Final 2008

THE EFFECTS OF SUPERVISORY SUPPORT, AGE AND GENDER ON SELF

EFFICACY AND METACOGNITIVE ACTIVITY IN A LEARNER

CONTROLLED TRAINING ENVIRONMENT

A Dissertation

Presented to the Faculty of the College of Business Administration

Touro University International

In Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy in Business Administration

By

James V. Polizzi

October 2008

ii

Copyright by James V. Polizzi

2008

All rights reserved

ii

BIOGRAPHICAL SKETCH

James Polizzi earned a Bachelors of Business Administration (Marketing) from The City College of New York in 1966. He received a Masters of Business Administration (Management) from Wagner College in 1996. He received a Doctor of Philosophy, Business Administration from Touro University International in 2008. He is currently an instructor in the Management Department at Berkeley College, New York City and Online Campuses and President of The Aegis Group – a strategic consultancy.

iii

DEDICATION

I dedicate this dissertation to my wife, Josephine. Her continuous support, understanding

and encouragement gave me the will to finally complete this endeavor.

iv

ACKNOWLEDGEMENTS

I would like to thank SimuLearn, Inc for their permission to use the Virtual Leader

leadership training software in the conduct of this study. Particular thanks to Mr. Pierre

Thiault for his advice and continuous support for this project.

v

Table of Contents

Page

List of Figures ................................................................................................................... vii

List of Tables ................................................................................................................. .viiii

Abstract………………………………………………………………………………...…ix CHAPTER I: INTRODUCTION........................................................................................ 1

Problem Background .............................................................................................. 1

Identification of the Issues ...................................................................................... 7

CHAPTER II: LITERATURE REVIEW ........................................................................... 9

Learning Theory...................................................................................................... 9

Metacognition and Training.................................................................................. 13

Learner Controlled Training and Metacognitive Interventions ............................ 18

Motivation and Self Efficacy................................................................................ 23

Age, Gender and Computer Self Efficacy ............................................................ 26

Supervisory Support.............................................................................................. 28

CHAPTER III: METHODOLOGY .................................................................................. 35

Research Design.................................................................................................... 36

Operationalization of Variables ............................................................................ 37

Sample................................................................................................................... 39

Procedure .............................................................................................................. 39

Research Questions............................................................................................... 40

Data Analysis Plan................................................................................................ 42

CHAPTER IV: ANALYSIS OF DATA........................................................................... 47

Preliminary Analyses ............................................................................................ 48

vi

Research Questions............................................................................................... 54

Research Question 1 ................................................................................. 55





CHAPTER V: DISCUSSION AND IMPLICATIONS.................................................... 65

Summary of Findings............................................................................................ 65

Implications........................................................................................................... 68

Metacognitive Activity for Children Versus Adults................................. 68

Gender, Age, and Computer Self-Efficacy............................................... 69

Supervisory Support and Self-Efficacy .................................................... 70

Metacognitive Activity and Self-Efficacy ................................................ 71

Recommendations for Future Research ................................................................ 72

Recommendations for Practice ............................................................................. 75

Conclusions........................................................................................................... 77

References......................................................................................................................... 79

Appendix A: Demographic Survey................................................................................... 88

Appendix B: Learner Control Self Efficacy Scale............................................................ 89

Appendix C: Computer Self-Efficacy Scale ..................................................................... 91

Appendix D: Supervisory Support Scale .......................................................................... 95

Appendix E: Metacognitive Activity Scale ...................................................................... 96

Appendix F: Full Regression Results for Path Models..................................................... 98

vii

List of Figures

Page Figure 1. Path Model for Research Questions 1-4 ............................................................46 Figure 2. Path Model for Research Question 5 .................................................................46 Figure 3. Path Model for Research Questions 1 and 2 with Regression Coefficients.......55 Figure 4. Path Model for Research Question 3 with Regression Coefficients for Males..................................................................................................................................57 Figure 5. Path Model for Research Question 3 with Regression Coefficients for Females ..............................................................................................................................57 Figure 6. Path Model for Research Question 4 with Regression Coefficients for Younger Participants..........................................................................................................60 Figure 7. Path Model for Research Question 4 with Regression Coefficients for Older Participants.........................................................................................................................61 Figure 8. Path Model for Research Question 5 with Regression Coefficients..................63

viii

List of Tables

Page

Table 1. Descriptive Statistics for Sample Demographic Characteristic (N=120)...........49 Table 2. Descriptive Statistics for the Composite Measures (N=120) ..............................50 Table 3. Correlations Between Composite Measures (N=120) ........................................51 Table 4. Correlations Between Composite Measures as a Function of Gender (N=120) ............................................................................................................................52 Table 5. Correlations Between Composite Measures as a Function of Age Group (N=120) .............................................................................................................................53

ix

ABSTRACT

THE EFFECTS OF SUPERVISORY SUPPORT, AGE AND GENDER ON SELF

EFFICACY AND METACOGNITIVE ACTIVITY IN A LEARNER CONTROLLED

TRAINING ENVIRONMENT

James V. Polizzi, Ph.D.

