significant) of the variance in post-
test paragraph scores. Therefore, the individual differences had a much stronger
effect on motivation than on learning outcomes.

When I examined the effects of each individual difference separately, only
one, mastery orientation, was related to learning. The results indicate that
trainees who focus on gaining competence are more likely to do well on learning
outcomes. This finding adds to the growing literature that reports strong support
for the role of mastery orientation in learning outcomes (e.g., Colquitt &

Simmering, 1998; Phillips 8: Gully, 1997).

102

l was surprised to find that none of the other individual differences were
positively related to motivation to learn and learning. Conscientiousness and
openness to experience, two of the Big Five factors of personality, were not
predictive of motivation to learn or learning. in their meta-analysis of motivation to
learn, Colquitt et al. (under review) found a positive relationship between
conscientiousness and motivation to learn. This was not replicated in my study.
However, I did find a significant correlation (r=.15, p<.05) between
conscientiousness and expectancy. Expectancy, which is an individual's belief in
his or her ability to perform a specific task, is indistinguishable from self-efficacy.
Colquitt et al. (under review) found a positive relationship between
conscientiousness and self-efficacy and between self-efficacy and motivation to
learn. Therefore, this study may provide indirect evidence for the relationship
between conscientiousness and motivation to learn.

The relationship between openness to experience and learning outcomes
has not been previously tested; however, Barrick 8. Mount (1991) indicated that
this factor may be strongly related to performance in training.

There are several possible explanations for these non-significant
results. The self-reported measures of motivation to learn appear to have been
weak. There is probable restriction in range due to the effects of social
desirability. There also is possible restriction of range in the individual difference
measures. Because the sample was students who were typically in their third or
fourth year of college, these individuals are more likely to be more conscientious

than the general public. This sample may not have had a different level of

103

openness to experience than other samples, but the lack of results may have
been due to the low reliability of this measure.

The non-significant results for the performance orientation measure may
be due to the nature of the training. The WBT was designed to be done
individually and privately. Although it is possible, it is unlikely that many of the
participants had others witnessing their performance in training. Therefore,
inclinations to perform well in front of others should not have had an effect in this
sfiuafion.

The results for the perceptions of legitimacy indicated partial support of
the hypothesized relationships. Two significant results were found. First, the
perceived legitimacy of Ieaming grammar was positively related to motivation to
learn as predicted. Second, learner control moderated the relationship between
the perceptions of legitimacy of WBT and motivation to learn such that when
trainees had high learner control, there was a stronger relationship. in other
words, trainees in the low learner control condition had average levels of
motivation to Ieam, despite their perceptions of the legitimacy of WBT. However,
those individuals who had high learner control and found the training to be
legitimate were much more motivated to learn.

These results present the first investigation of how perceptions of
legitimacy relate to motivation to learn. Because this concept was adapted from
organizational theory literature, it has not been previously examined in the
context of training. This dissertation, therefore, introduces an important new

element to understanding what motivates trainees to learn.

104

I predicted that trainees in the high learner control condition would have
more positive affective outcomes and spend less time in training. The results
indicated that trainees were not more likely to have positive affective outcomes.
Perhaps this is because the level of learner control in this study did not allow for
much trainee influence. In a situation in which learners have significantly more
control, particularly over the content of training, high learner control may be
related to affective outcomes. Similarly, there were no significant differences in
the time that trainees in the two conditions spent in training. Again, this was likely
due to the weak manipulation of learner control in this study.

My final hypothesis predicted that learning would be related to transfer of
training. This hypothesis was supported—trainees who scored better on the post-
test and the post-test paragraph made fewer errors in actual writing assignments.
Additionally, the post-test paragraph scores contributed slightly more incremental
variance explained than the post-test scores. This may be because transfer of
training tends to be stronger when there is a perceived similarity between the
training and job tasks (Ford & Kraiger, 1995). In this training, the requirements of
the post-test paragraph were more similar to the writing assignment than the

post-test was.

Limitations
One concern when using students for research participants is that their
characteristics are not necessarily representative of individuals in other

organizations. They may differ in the following ways: age, education, and

105

organizational constraints. First, the average age of this sample was estimated to
be approximately 21 years of age. Organizations are very likely to have a higher
mean age. Second, each participant had at least two years of college education,
but probably no more. Individuals in organizations will have a broader range of
education. Finally, although these participants experienced a number of
constraints related to training (e.g., coursework, jobs, etc), these may not be
similar to constraints experienced by individuals within an organization.

I used samples from two separate courses for this study. This afforded
more statistical power, but there were several significant differences between the
two groups. I statistically controlled for these group differences in analyses.
However, this indicates that the results of these analyses may have differed due
to the participants. Therefore, these results are not necessarily generalizable to
other samples.

Another limitation of this dissertation is that participation was voluntary,
and therefore, some individuals within the class did not participate in the study.
There may be some differences between those participants who did engage in
the study and those who did not. Those who did not participate may have had
lower conscientiousness, and therefore, were not interested in achieving a higher
score on their written assignments by going through the grammar lesson.
Alternatively, the highly conscientious students could have decided that they
were skilled enough in grammar to not need the lesson. Although there may be
differences between the group of participants and non-participants, the result of

any differences would be to restrict the range of the independent variables within

106

the sample. That is, there would be less variance on some individual difference
variables. This may have led to several of the non-significant findings in this
study.

Similarly, the two learner control conditions were, not very different. While
the definition of learner control indicates that control can vary on three
elements—content, pacing, and sequence—l only varied sequence. This
restriction was due to internal validity constraints, but still prevents a full
investigation of learner control. However, as discussed above, any results found
with learner control would only be stronger if the variance of this were to be
greater.

Missing data was a limitation of this study. I measured motivation to learn,
expectancy, valence, and legitimacy after each module. Although participants
were informed that they would be asked to answer the same questions several
times, only 38 percent completed the time 3 measures. In the analyses, I used
only the time 1 measures of these scales to avoid the missing data problem.
However, useful information regarding motivation and legitimacy was lost due to
participant fatigue. Additionally, the affective measures may have been affected
by dissatisfaction with the study requirements. In the future, one or two time
periods should be sufficient for this information.

Another concern is the method for handling missing data within the
individual differences scales. I used mean replacement to reduce the amount of
missing data, which is an acceptable strategy. However, I replaced each data

point with the mean of that item of all other participants. Another strategy is to

107

replace the data point with the mean of the other scale items for that participant.
The method that I used for mean replacement is more conservative and is more
likely to reduce the variance. Therefore, replacing the mean in another way may
influence the results of this study.

One potential limitation of this study is the absence of information due to
the data collection procedure. Because trainees were able to complete training
on their own time, outside of the view of any training administrator, there is no
guarantee that trainees gave their full attention to the course. For example,
trainees could have continued through the content without actually reading it.
Additionally, trainees could have accessed the pages of practice exercises
without actually completing them. Without knowing more about participation in
practice exercises, interpretations of those results must be conservative. Another
concern is that participants could have enlisted the help of others in answering
questions. Although this is unlikely, it cannot be discounted.

Finally, one result of concern is the small mean difference between the
pre-test scores and post-test scores. As a measure of Ieaming, these scores
were somewhat weak. I investigated the possibility that there were ceiling effects,
but did not believe that this was the case. I believe that a problem with the tests
is that they were dichotomous choice. It is likely that trainees can identify correct
and incorrect grammar usage in a dichotomous choice situation, but still have
problems with their own writing. However, the positive relationship between post-
test scores and the frequency of grammar errors indicates that there is a positive

relationship between the testing situation and the actual performance context.

108

Future Research
The findings in this dissertation emphasize the need for future research
into several key areas: motivation to learn, learner control, goal orientation,

legitimacy, transfer of training, and training media.

Motivation to Learn

The main implication from this dissertation is the need to empirically
investigate differences in measures of motivation to Ieam. Although there have
been comparisons of several self-reported measures (Mathieu & Martineau,
1997), my results indicate that there needs to be a comparison of self-reported to
unobtrusive measures. Web-based training is a medium that supports the
investigation of unobtrusive motivation to Ieam measures, and therefore, it may
be a useful platform for discovering more about the validity of these scales.

Another avenue of investigation in this area is the benefit of using
hyperlinks in WBT. In this study, trainees were offered three hyperlinks to aid
them in the WBT that provided no content to aid in the cognitive outcome
measures. Further studies should investigate the benefits of using hyperlinks with
training content. Because hyperlinks are unique to web-based training, their use
can provide information necessary to fully evaluate web-based training as an
appropriate medium for certain training content. Hyperlinks could be very useful
for training related to business, because they could link learners to current

business publications, organizations’ web pages, or government statistics.

109

Learner Control

One limitation of this study was the weak manipulation of learner control.
A possible confound with content prevented me from being able to emphasize
the differences in learner control between the two conditions. Future studies may
be able to emphasize the differences between high and low learner control
conditions. There are two possibilities for doing so.

One method may be to inform trainees of the level of learner control. In
this study, trainees were not only unaware of which learner control condition they
experienced, but they also did not know that the two conditions existed. In future
studies, informing trainees that they have either more or less control than others
may be a useful strategy. One limitation with this method is that the manipulation
would have to be checked by measuring the perceived learner control. However,
insight into the effects of perceived learner control would also contribute to the
literature.

The learner control manipulation could also be made stronger by using a
more involved training situation. The training in this study was very short (about
one hour), but in a longer, more intense training session, the sequencing and
pacing may become more important to learners. For example, if training requires
a full eight-hour day, trainees may be more sensitive to the sequencing of
material in order to relieve boredom and increase motivation. Additionally,
different training content may make sequencing more important to learners.

Pacing, another element of learner control, may also influence the

strength of the manipulation. In this study, there were not significant pacing

110

differences between the two conditions of learner control. However, if pacing can
save a significant amount of time, perhaps a manipulation of pacing may be
stronger. For example, if a low learner control situation would require a trainee to
engage in training for eight hours versus four hours in a high learner control
situation, the effects of learner control on motivation and learning might be
increased.

In summary, this study unfortunately has not contributed much information
to the already divided literature regarding learner control. Although the results
suggest that learner control makes little difference to motivation or Ieaming, the
weak manipulation may not have been effective in finding any actual effects for
the training conditions. In order to appropriately assess the effects of learner

control, a stronger manipulation of learner control is warranted.

Goal Orientation

Another topic that should be addressed by future research is goal
orientation. In this study, mastery orientation emerged as the only individual
difference related to motivation to learn. Additionally, although the results for the
performance orientation regression analyses were not significant, there is a
significant, positive correlation between performance orientation and the direct
measure of motivation to Ieam. This correlation indicates that predictions that
performance orientation would be negatively related to motivation to Ieam were
opposite of what was found. This may be due, as mentioned previously, to the

private nature of the task. One element of a performance orientation is the need

111

to demonstrate competence to others. Because the web-based training was
private, there was little or no opportunity for trainees to perform in front of others.
Future research should investigate the relationship between performance
orientation and motivation to Ieam in settings that are not private. For instance, if
this relationship had been studied using a classroom presentation of grammar
skills in which trainee participation was encouraged, perhaps a stronger, positive
effect would have been found for performance orientation predicting motivation to
Ieam.

Another avenue of research for goal orientation may be to analyze the
effects of mastery orientation and performance orientation together. In this study,
I examined the effects of these variables on the dependent variables separately.
However, because mastery and performance orientation are considered to be
independent of one another, it would be feasible to enter both into the same
regression equation to investigate their effects. In particular, an examination of
the interaction between mastery and performance orientation would be
interesting.

I investigated these effects with exploratory analyses. However, the
interaction of the goal orientation measures was not significant for any of the
dependent variables (motivation to Ieam, VIE, post-test score, and post-test
paragraph score). However, as stated previously, this task was private, and
therefore, it may be difficult to find effects for performance orientation. Future
research should be done to try to identify a possible interaction between mastery

orientation and performance orientation.

112

Legitimacy

Legitimacy is a new concept in training, and because of this, there should
be increased attention to understanding its role in training. This study indicates
that legitimacy is a significant predictor of motivation to learn. However, because
of the participants in the study and the content of the training, these effects need
to be replicated.

Although perceptions of legitimacy appear to be high in this group of
participants (mean=4.38 for grammar and 4.03 for WBT), this may be due in part
to range restriction. It is likely that university students are more likely to find
Ieaming situations, in general, to be more legitimate. Furthermore, university
students may use computers and the World \Mde Web more frequently than the
general public. Therefore, these perceptions of legitimacy may be lower with
employees in general than they are presented here. However, if there is range
restriction in this case, I would expect a stronger relationship between
perceptions of legitimacy and motivation to Ieam, as range restriction attenuates
correlations.

The sample in this study may have higher mean perceived legitimacy than
other samples because this training was voluntary. It is likely that only those
individuals who felt that the content and medium were legitimate would
participate. Therefore, perceptions of legitimacy should be investigated in
situations where trainees are assigned to training. The larger variance in

perceptions could produce stronger effects.

113

The content of the training is also likely to play a role in the mean level of
perceived legitimacy of training. Grammar skills may be seen as a remedial topic.
The perceptions of legitimacy may change for tasks that are new to the learner.
However, the use of web-based training for Ieaming new tasks may not be
believed to be as favorable. Therefore, these perceptions should be investigated
in other contexts.

In this study, I did not evaluate the origins of perceptions of legitimacy.
According to organization theorists, legitimacy can stem from regulations, norms,
or cognitive structures (Scott, 1991) or density (Harman & Carroll, 1992).
However, because legitimacy has not been previously investigated in a training
situation, there is not enough information about where it originates. Future
research should try to identify factors that are related to perceptions of
legitimacy. With this understanding, researchers and practitioners can attempt to

influence organizational and training characteristics to improve perceptions of

legitimacy.

Transfer of Training

In this dissertation, I evaluated the transfer of training and found a positive
relationship between learning and transfer. However, because this study was
conducted in a university environment, it may not be analogous to a work
sfluafion.

In their model of transfer of training, Baldwin 8. Ford (1988) indicated the

importance of the work environment on transfer of training. The work

114

environment, or the climate for transfer, may be more easily identified and
measured in an actual organizational setting. However, with a student sample,
this would be very difficult. This study may have provided a positive climate for
transfer, as the instructors of the course emphasized the importance of the web-
based training for course performance and use in other courses. However,
because the climate for transfer was not measured, this remains unknown.
Future research should focus on antecedents of transfer of training other
than just performance in training. The characteristics of the trainees, design
elements of the training, and the work environment may all play a part in the

degree to which Ieaming transfers to the actual job (Baldwin & Ford, 1988).

Training Media

Web-based training is a unique training medium that allows for
unobtrusive data collection not available through other media. The results of my
study indicate the importance of the unobtrusive measures. Therefore, I would
suggest that future studies make use of the benefits of this medium. However,
because WBT has not been compared to other training media, this is also a gap
in the training literature.

Web-based training is a relatively new medium, and as such, has not been
compared to other training media in published research. Two types of
comparisons would benefit research in this area. First, a test of WBT versus
another self—directed learning medium (e.g., printed materials) would provide

information regarding the benefits of each, independent of the level of learner

115

control. Second, a test of WBT versus a non-self-directed learning environment
(e.g., lecture) would provide a global comparison of each method.

One interesting construct that may provide insight into choice of training
medium is legitimacy. This dissertation suggests that perceptions of legitimacy
contribute to motivation to learn. It is likely that legitimacy will vary depending on
the training medium, as some training media are more common than others.
Additionally, the content of training may moderate the relationship between the

medium and perceptions of legitimacy.

Practical Application

There is limited practical application from this study, since there were few
significant effects and small effect sizes (Cohen, 1992). However, there are two
areas in which there may be practical application from the results of this
dissertation: the importance of the needs assessment and the perceptions of
legitimacy.