Touro University International 2008

The increase in costs and frequency of training have driven U.S. businesses to a

greater use of learner controlled training, i.e. training delivered in the absence of a live

instructor. Success in learning complex material has been positively related to

metacognitive activity, yet learner controlled training may present unique challenges to

the formation of learning strategies. This study investigated the relationships between

employee self efficacy, computer self efficacy, supervisory support, gender and age and

their effect on metacognitive activity. The research was conducted during

organizationally sponsored, learner controlled training among adults. The study results

suggest a positive role for supervisory support on self-efficacy and metacognitive

activity. Metacognitive activity increased with higher levels of learner control self

efficacy which, in turn, was associated with higher levels of computer self efficacy.

1

CHAPTER I: INTRODUCTION

Problem Background

As organizational efforts to improve productivity increase, employee training has

become an even more critical element of firm activities. Importantly, in addition to

productivity, the very nature of the business organization is shifting. As projected by the

RAND Corporation (2004), the required skills for a productive workforce in the 21st

century will include: problem solving skills, communication and collaborative ability.

The emergence of a knowledge-based workforce demands that education and training

become a continuous process throughout the life course, involving training and retraining

that continue well past initial entry into the labor market. Technology-mediated learning

is a promising tool for life-long learning, both on the job and through traditional public

and private education and training institutions. (RAND, 2004)

The American Society for Training and Development (ASTD; 2008) estimates

2006 learning and development spending for U.S. firms at $129.6 billion. Expenditures

per employee have risen to $1,040 in 2006, approximately 2% above 2004. Together,

managerial and executive development training totaled more learning content in 2004

than technology, business processes and industry-specific content (ASTD, 2004). A key

indicator of the trends in business organizations is the increasing use of terms such as

“workforce development” and “organizational effectiveness” as part of the titles of in-

house trainers and the establishment of a “Chief Learning Officer” (Rodriquez, 2005).

Human Resources Focus (“Despite Economy,” 2004) noted some significant

trends in training budgets and the nature of training methodologies: U.S. companies spent

more money on training, provided more hours of training and increased use of technology

2

for training in 2002 versus 2001. Training delivered via learning technologies increased

to 15.4% in 2002, from 10.5% in 2001; while training delivered via a traditional

classroom technique declined to 72%, versus 77% in 2001. More recently, 2004 saw 50%

of technology based delivery in an online format, with 75% of online learning classified

as “self-paced” (ASTD, 2004).

According to ASTD (2004), organizations with high levels of investment in

training aligned learning with business needs and achieved efficiency and effectiveness in

the learning function. Collins and Clark (2003) found that human resource practices (i.e.

training) were found to be positively correlated with creating organizational competitive

advantages. The increase in use of technology to deliver training, coupled with the

concurrent decline in traditional instructor-led training has been facilitated by the

widespread use of desktop computers and near universal access to the World Wide Web

in U.S. firms. Additionally, the rising costs of training have stimulated a move to more

efficient methods of delivering training in organizations.

Training without a live instructor encompasses many methods of instruction,

either as single or mixed method approaches, including Web based training, Intranet

training programs, and CD-ROM. Collectively called learner controlled training

(Schmidt, 2003), the benefits of self-pacing, flexible access and lower costs are driving

more firms to increase use of this design in training programs. The increased availability

of interactive training designs gives individuals increased control over the pace, sequence

and time spent on training (Tannenbaum, 1992).

Research on learner controlled training has shown generally positive, but mixed

results. Learners who are allowed to choose the sequence of learning, content, and time in

3

study have reported positive attitudes towards training and improved outcomes (e.g.,

Brown, 2001; Kinzie & Sullivan, 1989; Milheim & Martin, 1991; Morrison, Ross, &

Baldwin, 1992). Despite these observed advantages, increased learner control has not

shown consistent improvements in post training performance. Brown (2001) found that

learner control is associated with a number of negative processes, including lower time

on task and inadequate learning strategies. Eom and Reiser (2000) found a poorer posttest

result for learner control subjects than for program control subjects. Pollock and Sullivan

(1990) found that students with no control over practice had higher posttest scores than

students given control over the amount of practice. Gist, Schwoerer and Rosen (1989)

found that a modeling approach resulted in higher performance levels than computer-

assisted instruction among managerial trainees. Hannafin and Sullivan (1996) found that

learner ability affected the amount of control students applied in a learner controlled

program.