The results of this study emphasize the need for needs assessment
before training. Because some individual differences are related to motivation to
learn and learning outcomes, these should be assessed before training is
designed. In particular, a mastery orientation measure, if offered as part of the
person analysis, may indicate the degree to which trainees are ready to learn.

Improving perceptions of legitimacy may improve training. In order to
improve perceptions of legitimacy of Ieaming particular training content,

managers must emphasize the link between the trained content and the trainee’s

116

job. The perceptions of the legitimacy of web-based training may be dependent
on trainees’ experiences with web-based training. Therefore, increased exposure

to this new technology may improve motivation to Ieam.

117

APPENDIX A

WEB-BASED TRAINING

This appendix presents the web-based training in paper form. There are
three content modules: (1) Possessives and Contractions, (2) Sentence
Structure, and (3) Word Choice. Each module has three practice exercises. |
consulted four texts in writing this web-based training: Basic Language Skills
(1983), Kaufman (1985a, 1985b), Terban and Marvin (1993).

The presentation of this training on the World Vifide Web allowed for
access of pages in differing order in the high learner control condition. However,
in paper format, this is not possible. Therefore, the content is presented by
module in the order in which it was presented in the low learner control
treatment. The measures of motivation to Ieam and legitimacy that were given
after each module are presented in Appendix C.

In this appendix, each set of double lines indicates a new page, or screen,
of information in the web-based training. Within the WBT, a hyperlink at the
bottom of each page that read “continue” allowed the trainee to move to the next

Page of content.

118

Possessives and Contractions Module

 

Possessives

A possessive noun tells who or what owns something.

To make a singular word possessive, add an apostrophe and an “5.”
Example: That is the manager's briefcase.

Example: The file cabinet’s drawers are locked.

 

There are two ways to make a plural word possessive:

If the word ends in “3,” add only an apostrophe.
Example: Three managers' offices are being painted today.
Example: These four computers' keyboards are new.

If the plural does not end in an “5,” add an apostrophe and an “5.”
Example: The men's restroom needs to be cleaned. I

Example: The new children's wing of the hospital opens this week.

 

 

If the word is a name or a proper noun, the same rules for possessives apply.
Example: Pat's assistant answered the telephone.
Example: We have to consider Korea's import tax.
If a person’s name ends in an “3”, add an apostrophe and an “s.” The same rules
apply to last names.
Example: Chris’s schedule is very busy today.

Example: Mrs. Davis’s computer hasn’t been working correctly.

119

If you are referring to a group of people by their family (or last) name, add an
apostrophe after the “s.”
Example: The Robinsons’ house isn't for sale.

Example: The reception will be held at the Burrows’ house.

 

Pronouns

Pronouns are words that take the place of nouns.

Possessive pronouns show ownership but do not use apostrophes.
Example: I put his report on the desk.
Example: That computer program is theirs.
Example: That diskette is mine; its label has my name on it.

Example: I'm not sure whose it is.

 

Contractions
A contraction is when two small words are combined into one with an
apostrophe to form one word.
Example:
It is very nice of you to help me with this project.
It’s very nice of you to help me with this project.
Example:
They are coming to the meeting today.

They’re coming to the meeting today.

120

Example:
Who is in charge of scheduling the meeting?

Who's in charge of scheduling the meeting?

Click here for a helpful hint.

 

Helpful Hint:
The MSU Writing Center helps students improve their grammar and writing style
in their class assignments. This is an excellent way for students to get hands-on
help with their writing. You can visit the Writing Center in 300 Bessey Hall or
contact them on the Internet. You can find the MSU Writing Center's web page
by linking to it from the MGT 310 web page.

Return to lesson

 

To decide whether or not an apostrophe is necessary in a word, follow this rule:
Read words with an apostrophe as if it were two separate words. For example,
read “it's” as “it is” in the sentence, and see if the sentence still makes sense.
Here are some comparisons between contractions and possessive pronouns.

It’s vs. its

“it’s” is a contraction of “it is”
It’s an excellent compensation plan.

We have to decide if it’s a program that is worth pursuing.

121

“its” is a possessive pronoun
Its main purpose is to provide a wide range of benefits.
We have to decide if its cost savings are large enough to continue
the program.
Who’s vs. whose
“who’s” is a contraction of “who is"
Who's supposed to make a presentation at the meeting?
I'm not sure who's responsible for the mistake.
“whose” is a possessive pronoun
Whose training manual is this?
I don’t know whose mistake it was.
they’re vs. their
“they’re” is a contraction of “they are”
They’re recommending that we move the plant to Michigan.
If they think we need more training, they’re wrong.
“their" is a possessive pronoun
Their plan for growth is outstanding.

Retaining 90% of employees is their goal.

122

Possessives and Contractions

Exercise One

Directions: For each sentence, choose the correct word to fill in the blank by
using the pull-down menu. Then click on “Check Answer” to see if you are
correct. A small gray box that says “JavaScript Alert” will pop up, indicating

whether or not your answer was correct.

 

 

I think a great plan. (pull-down menu)
0 Its
0 it's

Check Answer

 

 

 

 

The restroom is around (pull-down menu)
the corner. 0 womens Check Answer
. women’s
. womens’

 

 

 

 

 

 

fifteen minutes late for (pull-down menu)
the meeting. . You’re
. Your

 

Check Answer

 

 

 

I need to give the memo to Ms. (pull-down menu)
. 0 Jones

. Jones’

. Jones’s

 

 

 

Check Answer

 

 

 

I’m not sure going to (pull-down menu)
attend the meeting. 0 Whose Check Answer
. who’s

 

 

 

 

 

 

 

books are on the table. (pull-down menu)
. Marcus Check Answer

. Marcus’s
o Marcuses’

 

 

 

 

 

 

 

123

 

Possessives and Contractions

Exercise Two

Directions: For each sentence, choose the correct word to fill in the blank by
using the pull-down menu. Then click on “Check Answer" to see if you are

correct. A small gray box that says “JavaScript Alert” will pop up, indicating

whether or not your answer was correct.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The computer malfunctioned (pull-down menu)

because I spilled water on 0 its Check Answer

keyboard. . it’s

The office is down the hall (pull-down menu)
0 boss
. boss’s Check Answer
. bosses
. bosses’

The decision was a good (pull-down menu)

one. . groups Check Answer
. group s
0 groups’

Those three projects are (pull-down menu)

due today. 0 employees Check Answer
a employee’s
- employees’

I think planning a (pull-down menu)

meeting today. . there Check Answer
. they’re
0 their

The car is in the repair (pull-down menu)

shop. . Reynolds Check Answer
- Reynold’s
- Reynolds’

 

124

 

 

 

 

 

 

 

Possessives and Contractions

Exercise Three

Directions: For each sentence, choose the correct word to fill in the blank by
using the pull-down menu. Then click on “Check Answer” to see if you are

correct. A small gray box that says “JavaScript Alert” will pop up, indicating

whether or not your answer was correct.

 

The book isn’t mine; it’s

(pull-down menu)
. yours

. your’s
. yours’

 

 

Check Answer

 

 

I‘m not sure where office

will be.

(pull-down menu)
. you’re
. your

 

 

Check Answer

 

 

Several names were left

off the list.

(pull-down menu)
0 peoples
o people’s
0 peoples’

 

 

Check Answer

 

 

computer wasn’t
working?

(pull-down menu)
. Whos

Who's
o Whose
o Whos”

 

 

Check Answer

 

 

The pages are numbered.

(pull-down menu)
0 books

- book’s
. books'

 

 

Check Answer

 

 

That desk is

 

 

(pull-down menu)
0 Ross

. Ross’s
o Rosses
o Rosses’

 

 

Check Answer

 

 

 

 

125

 

 

 

 

 

 

 

Sentence Structure Module

 

Subject and Verb Agreement
A singular subject is one in which one person or thing is named. The verb must
match the subject.

Example: She works eight hours each day at the office.
A plural subject is one in which more than one person or thing is named. The
verb must match the subject.

Example: They work eight hours each day at the office.