In a review of interactive learning environments (Aleven, Stahl, Schworm,

Fischer & Wallace, 2003), the authors identify the need for further investigation of the

effects of the context in which learning occurs. Specifically, they identify the physical,

social and institutional environment as factors potentially affecting the learning process.

Thus, an important area for research is to identify the variables that might influence

learners in learner controlled training. One result of these findings has been an increased

focus on the learner as an active participant in the learning process. A significant

conclusion is that not all learners are capable of successfully directing their learning; they

fail to take advantage of the control they are given.

4

A review of the literature suggests that learner controlled training can be

improved by incorporating learning theory to increase the understanding of how people

learn. Many employees, especially older workers, approach these nontraditional training

environments with some trepidation and are easily frustrated with the inability to directly

question a live instructor. Thus, training design can be improved by understanding the

attitudes and capabilities of workers to utilize the technology. Most empirical studies of

learner control have been conducted among students in academic environments.

Developing a greater understanding of adult learners in an organizational setting can

improve the effectiveness of training.

A significant criticism of learner controlled training has been the lack of teaching

of higher order cognitive skills. Live instruction has traditionally been viewed as the best

method for communicating these skills. A growing number of training designs utilize in-

training interventions to guide the learner in the cognitive processes need to master the

training material. These can be broadly classified as metacognitive interventions;

explained as guiding the learner to think about their thinking process. Metacognitive

interventions have been associated with positive outcomes in learner controlled training

environments and holds promise for improving organizational training designs (Schmidt,

2003). Antecedents of metacognitive activity have been identified as learner mastery and

performance orientations, where performance orientations include the learner’s

motivation and perception of self efficacy (Schmidt).

Training outcomes can also be influenced by the potential effects of an

employee’s perception of their ability to learn without a live instructor. This general

construct of perception of ability is identified as self efficacy (Bandura, 1986). Self

5

efficacy refers to an individual’s belief in his/her ability to accomplish a given task

(Bandura). Self efficacy relates to effort expended on a task, and persistence in achieving

a positive task result (Gist & Mitchell, 1992). In a training environment, self efficacy

beliefs are likely to contribute positively to successful outcomes. Research has

consistently demonstrated a positive relationship between self efficacy, positive

motivation and learning (e.g. Gist, Stevens & Bavetta, 1991; Martocchio, 1994; Mathieu,

Tannenbaum & Salas, 1992). However, the literature reveals very few studies of the

relationships between self efficacy and metacognitive interventions in business

organizations (e.g., Schmidt, 2003). The first question raised here, then, is whether an

adult learner’s self efficacy perception affects metacognitive activity in a learner

controlled training environment in the context of a business organization.

While Bandura (1986) posited an individual’s behavior as a result of external

approval or disapproval, the landmark Ohio State Leadership research (Halpin & Winer,

1957) studied the link between supervisory behavior and subordinate attitudes and

performance. The external influences the employee observes – namely, the attitude of

their supervisor for this alternative training design can affect perceptions of self efficacy

(e.g., Bandura, 1986; Illeris, 2003; Martocchio, 1992).

To summarize, two variables may affect the level of metacognitive activity in a

learner controlled training environment: the learner’s perception of self-efficacy in the

absence of a live instructor, and the degree to which the individual’s supervisor exhibits

support/trust for the learner in learner controlled training. The nature of the delivery of

learner controlled training in the contemporary organization (e.g., via desktop

computers), raises the possibility that two further variables may affect learning: age and

6

gender. Given recent trends, we can presume continued and increasing use of computers

to deliver training, thus examination of these variables may prove useful.

It is a general perception among those in business and in the general population

that older workers are less comfortable and less proficient than younger workers in the

skillful use of desktop computers. Specifically, older users were found to have low

confidence in their ability to use computer technology (Comber, Hargreaves & Dorn,

1997). We can speculate that older workers will exhibit lower self efficacy in training

delivered by computer than training delivered with traditional instructor-led training.

While gender has generally been shown to have mixed effects on computer

competency (e.g., Ford, Miller & Moss, 2001; Henry & Stone, 1999); the differences in

masculine sex role traits (e.g., independence, assertiveness and competitiveness) and

feminine sex role traits (e.g., dependence and interpersonal relationships) may have an

effect in a learner controlled training environment where the absence of a live instructor

prevents interaction. Gilley (2002) reports that females do not perceive themselves as

manipulators of computer technology; but merely as end users of pre-designed programs.

The American Association of University Women (AAUW; 2000) studies found that

females have been encouraged to participate in computer technology through the use of

productivity software such as graphics programs, databases, page layouts, and so forth,

whereas males are more adventurous in their learning with respect to computer

technology. Gender differences in computer self efficacy is revealed in Brosnan’s (1998)

study which found that 64% of females agreed that computing was a “male activity” and

that “men were better at computing than women” (1998, p. 63). Gender, therefore, may

7

have an effect on a learner’s perception of self efficacy in training delivered in the

absence of a live instructor.

Identification of the Issues

The increasing complexity of job requirements has fostered a continuing increase

in training activities and related costs for U.S. firms. Training designs have evolved to

meet these challenges by increasing use of training without a live instructor and are

increasingly utilizing in-learning interventions to improve acquisition of thinking skills

(metacognitive activity). There is empirical evidence that increased metacognitive

activity improves the effectiveness of learner controlled training, yet not all learners show

consistently positive results. What, then, are the key variables affecting metacognitive

activity? Self efficacy perceptions and supervisory support have been shown to affect

learning outcomes; with self efficacy viewed as both an independent variable affecting

learning outcomes and a dependent variable affected by supervisory support. With the

trend to even greater use of learner controlled training, age and gender are two

moderating variables important to evaluate when researching self efficacy in this

environment.

The research questions guiding this study are:

1. Is supervisory support related to learner control self efficacy and computer

self-efficacy in a learner controlled training environment?

2. Are computer self-efficacy or learner control self-efficacy related to

metacognitive activity in a learner controlled training environment?

8

3. Do the relationships between supervisory support, learner control self-

efficacy, and metacognitive activity vary as a function of the gender of the

learner?

4. Do the relationships between supervisory support, learner control self-

efficacy, and metacognitive activity vary as a function of the age of the

learner?

5. Does computer self-efficacy have an effect on learner control self-efficacy

which subsequently has an effect on metacognitive activity?

In order to address these questions, a quantitative study to test the effects of supervisory

support, age and gender on an adult learner’s perception of self-efficacy and

metacognitive activity when metacognitive interventions are utilized in a learner

controlled training environment was performed. The results of this study provide insights

potentially valuable in improving the effectiveness of learner controlled training.

9

CHAPTER II: LITERATURE REVIEW

This section will review and apply the significant theories and empirical research

encompassing learning theory, metacognitive processes, an individual’s perception of

self-efficacy, the effects of supervisory influences, and experience with metacognitive

interventions as a training strategy. This review will also establish the relationship of

metacognitive activity to potential successful training outcomes and metacognitive

interventions as a factor which improves individual metacognitive activities. Individual

self-efficacy will be shown to be influenced by supervisory support for the specific

training activity. The present study examined the interaction of the internal effects of

metacognition and self-efficacy, as well as the external influence of supervisory support

on self efficacy in a learner controlled training environment.

Learning Theory

Skinner (1968, 1969) proposed that learning is a result of patterns of behavior

developed as a response to a stimulus. Skinner built on Watson’s work from the

beginning of the twentieth century using an empirical approach with animals, termed

stimulus-response behaviorism (DeMar, 2004). Ultimately termed classical behaviorism,

the theory viewed learning as changing the behaviors of individuals, sometimes through

trial and error experiences until a positive reinforcement was obtained (Semple, 2000).

Skinner’s experiments led him to modify Watson’s original view of behavior by adding

the concept of intermediary purposefulness to the stimulus – response formula (DeMar).

This concept is now described as operant conditioning, i.e. people behave in a particular

way because of the past consequences of that behavior, and thus one acts in expectation

of a certain outcome (DeMar). Skinner’s research with rats showed that punishment

10

halted a previously rewarded behavior almost immediately; but previously rewarded

behavior continued for some time when only the reward was withheld (Naik, 2004).

Behaviorists embrace four main steps regarding learning: first, each step should

be brief and follow from previously learned behavior; second, behavior is shaped by the

pattern of reinforcements, so learning should be regularly rewarded; third, provide

immediate feedback; fourth, the learner should be given direction to the most successful

path (Semple, 2000). Behaviorist theories of learning led to the introduction of

“programmed learning” (also programmed instruction) by machines in the 1950’s and

1960’s (Semple). In a learning environment, behaviorism relies on an instructor centered

approach where the learner is largely passive and controlled by the instructor’s processes

(Constructivist Learning Theory, n.d.).

Constructivist Learning Theory (n.d.) views learning differently from the

behaviorist stimulus-response phenomena. Constructivism posits the concepts of self-

regulation and acquisition of conceptual cognitive structure through reflection and

abstract thought (Constructivist Learning Theory, n.d.). Two major themes of

constructivism relate to how people learn: order and self (Mahoney, n.d.). Mahoney

explains that order reflects a person’s activities devoted to establishing a pattern to prior

experiences using emotional “meaning-making” processes (p.3). Constructivists further

posit that the organization of activity is fundamentally self-referent and self-repeating;

people continually experience and monitor their sense of personal identity (Mahoney).