 

Subject and Verb Agreement
Subject and verb agreement becomes more difficult when there is a phrase
between the subject and verb.
For correct agreement, read the subject and verb without the phrase to see if it
makes sense. 1
Example: The list of options presented in the meeting is not complete.
Example: Neither of the invited speakers is able to attend.
Sentence Fragments
A sentence fragment is a group of word that does not express a complete
thought and cannot stand by itself. A complete sentence must have a subject and

a main verb.

126

Examples of sentence fragments:
Sending the paychecks in the mail today.
(There is no subject in this sentence. This doesn’t tell who sends
the paychecks.)
Omar, who is the new VP of Marketing for our department.
(There is no main verb for this sentence. This describes who Omar
is, but not what action he takes.)
Sometimes a complete sentence appears to not have a subject, but is still
correct.
Example: Give me that pen.
This is an imperative sentence—it makes a request or gives a command. In an
imperative sentence, the subject is understood to be you even though it is not
written.

Example: (You) Give me that pen.

 

Sentence Fragments

Often clauses are mistaken for complete sentences. A clause is a group of words
that has a subject and a predicate.

Subjects are nouns, pronouns, or phrases used as nouns. They tell what the
sentence is about—a person, thing, or idea. Predicates tell us about the subject,

what the subject does or is.

127

There are two types of clauses:
Main clauses can stand alone as sentences. Subordinate clauses are
used with main clauses to express a related idea.
Examples of clauses:
My supervisor let me have the day off.
(This is a main clause.)
My supervisor let me have the day off so that I could take my son to the
doctor.
(This is a main clause plus a subordinate clause.)
So that I could take my son to the doctor.
(This is a subordinate clause, and alone, is a sentence fragment.)

Click here for a helpful hint.

 

Helpful Hint:
The MSU library has many books related to grammar. Use the computerized
library search system to find reference books on grammar skills, writing style,
and citation formats. You can find the library’s resources with a link on the MGT
310 web page.

Reign to lesson

 

128

Sentence Fragments
Sentence fragments can be corrected by adding a main subject and predicate.
Here are some examples of sentence fragments and how they can be corrected.
Sentence Fragment: Unless I get a call from you.
Complete Sentence: I won’t go unless I get a call from you.
Sentence Fragment: Assuming you have the information you need.
Complete Sentence: You can finish the project, assuming you have the
information you need.
Sentence Fragment: A large project, with many employees, clients, and
other stakeholders involved.
Complete Sentence: This is a large project, with many employees, clients,
and other stakeholders involved.
Sentence Fragment: Julie, who was working in that department for months
as a part-time employee.
Complete Sentence: Julie, who was working in that department for months

as a part-time employee, was just hired as a full-time employee.

 

 

Run-on Sentence
A run-on sentence is a sentence that is really two or more sentences that run

together without the proper punctuation to join them.

129

Here are some examples of run-on sentences:
Example: Bring a pencil you will have to take notes.
Example: I want to reduce the budget for research, however, my colleague
does not.
Example: I will ask Bill he is a friend of mine.

Example: Several of us are going, we want to get involved.

 

Run-on Sentences
There are several ways to fix run-on sentence.
First, you must decide what the separate sentences are. Break them apart into
two individual sentences.

Run-on Sentence: Take this pen I don’t need it.

Sentence 1: Take this pen.

Sentence 2: I don't need it.
You can leave the two sentences as separate to eliminate the run-on sentence.
Second, choose how you will join the sentences: with a semi-colon or with a
comma and a conjunction.

Run-on Sentence: Take this pen I don’t need it.

Sentence with a semi-colon: Take this pen; I don’t need it.

Sentence with a comma and a conjunction: Take this pen, because I don’t

need it.

130

Using a comma to join two sentences is not correct. If you can break the
sentence into two complete sentences where the comma is, you should use a
semi-colon instead of a comma.
Run-on Sentence: I want to start the meeting, however, she would like to
wait five minutes.
Comect Sentence: I want to start the meeting; however, she would like to

wait five minutes.

 

_ @
—_

131

Sentence Structure

Exercise One

Directions: For each sentence, decide whether it is grammatically correct or
incorrect. Then click on “Check Answer” to see if you are right. A small gray box

that says “JavaScript Alert” will pop up, indicating whether or not your answer

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

was right.

Generally speaking, I don’t like (pull-down menu)

using e-mail. 0 Correct Check Answer ‘
. Incorrect

I don't think that’s a good idea (pull-down menu)

it’s too time consuming. . Correct Check Answer .
o Incorrect

I want to go to the meeting; (pull-down menu)

however, I have another 0 Correct Check Answer .

commitment. . Incorrect

The list of options are too long; I (pull-down menu)

can’t decide. . Correct Check Answer
0 Incorrect

Neither of the managers is (pull—down menu)

willing to lead this project. . Correct Check Answer
0 Incorrect

Assuming you don’t have a lot of (pull-down menu)

work to do today. 0 Correct Check Answer
0 Incorrect

 

 

 

 

 

 

 

 

132

 

 

 

 

 

 

 

Sentence Structure

Exercise Two

Directions: For each sentence, decide whether it is grammatically correct or
incorrect. Then click on “Check Answer” to see if you are right. A small gray box
that says “JavaScript Alert” will pop up, indicating whether or not your answer

was right.

 

A group of managers is waiting (pull-down menu)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

in your office. . Correct Check Answer
- Incorrect

The project will be difficult, (pull-down menu)

however, I have confidence that . Correct Check Answer

we can do It. , Incorrect

Because I have to go through (pull-down menu)

several weeks worth of memos . Correct Check Answer

that accumulated during my . Incorrect

vacation.

Give me a chance to review the (pull—down menu)

report. 0 Correct Check Answer
a Incorrect

We began the project last month, (pull-down menu)

new we have to rush to finish it. 0 Correct Check Answer
- Incorrect

Permission to enter the (pull-down menu)

executive suite of offices is rarely 0 Correct Check Answer

9W9": o Incorrect

 

 

 

 

 

 

133

Sentence Structure

Exercise Three

Directions: For each sentence, decide whether it is grammatically correct or
incorrect. Then click on “Check Answer” to see if you are right. A small gray box

that says “JavaScript Alert” will pop up, indicating whether or not your answer

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

was right.

A set of keys are missing from (pull-down menu)

the office. . Correct Check Answer
. Incorrect

Let me do it. (pull—down menu)
0 Correct Check Answer
. Incorrect

There are many factors to (pull-down menu)

consider like the cost which I 0 Correct Check Answer

think is extremely high. . Incorrect

For which we want to thank you. (pull-down menu)
. Correct Check Answer
a Incorrect

We might need the report take it (pull-down menu)

with you to the meeting. 0 Correct Check Answer
. Incorrect

Neither of us needs to attend the (pull-down menu)

meeting. . Correct Check Answer
0 Incorrect

 

 

 

 

 

 

134

Word Choice Module

 

Homonyms
Homonyms, or homophones, are words that are spelled differently and have
different meanings, but are pronounced alike.
The following are several common homonyms:
to is a preposition that means “toward”
Example: I will go to the meeting.
too is an a verb that means “also” or “in addition to”
Example: I should go to the meeting too.
two is a noun that means the number between one and three

Example: There are two meetings today.

 

f —
_ 4— —

 

Homonyms
weather is the condition of the atmosphere
Example: The weather is supposed to be nice tomorrow.
whether means “if" or “either"

Example: I’m not sure whether or not I’ll go to the meeting.

_ —
- —

135

Homonyms
role is a position in a company or a part in a movie or play
Example: My role as a manager is limited.
roll is a list of names

Example: I will call roll before we begin.

 

Homonyms
through means “in one side of something and out the other”
Example: I will go through the meeting agenda before we begin.
threw is the past tense of the verb “to throw”

Example: I threw a party for the employee who is retiring.

 

 

Homonyms
, brake is a device for stopping a vehicle
Example: I had to press on the brake to stop.
break is a verb that means to make something come apart or a noun that
means “to rest”

Example: After this presentation, we can take a break.

136

Homonyms
there means “at or in that place”
Example: I think we should meet there.
their means “belonging to them”

Example: We need their permission to buy the computer.

 

 

Related Words
These words sound and look very much like each other. Because of this they are
easily confused and misused.
The following are pairs of words that are commonly misused:
affect is a verb that means “to influence or change”
Example: This plan should affect our productivity.
effect is a noun that means “a result or consequence”

Example: The effect of the plan was an increase in productivity.

 

Related Words
moral is an adjective that means “good behavior or character”
Example: It is moral to donate money to charity.
morale is a noun that means “the attitude or spirit of a person or group”
Example: The holiday bonus checks should raise the morale of the

employees.

II

137

Related Words
lose is a verb that means “to misplace or fail to win"
Example: If we lose the account, I won’t get a commission.
loose is an adjective that means “not firmly attached”
Example: The parts on this widget are loose.

Click here for a helpful hint.

 

 

Helpful Hint:
If you are using a computer to write your paper, you can take advantage of the
spell checking and grammar checking features. However, remember that the
spell checker does not catch words that are spelled correctly, but misused. Make
sure that you proofread your paper even if you use the spell checker.

Rej_urn to lesson

 

 

Related Words
personal is an adjective meaning “private"
Example: I need to miss work for a personal reason.
personnel is a noun that refers to people working in a business
Example: We had several complaints from personnel in the branch

office.

 

138

Related Words
accept is a verb meaning “to take what is offered or given”
Example: I will accept a job offer if they make one.
except is a preposition that means “leaving out” or “other than”
Example: I agree with everything except your last point.
Related Words
then is an adverb meaning “at that time”
Example: We will discuss the project specifications then.
than means “in comparison with”

Example: I like the first idea more than the second one.

 

 

139

 

Word Choice

Exercise One

Directions: For each sentence, choose the correct word to fill in the blank by
using the pull—down menu. Then click on “Check Answer" to see if you are
correct. A small gray box that says “JavaScript Alert” will pop up, indicating

whether or not your answer was correct.

 

 

 

 

 

 

 

 

 

I need to “put the ” on (pull-down menu)

this type of behavior before it . breaks Check Answer

reduces productivity. , brakes ' ‘ ‘

What do you think the of (pull-down menu)

the labor strike will be? . effect Check Answer
. affect

The proposal is a (pull-down menu)

interpretation of what we did last 0 lose Check Answer

year. . loose

 

 

 

 

After the three-hour meeting,

(pull-down menu)

 

 

 

 

 

 

 

 

 

 

 

 

we’ll need a . break Check Answer
. brake

Can you call before we (pull-down menu)

beginthe training session? 0 roll Check Answer
. role

I hope we don’t need it, because (pull-down menu)

I that document away. . through Check Answer
- threw

 

 

 

 

 

140

 

 

 

 

 

 

 

Word Choice

Exercise Two

Directions: For each sentence, choose the correct word to fill in the blank by
using the pull-down menu. Then clickon “Check Answer" to see if you are
correct. A small gray box that says “JavaScript Alert” will pop up, indicating

whether or not your answer was correct.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

or not you agree, you (pull-down menu)

have to support me on this. 0 Whether Cheek Answer

. Weather
isn’t any reason to (pull-down menu)

terminate his employment. o Their Check Answer
0 They’re
c There

I think the first applicat we (pull-down menu)

interviewed would make a better 0 than Check Answer

manager the second , then

applicant.

Many of the employees feel that (pulldown menu)

they are busy since we . to Check Answer

took on the new account. , too
0 two

Do you think that we can (pull-down menu)

the conditions of their contract? 0 except Check Answer
. accept

We have a obligation to (pull-down menu)

provide health benefits to our 0 morale Check Answer

employees. , moral

 

 

 

'!

141

 

 

 

 

 

 

 

Word Choice

Exercise Three

Directions: For each sentence, choose the correct word to fill in the blank by
using the pull-down menu. Then click on “Check Answer” to see if you are
correct. A small gray box that says “JavaScript Alert” will pop up, indicating

whether or not your answer was correct.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

We have too few to (pull-down menu)

finisht he project on time. . personal Check Answer
0 personnel *

If we don’t submit a lower bid, (pull-down menu)

we could the client. 0 loose Check Answer
. lose

Employee has suffered (pull—down menu)

ever since the economic e morale Check Answer

recession. , moral

We are having a meeting about (pull-down menu)

that topic next week, so we can c then Check Answer

discuss it , than

for Joe, everyone is (pull-down menu)

here. 0 Except Check Answer
. Accept

I hope that this merger (pull-down menu)

our position in the market. . affects Check Answer
. effects

 

 

 

 

 

 

 

142

 

 

 

 

 

 

 

APPENDIX B
INFORMED CONSENT FORM

This study is an examination of factors that lead to effective learning in a
web-based training program. As a part of this study, you will be asked to
complete a personality survey, participate in a web-based training program to
Ieam more about English grammar, and take a short ability test in class. The
web-based training program will take approximately 1 hour. Additionally, the
grade from your individual paper assignment will be collected (by student

number) at the end of the term.

Your participation in this study is purely voluntary, and you may stop at
any time. If you choose to participate in this study, you will receive credit for
participation points in your MGT 310 class. If you choose not to participate, you

can receive participation points for MGT 310 with in-class activities.

In this study, your responses will be confidential. Although you will be
identified by your PID in the study database, your name will not be associated
with your survey responses or your performance in the training. All data collected
from this survey will be presented in aggregate only, and your name will never be
attached to the information that you provide. Furthermore, although you will
provide answers on the computer, the database for that information is password

protected and available only to the study investigators.

143

As a part of the study, it will be necessary to obtain your grade point
average and SAT/ACT score from the registrar. These are only to be used as
statistical controls, and only the study investigators will see these data. Your
instructor will not see any information related to your grade point average and
SAT/ACT scores. Additionally, your instructors will not see any of your answers
to any of the questions in the personality survey nor in the training program itself.
Any data presented from this study will be in aggregate only, and your name will

not be attached to your responses.

If you participate in this study, you will be debriefed approximately 1-3
weeks after your training to explain the purpose of the study and its preliminary
results. If you have any questions regarding this project, please contact Marcia

Simmering at 517-353-6788, or simmerin@pilot.msu.edu.

Please sign your name below if you would like to participate in this study. By

signing, you indicate your voluntary consent to participate in this study.

Signature

 

 

Name (please print)

Student Number Date

 

 

144

APPENDIX C

MEASURES

All items will be measured on a five-point Likert scale ranging from

"strongly disagree to "strongly agree." Reverse-scored items are indicated by (R).

1 2 3 4 5
Strongly Disagree Neutral Agree Strongly
Disagree Agree

Conscientiousness

1.

2.

9.
10.
11.

12.

I keep my belongings clean and neat.

I'm pretty good about pacing myself so as to get things done on time.

I am not a very methodical person. (R)

I try to perform all the tasks assigned to me conscientiously.

l have a clear set of goals and I work toward them in an orderly fashion.
| waste a lot of time before settling down to work. (R)

I work hard to accomplish my goals.

. When I make a commitment, I can always be counted on to follow

through.

Sometimes I’m not as dependable or reliable as I should be. (R)
I am a productive person who always gets the job done.
I never seem to be able to get organized. (R)

I strive for excellence in everything I do.

145

Openness to Experience

1.
2.

10.

11.

12.

I don’t like to waste my time daydreaming. (R)
Once I find the right way to do something, I stick to it. (R)
I am intrigued by the patterns I find in art and nature.
I believe letting students hear controversial speakers can only confuse
and mislead them. (R)
Poetry has little or no effect on me. (R) (dropped)
I often try new and foreign foods.
I seldom notice the moods or feelings that different environments produce.
(R)
I believe we should look to our religious authorities for decisions on moral
issues. (R)
Sometimes when I am reading poetry or looking at a work of art, | feel a
chill or wave of excitement.
I have little interest in speculating on the nature of the universe or the
human condition. (R)
l have a lot of intellectual curiosity.

I often enjoy playing with theories or abstract ideas.

146

Mastery Orientation

1.