Flavell (1977) posits that a person’s knowledge affects and is affected by how one

perceives things; and how one classifies and conceptualizes influences the way a person

reasons about those things. Cognition can be described as a system of “interacting

11

processes which generate, code, transform and otherwise manipulate information”

(Flavell, p. 14). Viewed more narrowly, cognition addresses physical and mathematical

objects while social cognition concerns human affairs and social interactions (Flavell).

Social cognition explains that courses of action are chosen as a result of a person’s

perceived capabilities and sustained partly on the basis of expected outcomes (Bandura,

1986).

In expanding the constructivist learning theory, Bandura (1986) explains that, in

the social cognitive view, humans are not driven solely by inner forces or by external

stimuli. Rather, the interaction of behavior, cognitive and personal factors, and

environmental events describe a model of reciprocity of these elements that seeks to

explain human functioning (Bandura). Each of these factors can be of different strengths

and can occur at different times. The influence of any factor can take time to develop and

to trigger a reciprocal influence.

Bandura (1986) describes the nature of social cognition, and its differences from

Behaviorism, in terms of “capabilities” (p. 18-21). Symbolizing capability refers to the

human capacity to transform experiences into internal models that serve as guides for

future action (Bandura). This suggests that experience mediates the classical stimulus-

response view. Forethought capability is explained (Bandura) as the use of a visualized

future which is affected by goals and potential courses of action; suggesting that

stimulus-response is also mediated by anticipated future outcomes – not necessarily an

immediate outcome. Bandura also posits an external influence on learning: vicarious

capability, i.e. the ability to learn through observation of actions of others and

consequences of those actions. Self-regulatory capabilities are, perhaps, central to social

12

cognitive theory (Bandura). Behavior is motivated and regulated by a person’s internal

goals and standards as well as their assessment of their performance towards those goals

(Bandura). Thus, self-produced influences mediate the stimulus-response model. Bandura

describes the distinctively human characteristic of self-reflective capability:

This (self-reflective capability) enables people to analyze their experiences and to think about their own thought processes. By reflecting on their varied experiences and on what they know, they can derive generic knowledge about themselves and the world around them. People not only gain understanding though reflection, they evaluate and alter their own thinking. In verifying thought through self-reflective means, they monitor their ideas, act on them or predict occurrences from them, judge the adequacy of their thoughts from the results, and change them accordingly. (p.21)

With regard to the nature of cognitive and personal factors, Wood and Bandura (1989a)

discuss the role of cognitive, vicarious, self-regulatory and self-reflective processes as

central to people’s behavior in organizations. Wood and Bandura explain that people

develop competencies through behavior modeling, cultivation of beliefs in their

capabilities, and enhancement of motivation through goals.

Gagne´ and Briggs (1974) describe the act of learning as composed of three

internal states: information, intellectual skills and strategies. Information can be stored in

memory for retrieval as required or accessed directly as in printed directions. Intellectual

skills are described as the ability to learn new things based upon cues that must be

previously learned and recalled. A learning situation often requires the use of strategies

for learning and remembering. These strategies are very general and apply to a wide

range of learning situations. Referred to as “self-management” (Gagne´ & Briggs, 1974

p. 9), the concept embodies a learner’s individual process for solving problems and

recalling previously learned methods of cognitive paths.

13

Variations in adult learning – both inter personal and intra-personal – have been

attributed to differences in prior knowledge, cognitive processes, and learning and

memory strategies (Weinert & Kluwe, 1987). The identification and explanation of the

role of learning strategies in organizational training are examined in detail in the

following section - Metacognition and training.

Metacognition and Training

In general, Metacognitive theory focuses on first, the awareness and management

of one’s thinking; second, differences in self-efficacy perceptions; third, knowledge and

knowledge and development of thinking strategies from one’s experiences and fourth,

strategic thinking (Paris & Winograd, 1990). Cognitive strategy is an internal skill in

which the learner consciously or unconsciously selects a mode of thinking about and

solving a problem. The object of the skill is to manage thinking behavior (Gagne´ &

Briggs, 1974). The quality of one’s cognitive strategies affects the degree of creativity,

fluency and criticality of the learning process (Bruner, Goodnow & Austin, 1956, Gagne´

& Briggs).

Flavell is most often cited as the developer of original propositions about what are

called metacognitive processes. Flavell (1976) attempted to explain why children could

not solve problems although they were given correct solution procedures. He believed

that this was “the central problem in learning and development, namely, how and under

what conditions the individual assembles, coordinates or integrates his already existing

knowledge and skills into new functional organizations” (p. 231). In examining the

inability of children to solve problems consistently, Flavell posed two questions: “what

problem-adaptive things might they be failing to do, or what problem mal-adaptive things

14

might they be doing instead?” (p. 232). From these questions, he developed the construct

of metacognition. Flavell described the construct as follows:

Metacognition refers to one’s knowledge concerning one’s own cognitive processes and products or anything related to them, e.g., the learning-relevant properties of information or data . . . Metacognition refers, among other things, to the active monitoring and consequent regulation and orchestration of these processes in relation to the cognitive objects or data on which they bear, usually in the service of some concrete goal or objective. (p.232) Flavell (1979) explained metacognitive experiences as “any conscious cognitive

or affective experiences that accompany and pertain to any intellectual enterprise” (p.