2.

The opportunity to do challenging work is important to me.
When I fail to complete a difficult task, I plan to try harder the next time I
work on it.

I prefer to work on tasks that force me to learn new things.

. The opportunity to learn new things is important to me.

I do my best when I’m working on a fairly difficult task.

When I have difficulty solving a problem, I enjoy trying different
approaches to see which one will work.

On most jobs, people can pretty much accomplish whatever they set out
to accomplish.

Your performance on most tasks or jobs increases with the amount of

effort you put into them.

Performance Orientation

1.

2.

The things I enjoy the most are the things I do the best.

The opinions of others have about how well I can do certain things are
important to me.

I feel smart when I do something without making any mistakes.

I like to be fairly confident that I can successfully perform a task before I
attempt it.

I like to work on tasks that I have done well on in the past.

I feel smart when I can do something better than most other people.

147

7. Even if I know that I did a good job on something, I’m satisfied only if
others recognize my accomplishments.

8. It‘s important to impress others by doing a good job.

Legitimacy

1. I think training on the computer is a good idea.
This training is a legitimate way to learn more about grammar.

I think grammar lessons should be available for college students.

sew

It is legitimate to have college students trained in grammar skills.

9‘

Web-based training is a useful training method.

9’

Grammar skills are important for college students to learn.

Motivation to Learn (direct measure)

1. I will exert considerable effort to learn the training material.
2. Learning the training material is important to me.
3. I will try hard to learn the training material.

4. It is very important to me to learn the training material.

148

Valence

Listed below are a number of events which are related to this training. For each
one, please indicate how desirable it is to you. Please use the following scale.

1 2 3 4 5
Undesirable Neutral , Desirable
1. Getting a good grade in MGT 310/COM 440.
2. Doing well on writing assignments.
3. Achieving success in the training.
4. Doing well in training.
5. Learning the training material.
6. Trying hard.

7. Putting forth effort.

lnstrumentality
Below you will see a number of pairs of factors that look like this:
Example: Warm weather -+ Sweating
Please indicate how true you feel that it is for you personally that the first word
leads to the second word. Please use the following numbers to represent
different feelings about how often the first word leads to the second:
1 2 3 4 5
Rarely Sometimes Usually
1. Learning the training material -> Performing well in training

2. Learning the training material —> Achieving success in training

3. Learning the training material _. Doing well in training

149

Expectancy

Below you will see a number of pairs of factors that look like this:

Example: Warm weather -» Sweating

Please indicate how true you feel that it is for you personally that the first word
leads to the second word. Please use the following numbers to represent
different feelings about how often the first word leads to the second:

1.

1 2 3 4 5
Rarely Sometimes Usually

Trying hard _. Learning the training material

2. Putting forth effort -> Understanding the training material

Affective Outcomes

1.

2.

I am satisfied with this training.

I would be interested in participating in web-based training in the future.
I believe that this training was worthwhile.

I enjoyed this training.

I would be interested in participating in web-based training in the future.
I think other writing classes should use this web-based training program.
I would recommend this training program to students in other writing
classes.

I think web-based training should be offered more frequently.

150

Other Variables

1.

8.
9.
10.What is your major?

11.What is your grade point average (e.g., 2.85)

Approximately how many hours per week do you spend using a computer for

any reason?

 

Approximately how many hours per week do you use the internet (world wide

web) for any reason?

 

Have you ever participated in web-based training before?

 

Please rate your computer skills:

1 2 3 4 5
Poor Average Excellent

How comfortable are you using a web browser (like "Netscape" or "Internet

Explorer")?
1 2 3 4 5
Poor Average Excellent
What is your gender? __ Male _ Female
Where are you right now? Home __ MSU computer lab
_ other

What is your primary (native) language?

 

If you are bilingual, what is your secondary language?

 

 

 

151

Computer Anxiety

1.

2.

I am confident that I can learn computer skills. (R)

I am unsure of my ability to interpret computer output.

I will be able to keep up with the important technical advances of
computers. (R)

I can understand most technical matters. (R)

I feel apprehensive about using a computer.

I'm afraid I might damage a computer in some way.

Computer terminology makes sense to me. (R)

. I have avoided computers because they are unfamiliar to me.

I am confident that I can correct most mistakes that I make on a computer.

(R)

152

APPENDIX D
PRE-TESTSIPOST-TESTS

Dichotomous Choice Pre-Testhost-Test
Please indicate whether or not the italicized word is used correctly in the
sentences below. Use the mouse to click on the circle for the answer that you
choose.
1. If we throw a picnic, we can boost team moral.

correct

incorrect
2. I suggest that we turn down the contract; it's too risky to accept a project that
large with our current shortage of workers.

correct

incorrect
3. The five students' papers were in the office.

correct

incorrect
4. The project will loose money if it isn't managed well.

correct

incorrect

5. We can take a brake after the next speaker presents his proposal.

correct
incorrect

153

6. I wonder who's notebook this is.

correct

incorrect
7. I don’t know where your book is.

correct

incorrect
8. l assigned four more people to the development team because they're projects
are behind schedule.

correct

incorrect
9. I want to proofread my paper one more time before I attend my professor’s
office hours tomorrow.

correct

incorrect
10. In recent years, there has been an emphasis on development of new
products rather then on updating previous ones.

correct

incorrect
11. The dog left its footprints on the carpet.

correct

incorrect

12. Please inform all personal that the new casual Friday policy begins this week.

correct
incorrect

154

13. We hope that the wage raise will effectjob satisfaction.
correct
incorrect
14. All of the students are here accept Mary.
correct
incorrect
15. I think that project will be to expensive.
correct
incorrect
16. The bus's windows are broken.
correct
incorrect
17. The women's restroom is on the third floor.
correct
incorrect
18. After the company was acquired last year, my roll in the department changed
to one of a consultant.
correct
incorrect
19. That is Mr. Jones’s wallet.

correct
incorrect

155

20. The book's title is too long.

correct
incorrect

Indicate whether or not the sentence is grammatically conact or incomect by
clicking on the appropriate circles below.
21. Considering all that we have to do today.
correct
incorrect
22. Recruitment has improved; however, we still don't have enough applicants for
the position.
correct
incorrect
23. Access to floors three, four, and five are limited after 5 pm.
correct
incorrect
24. In order to fully understand the proposal that Laura submitted.
correct
incorrect
25. The list of students who want to participate are on the desk.
correct
incorrect
26. I liked having her on my team, she was a good employee.

correct
incorrect

156

27. Sitting at the corner, waiting for a ride.
correct
incorrect
28. Neither of us wants to have to fire him.
correct
incorrect
29. I wrote my paper last night it was really hard to finish.

correct
incorrect

30. Ann, a capable student that I had in class.

correct
incorrect

157

Pre-Test Paragraph

Read the following memo and answer the questions below.

I propose that we begin a new recruitment program to attract college students to
our company. The group of students' at the local. university have a reputation for
working hard, which is impressive. Considering all the pressures they have. I
know that the local college is Mr. Jones' alma mater, he always has good things to

say about the students.

We should get a list of student's who are interested in the positions our
department has. I suggest that we ask for there resumes ahead of time and then
invite some of them for visits to our company. I thinkthe job applicants' visits
should take place on Fridays so they can see our relaxed, casual atmosphere.
Bringing in students should also raise the moral of the employee's. Since they
want more personnel because they have been complaining about to much work. It
will also be a nice brake for the employees because we can have some of them

give the applicants' tours.

Its not to late to invite some students this fall, we can request resumes this week.
We may loose some applicants because other companies have started to recruit,
but not too many. If you like this idea, tell me, I will send a memo to our
supervisors secretary. We can than work threw the human resource management

office to set up the visits.

158

Please answer the following questions using numbers only (example: 3). If there

are none, enter a '0'.

How many sentence fragments are in the memo?
How many run-on sentences are in the memo?

How many incorrect examples of subject-verb agreement are in the
memo?

How many times were apostrophes used incorrectly in the memo?

How many times were necessary apostrophes omitted in the memo?