906). These experiences are conscious and are generally accompanied by emotions such

as anxiety, feeling of knowing, or judgments of learning. Flavell (1987) explained

metacognitive experiences with the following:

If one suddenly has the anxious feeling that one is not understanding something and wants and needs to understand it, that feeling would be a metacognitive experience. One is having a metacognitive experience whenever one has the feeling that something is hard to perceive, comprehend, remember or solve; if there is a feeling that one is far from the cognitive goal; if the feeling exists that one is, in fact, just about to reach the cognitive goal; or if one has the sense that the material is getting easier or more difficult that it was a moment ago. (p. 24)

Metacognitive experiences aid in the assessment of metacognitive goals, modification of

metacognitive knowledge and in the utilization of strategies (Flavell, 1979).

Flavell (1979) developed a model of metacognition and cognitive monitoring that

contained four classes of cognitive phenomena: metacognitive knowledge, metacognitive

experiences, tasks and actions (strategies). Flavell described metacognitive knowledge as

“that segment of your stored world knowledge that has to do with people as cognitive

creatures and with their diverse tasks, goals, actions, and experiences” (p. 906).

Metacognitive knowledge consisted of three factors: (a) person, (b) task and (c) strategy.

15

The person factor of metacognitive knowledge concerns knowledge and beliefs

about one’s self and others as cognitive processors. Flavell (1987) identified three

subcategories of the person factor: intraindividual, interindividual and universal.

Intraindividual knowledge relates to the variation in interests, propensities and aptitudes.

Interindividual knowledge concerns comparisons between persons. Universal knowledge

is concerned with “intuitions about the way the human mind works – knowledge of such

universal mental phenomena” (Flavell, p. 22).

The task factor of metacognitive knowledge relates to the availability of

information and the use of that information in the context of task demands or goals.

According to Flavell (1987), the task factor concerns how information “affects and

constrains how one should deal with it” (p. 22). Flavell explains that if information is

very difficult, one proceeds slowly and carefully to insure deep and comprehensive

understanding. The strategy variable concerns “what strategies (means, processes, and

actions) are likely to be effective in achieving what subgoals and goals in what sorts of

cognitive undertakings” (Flavell, 1979, p. 907). In 1982, Kluwe expanded the concept by

identifying two common attributes of metacognitive activities: the subject has some

knowledge of his own thinking and the subject may monitor and regulate the course of

his own thinking.

Metacognition has been defined in various ways by different researchers;

however, the various approaches contain the following concepts: knowledge of one’s

knowledge, thought processes, and cognitive and affective states; the ability to

consciously and deliberately monitor and regulate one’s knowledge, processes, and

cognitive and affective states (Hacker, 2003, p. 6). Metacognition can also be explained

16

as the ability to control one’s cognitive processes, viewed as self-regulation (Livingston,

1997) or self-management (Gagne´ & Briggs, 1974).

Metacognitive skill has been found to distinguish successful learners from

unsuccessful learners (Tannenbaum & Yuki, 1992). Metacognitive interventions (in-

learning training strategies) have been found to increase the amount and accuracy of

learner’s knowledge and to improve strategies for allocating time and effort (Schmidt &

Ford, 2003). In one study among students (Relan, 1995) subjects receiving learning

strategy training in a computer based instruction environment performed better in posttest

results than those who received no strategy training.

The increased availability of interactive training designs gives individuals

increased control over the pace, sequence and time spent on training (Tannenbaum &

Yuki, 1992). Brown (2001) concludes that learner choices regarding study time and

practice positively affected knowledge acquisition in a computer based training program.

Learners with increased control can consciously tailor training, leading them to learn the

task more effectively (Ford, Smith, Weissbein, Gully & Salas, 1998). However, Brown

also found that learner control is associated with a number of negative processes,

including lower time on task and inadequate learning strategies.

Gagne´ and Briggs (1974) proposed that cognitive strategies can be learned by

organizing external interventions that foster the development of internal processes. They

posit a design whereby “favorable conditions” (1974, p. 72) must be designed and present

for instruction in cognitive strategy development. Those conditions suggest that in order

to learn to think, a learner must be guided into opportunities to think.