 

How many incorrect word choices (e.g., homonyms and related words)
were made in the memo?

159

APPENDIX E

INDIVIDUAL PAPER ASSIGNMENT

The individual paper serves two purposes: to develop your skills in finding and
analyzing information about HRM issues, and to improve your business writing
skills. In order to meet this second goal, you will submit a first draft of the paper,
which will be evaluated with respect to both content and writing style. You will
then be asked to revise the paper. Grades will be assigned to each version

based on writing style and HR content.

The individual writing assignment involves analysis of a current event that has
implications for human resource management. The HR world is constantly

changing. Globalization, increasing diversity of the workforce, and competitive
pressures all cause organizations to search for (and, to an extent, experiment
with) new HR techniques. At the same time, legislation in the US. and abroad

has direct and indirect influences on HR

For this assignment, you are to select an article from your readings of the popular
or business press that deals with an HR issue, and prepare a brief “position
paper” describing how (or whether) this issue should influence human resource
management practices in business. M38 Week, Fortune, The Wall Street
W, The New York Times, Workforce. and HRMagazine are examples of
publications that frequently carry articles about HR-related issues. We will give

examples of appropriate articles in class.

160

The format for this assignment is as follows: You are to take the role of a newly
hired graduate in a large firm with facilities in the United States and abroad. One
day, as you are riding the elevator with your boss, helshe says: “You know, I saw
something on the news the other day about (your topic). Aren’t you something of
an expert in this area? What do you know about this issue? How do you think it
will affect this company? Get back to me with an update. If it’s good, I’ll send it

out to our HR staff and mangers across the company”

You are to respond to this manager with a memo, no longer than 6 pages
double-spaced, that involves the following components: (1) a brief summary of
the issues as reported in the article; (2) a more in-depth discussion of the issue,
based on your own research into the topic (i.e., you will need to gather more
resources than just the one article); (3) your analysis of howthis issue might
affect a specific company (feel free to make assumptions about the company’s
business, workforce, locations etc. as needed); and (4) your opinion regarding

what steps, if any, the company should currently take to address this issue.

Please submit a copy of the original article along with your analysis. The first
draft of this written analysis will be due February 25 for all students. Remember,
it is worth 15% of your grade, so it is not in your best interest to take the first draft

lightly. The second draft will be due March 30.

161

REFERENCES

Arvey, R.D., & Cole. 1989. Evaluating change due to training. In I.L.

Goldstein (Ed.) Training and Development in Organizations. San Francisco:
Jossey-Bass.

Avner, A., Moore, 0., & Smith, S. (1980). Active external control: A basis
for superiority of CBI. Journal of Computer-Based Instruction, 6, 115-118.

Baldwin, T.T., & Ford, J.K. (1988). Transfer of training: A review and
directions for future research. Personnel Psychologys41, 63-105.

Baldwin, T.T., Magjuka, R.J., & Loher, B.T. (1991). The perils of
participation: Effects of choice of training on trainee motivation and learning.
Personnel Psmhologv. 44. 51-65.

 

Barrick, MR. & 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 personality dimensions and job performance.
Journal of Applied Psychology. 7s, 111-118.

Basic language skills. (1983). Englewood Cliffs, NJ: Regents/Prentice
Hall.

Button, S.B., Mathieu, J.E., & Zajac, 0M. (1996). Goal orientation in
organizational research: A conceptual and empirical foundation. Organizational
Behavior and Human Degion Processes 67, 26-48.

 

Canterucci, J. (1997). Intranet security. Training & Development,
February, 47.

Cohen, J. & Cohen, P. (1983). Applied multiple regression/correlation
analysis for the behaybralsciences (2" edition). Hillsdale, NJ: Lawrence
Erlbaum Associates.

 

Cohen, J. (1992). A power primer. Psychological Bulletin. 11(1), 155-159.

Cohen, S. (1997). Intranets uncovered. Training & Development.
February, 48-50.

Colquitt, J.A., LePine, J.A., & Noe, RA. (1999). Trainee attributesfiand

attitudes revisited: A meta-analytic structural equation modeling analvsis_of
research on training motivation. Manuscript submitted for publication.

162

Colquitt, J.A. & Simmering, M.J. (1998). Conscientiousness, goal
orientation, and motivation to Ieam during the learning process: A longitudinal
study. Journal of Applied Psvcholgy, 83(4).

Cook, T.D. & Campbell, D.T. (1979). Quasi-experimentation: Design and
analysis issues for fiefisettings. Boston, MA: Houghton Mifflin Company.

Cornell, R. & Martin, BL. (1997). The role of motivation in web-based
instruction. In B.H. Khan (Ed.), Web-Based Instruction (pp. 93-100). Englewood
Cliffs, NJ: Educational Technology Publications.

Costa,P.T. Jr., & McCrae, RR. (1992). The NEO-PI professional manfi
Odessa, FL: Psychological Assessment Resources.

Craiger, JP. 8. Weiss, R.J. Traveling in cyberspace: Web-based
instruction. TIP.

Curtin, C. (1997). Getting off to a good start on intranets. Training &
Development, February, 42-46.

Dillon, W.R., & Goldstein, M. (1984). Multivariate analysis: Methods and
applications. New York: John Wiley 8. Sons.

Dweck, CS. (1975). The role of expectation and attributions in the
alleviation of learned helplessness. Journal of Personalitv and Social

Psycholggy, 31, 674-685.

Dweck, CS. (1986). Motivational processes affecting learning. American
Psychologist, 41 , 1040-1048.

Elliott, E.S., & Dweck, CS (1988). Goals: An approach to motivation and
achievement. Jourpal of Personalig and SorflPsvcholmy, 54, 5-12.

Filipczak, B. (1996). Training on intranets: The hope and the hype.
Training, September, 24-32.

Fisher, S.L., & Ford, J.K. (1998). Differential effects of learner effort and
goal orientation on two Ieaming outcomes. Personnel Psychology, 5, 397-419.

Fleishman, EA. 8. Mumford, MD. (1989). Individual attributes and training

performance. In I.L. Goldstein and associates (Eds) Training and development in
organizations (pp. 183-255). San Francisco, CA: Jossey-Bass Publishers.

163

Ford, J.K., 8. Kraiger, K. (1995). The application of cognitive constructs
and principles to the instmctional systems model of training: Implications for
needs assessment, design, and transfer. lntemational Review of Industrial and

Organizational Psycholpgy, 10, 1 - 48.

Gagne, RM. (1985). The corfiitions ofLearningand theorv of instruction.
New York: Holt, Rinehart, and Winston. .

Gagne, R.M., & Medsker, KL. (1996). The conditions of Ieaming: Training
applications. Forth Worth, TX: Harcourt Brace.

Gellatly, LR. (1996). Conscientiousness and task performance: Test of a
cognitive process model. Journal of Applied Psvcholpgwm (5), 474-482.

Golstein, l.L. & Gilliam, P. (1990). Training system issues in the year
2000. American Psycholpgist, 45(2), 134-143.

Hannan, M.T. & Carroll, GR. (1992). Dynamics of organizational
populations. Oxford: Oxford University Press.

Howard, A. (1995). A framework for work change. In A. Howard (Ed.), T_he_
Changing Nature of Work, (pp. 344). San Francisco, CA: Jossey-Bass, Inc.

Howell, A. E., Mullins, M. E, Fisher, S. L., Schmitt, N., & Kozlowski, S...WJ
(1998, April). Using the internet for competitive advantage. Paper presented at
the 13th Annual Meeting of the Society for Industrial and Organizational
Psychology, Dallas, TX.

Kaufman, J. (1985a). Mastering grammar: Book1.Cambridge Review
Series. Prentice Hall.

Kaufman, J. (1985b). Mastering gLammar: Book 2. Cambridge Review
Series. Prentice Hall.

Keller, J.M. (1983). Motivational design of instruction. In C.M. Reigeluth
(Ed.), Instructional-Design Theories and Models: An Overview of their Current
Status (pp. 383-434). Hilldsdale, NJ: Lawrence Erlbaum Associates.