17

Glaser and Pellegrino (1987) suggest that the improvement of the skills of

learning will take place through the development of procedural (problem-solving)

knowledge. While their research attempted to identify cognitive components of

performance on tasks used to assess aptitude, their ultimate goal was to use the

knowledge gained to design instruction to directly or indirectly teach the processes that

facilitate learning. This research is one of the early investigations into what is now known

as metacognitive intervention in training design. The authors analyzed the processes used

by high and low performing individuals and concluded that the problem solving strategies

employed differed for each group. An oral problem solving technique was employed to

identify the processes used by each group in a standardized analogy test. The findings

suggest that high-ability individuals limit their approach to a few plausible mathematical

relationships; whereas, the low-ability individuals do not solve analogies with a

systematic approach. The implications which can be drawn involve the possibility of

influencing mental processing skills by teaching individuals to employ better methods of

searching memory and seeking connections.

Livingston (1997) posits that learners with greater metacognitive abilities tend to

be more successful in their cognitive activities and that individuals can learn how to

improve cognitive activities. Metacognition can enable learners to gain greater benefit

from instruction and influences the use and maintenance of cognitive strategies

(Livingston). Cognitive Strategy Instruction is a technique that emphasizes the

development of thinking skills and processes as a means to enhance learning

(Livingston).

18

Schmidt and Ford (2003) studied the effect of trainee characteristics on

metacognitive activity in a learner controlled training environment and found that

metacognitive activity was mediated by the level of the trainee’s goal avoidance

orientation. The implication of this finding for the present study is that the outcomes of

metacognitive interventions are not consistently positive among all learners, but that

individual differences may account for variations in metacognitive activity. These

individual differences may include trainee motivation and perceptions of self-efficacy.

Flavell (1979) explains that metacognitive processes can lead a learner to select, evaluate

and revise cognitive strategies with regard to a learner’s ability and interest in what is

being learned.

Relevant to this current study, metacognition had been linked to self-efficacy

perceptions early in Flavell’s (1987) thinking about the construct. Flavell referred to

metacognition as having a psychological space. He hypothesized the interactions that

may link metacognition and other constructs. The constructs included “executive

processes; formal operations; consciousness; social cognition; self-efficacy; self-

regulation; reflective self-awareness and the concept of psychological self or

psychological subject” (Flavell, p. 25).

Much of the empirical research on interventions has been conducted among

children and/or in academic settings, suggesting the opportunity for a more generalizable

study among adult learners in an organizational training environment.

Learner Controlled Training and Metacognitive Interventions

Learner controlled training has increased in usage as a result of widespread

availability of workplace and personal computers (Schroeder, 1994). Learner control

19

refers to the degree to which learners are able to choose the method, timing, practice and

feedback during training (Milheim & Martin, 1991). A major advantage of learner

controlled training over traditional forms of training is its potential to allow trainees to

proceed through training at their own rate, controlled by their own needs and preferences

(Eom & Reiser, 2000).

However, empirical research as shown mixed results for learner controlled

training (Eom & Reiser, 2000). This section will briefly present the underlying theory

and research for learner controlled training, leading to the assessment that metacognitive

interventions may have the potential to improve outcomes in learner controlled training

environments.

Hilgard and Bower (1966) explain that the beginnings of learner controlled

training, then referred to as “programmed learning”, emerged from a behaviorist

perspective. Programmed instruction is characterized by having information broken down

into smaller, simpler groups of information. While considered a more effective teaching

method than historical methods, several weaknesses emerge: behavioral models of

programmed instruction isolate factual information thus, learners learn in isolation, not in

the context of the interrelationships of the material (Hilgard & Bower). The programmed

instruction technique evolved into many tools, the most common today called Computer-

Assisted Instruction (CAI), a technique which can be designed to incorporate both the

original behaviorist view of learning (e.g., Skinner, 1968, 1969) and the more widely

accepted cognitive view (e.g., Bandura, 1986). Learner controlled training does not

presume the use of the computer as the training delivery method, but the computer’s near

20

total availability in organizations today has made its use more widespread than traditional

paper, audio and video techniques.

Eom and Reiser (2000) posit that the conflicting or mixed results of learner

control are possibly due to the characteristics of the learner population. Kinzie (1990)

found that the degree of experience and comfort with learner control instruction

influences the effectiveness of the instruction. Kinzie, Sullivan and Berdel (1988) found

that pre-test reading levels were a more significant predictor of performance than the

level of learner control and called for an examination of self regulatory skills in learner

controlled environments.

Bandura (1986) describes the widely accepted view that social learning practices

are improved by structuring the learning environment in such a way as to allow learners

to judge themselves in reference to their own capabilities and standards, rather than in

comparison with others.

Self regulatory skills are a learner characteristic that may affect a learner’s ability

to benefit from learner controlled instruction (Armstrong, 1989; Eom & Reiser, 2000).