Khan, EH (1997). Web-based instruction (WBI):What is it and why is it?
In B.H. Khan (Ed.) Web-Based Instruction (pp. 5-18). Englewood Cliffs, NJ:
Educational Technology Publications.

Kinzie, MB. (1990). Requirements and benefits of effective interactive

instruction: Learner control, self regulation, and continuing motivation.
Educational TechnologLResearch and Design, 381-21.

164

Kinzie, M.B., Sullivan, H.J., & Berdel, R.L. (1988). Learner control and
achievement in science computer-assisted instruction. Jourrgl of Educational

Psycholpgy, 80, 299-303.

Kraiger, K., Ford, J.K., & Salas, E. (1993). Application of cognitive, skill-
based, and affective theories of Ieaming outcomes to new methods of training
evaluation. Journal of Applied Psychology, 78(2). 31 1-328.

Lawler, E.E., Ill, & Suttle, J.L. (1973). Expectancy theory and job behavior.
Organizational Behavior and Human Performance. 9, 482-503.

Lee, S.S. & Lee, Y.H.K. (1991). Effects of learner-control versus program-
control strategies on computer-aided Ieaming of chemistry problems: For
acquisition or review? Journal of Educational Psvcholggy, 83, 491-498.

London, M. & Bassman, E. (1989). Retraining midcareer workers for the
future workplace. In I.L. Goldstein and associates (Eds) Training and
development in organizations (pp. 333-375). San Francisco, CA: Jossey-Bass
Publishers.

Martocchio, J.J. (1992). Microcomputer usage as an opportunity: The
influence of context in employee training. Personnel Psychology, 45, 529-552.

Martocchio, J.J. (1994). Effects of conceptions of ability on anxiety, self-
efficacy, and learning in training. gouml of Applied Psychology. 79(6), 819-825.

Martocchio, J.J. & Judge, TA. (1996). The role of conscientiousness in
employee training: A test of alternative models. Paper presented at the Eleventh
Annual Conference of the Society for Industrial and Organizational Psychology,
San Diego, CA.

Martocchio, J.J. 8. Webster, J. (1992). Effects of feedback and cognitive
playfulness of performance in microcomputer software training. Personnel

Psycholggy, 45, 553-578.

Mathieu, J.E., & Martineau, J.W. (1997). Individual and situational
influences on training motivation. Improving Training Effectiveness in Work
Organizations. Mahwah, NJ: Lawrence Erlbaum Associates.

Mathieu, J.E., Tannenbaum, S.l, & Salas, E. Influences of individual and
situational characteristics on measures of training effectiveness. Academy of

Management Journal, 35, 828-847.

Milhelm, W.D. & Martin, BL. (1991). Theoretical bases for the use of
Ieamer control: Three different perspectives. Journal of Cmputer-Bassg
Instruction 18, 99-105.

 

165

Mischel, W. (1977). The interaction of person and situation. In D.
Magnusson & N.S. Endler (Eds.), Personalipr at the crossroads: Current issues in
interactional psycholpgy (pp. 333-352). Hillsdale, NJ: Erlbaum.

 

Murphy, MA. 8 Davidson, G.V. (1991). Computer-based adaptive
instruction: Effects of learner control on concept Ieaming. Journal of Computer-
Based Instruction 18, 51-56. .

 

Murrell, A.J. & Sprinkle, J. (1993). The impact of negative attitudes toward
computers on employees' satisfaction and commitment within a small company.
Computers in Human Behavior. 9, 57-61.

Noe, RA. (1986). Trainees’ attributes and attitudes: Neglected influences
on training effectiveness. Academv of Managment Review. 11, 736-749.

Noe, R.A. & Ford, J.K. (1992). Emerging issues and new directions for
training research. Basearch in Personnel and Human Resources Management,
E. 345-384.

Noe, R.A., & Schmitt, N. (1986). The influence of trainee attitudes on
training effectiveness: Test of a model. Pepsonnel Psychology. 39. 497-523.

Pear, J.J. & Novak, M. (1996). Computer-aided personalized system of
instruction: A program evaluation. Computers in Teaching. 23(2), 119-123.

Phillips, J.M. 8. Gully, SM. (1997). Role of goal orientation, ability, need
for achievement, and locus of control in the self-efficacy and goal-setting
process. Journal of Applied PsvchoLoqv. 82(5), 792-802.

Pollack, C. & Masters, R. (1997). Using internet technologies to enhance
training. Perfon'nance Improvement, February, 28-31.

Pollack, J.C. & Sullivan, H.J. (1990). Practice model and learner control in
computer-based instruction. Contemporagy Educational Psycholmv. 15, 251-
260.

Pridemore, D.R. & Klein, JD. (1991). Control of feedback in computer-
assisted instruction. Educational Technology Research and Development. 39.
27-32.

Quinones, MA. (1995). Pretraining context effects: Training assignment

as feedback. Jogmal of Applied Psychology, 80, 226-238.

Ray, N.M. & Minch, RP. (1991). Computer anxiety and alientation:
Toward a definitive and parsimonious measure. Human Factors. 32, 477-491.

166

Reiser, R.A., & Gagne, RM. (1983). Selecting media for instruction.
Englewood Cliffs, NJ: Educational Technology Publications.

Ross, S.M. & Morrison, G.R. In search of a happy medium in instructional
technology research: Issues concerning external validity media replications, and
learner control. Educational Technolmv Research and Development. 37. 19-33.

Roth, PL (1994). Missing data: A conceptual review for applied
psychologists. Personnel Psvcholong47, 537-560.

Salthouse, TA. (1991). Mediation of adult age differences in cognition by
reductions in working memory and speed of processing. Psycholpgical Science,
2(3), 179-183.

Scott, W.R. (1991). Institutions and organizations. Thousand Oaks, CA:
Sage Publications, Inc.

Shaw, DS (1992). Computer-aided instruction for adult professionals: A
research report. Journal of Computer-Based Instruction. 19(2), 54-57.

 

Snow, RE. (1978). Theory and method for research on aptitude
processes. Intelligence, 2, 225-278.

Snow, RE, 8. Lohman, OF. (1984). Toward a theory of cognitive aptitude
for Ieaming from instruction. Journal of Educational PsvcholongQ, 347-376.

Steinberg, ER. (1977). Review of student control in computer-assisted
instruction. Journal of Computer-Based Instruction, 3, 84-90.

Stevens, L. (1996). The intranet: Your newest training tool? Personnel
Journal, July, 2-32.

Terban, M. (1993). Checking your grammar. New York: Scholastic, Inc.

Tumage, J.J. (1990). The challenge of new workplace technology for
psychology. American Psychologist. 45(2), 171-178.

Vroom, V.H. (1964). Work and motivation. New York: VVIley.

Warr, P. & Bunce, D. (1995). Trainee characteristics and the outcomes of
open Ieaming. Personnel Psychology, 48, 347-375.

Webster, J. & Martocchio, J.J. (1995). The differential effects of software
training previews on training outcomes. Jourfll of Management, 21(4), 757-787.

167

Wexley, K.N., & Latham, GP. (1981). Developing and training human
resources in organizations. Glenview, IL: Scott Foresman.

Vifilder, G., Mackie, D., & Cooper, J. (1985). Gender and computers: Two
surveys of computer-related attitudes. Sex Roles. 13(3/4), 215-228.

Wine, J. (1971). Test anxiety and direction of attention. Psychological
Bulletin 76(2), 92-104.

 

Worthington, E.L., Jr., Welsh, J.A., Archer, C.R., Mindes, E.J., & Forsyth,
DR. (1996). Computer-assisted instruction as a supplement to lectures In an
introductory psychology class. Computers in Teaching. 23(3), 175-181.

Wulf, K. (1996). Training via the internet. Mining & Development. May,
50-55.

Yang, C.S. & Moore, D.M. (1995-6). Designing hypermedia systems for
instruction. Journal of Educational Technology Systems, 24(1), 3-30.

168

IIIIIIIIIIIIIIIIIIIIIIIIIIIII

willilllliljjljjllljjl