Self regulated learning strategies have been defined as metacognitive, motivational, and

behavioral techniques that a learner can use to control his or her learning process

(Zimmerman & Martinez-Pons, 1988). Eom and Reiser (2000) explain that intrinsic

motivation and self efficacy have an impact on self regulated learning. Boekaerts (1995)

indicated that self regulated learning strategies involve affective variables (e.g., anxiety)

as well as cognitive variables. Bandura (1986) viewed self regulation as composed of

multiple processes such as self observation, self judgment and self reaction. Motivational

factors such as attribution and self efficacy influence self regulated learning strategies;

21

thus self regulated learners can be considered self motivated and are self directed in a

metacognitive sense as well (Eom & Reiser, 2000). Jegede, Taplin, Fan, Chan and Yum

(1999) found a higher level of use of metacognitive strategies among students describing

themselves as high achievers in a learner controlled environment.

Computer based training designs allow users to exert significant control over

sequence of learning, content and pace of instruction (Bell & Kozlowski, 2002). In a

review of the literature examining effectiveness of learner control in CAI, Lunts (2002)

reports that the amount of learner control affects the effectiveness of the method, with

greater control associated with improved creativity and learner initiative. Lunts further

reports that, generally, the literature suggests that learner control is a useful tool for

adapting a learning environment to students’ needs. Perceived learner control positively

affects motivation and the amount of effort invested in the learning task (Perez, Kester &

Van Merrienboer, n.d.). Eom and Reiser (2000) explain that poor performance under

learner control appears due to the learners’ failure to use effective learning strategies and

poor metacognitive skills. However, when learner control is supplemented with in

learning interventions, individual performance increases (Bell & Kozlowski). In his

summary of five meta-analyses of the impact of technology on student achievement,

Schacter (1999) reports that CAI, integrated learning systems and instruction in higher

order thinking show positive gains in researcher constructed tests, standardized tests and

national tests.

Gagne´ (1977), in reporting a series of experiments of in-training interventions ,

proposed that learners are able to exercise more successful control over their own

learning process by using a cognitive strategy that is presented to them during the

22

learning experience or by using a cognitive strategy that may have previously been

learned. The use of frequent numerical or technical questions interspersed in a long

reading passage resulted in an improved retention of the information compared to those

not exposed to interruptive questions. Gagne´ (1977) suggests that the question

interventions had the effect of “activating a strategy of attending” (p. 168) to the facts to

be learned. This anticipated Flavell’s (1979) theory of metacognitive processes and the

use of in-training interventions to stimulate a learner’s ability to increase learning

effectiveness.

Watson (n.d.) reported significant positive performance improvement among

students receiving metacognitive prompts during a computer based learned controlled

tutorial. Embedded metacognitive training resulted in a significant increase in

performance versus both strategy training and a no-training control group among primary

school students (Mevarech, 1999). Hill and Hannafin (1997) report improvements in

posttest performance as a result of embedded cues. Metacognitive training for math

students resulted in increased performance versus traditional learning methods in a two-

year study among eighth-grade students (Mevarech & Kramarski, (2003).

As with metacognitive studies, much of the empirical research on learner control

has focused on students in a school learning environment. In fact, this situation led Lunts

(2002) to characterize learner control research as “excessively targeting younger and

inexperienced learners” (p. 68). Lunts further implies that learner control should have a

greater chance for success with adult learners, as they are likely to be more motivated and

able to comprehend the higher order skills (versus factual information) contained in many

organizational training programs.

23

Motivation and Self Efficacy

Motivation has been described as a cognitive process which directs choices

among alternative paths of voluntary actions (Vroom, 1964). A number of theorists have

explained motivation in terms of the expectancy-valence model (Atkinson, 1964;

Fishbein, 1967; Vroom). This model suggests that one’s degree of motivation is

dependent upon both the belief that specific actions will produce particular outcomes and

the value of those outcomes to the individual. Valence is described as the anticipated

satisfaction (positive or negative) of an outcome, whereas value refers to the actual

satisfaction derived. A learner’s perception of self-efficacy can be measured in terms of

their judgments of capabilities and the strength of that belief (Bandura, 2003).

Bandura (1988) joins motivation and self-efficacy as follows:

People’s beliefs in their capabilities affect their motivation as well as the activities they undertake. Significant human accomplishments require perseverant effort. It is renewed effort in the face of difficulties and setbacks that usually brings success. … The important matter is not that difficulties arouse self-doubt –which is a natural immediate reaction – but the recovery from difficulties. Some people quickly recover their self-confidence; others lose faith in their capabilities. It is resiliency of self-belief that counts. (p.282)

Evaluations of self-efficacy affect an individual’s initiation of behavior, the amount of

effort to be expended, and the duration of that effort in the face of disconfirming evidence

(Bandura, 1977). Wood and Bandura (1989a) explain self-efficacy as a regulatory

mechanism affecting motivation:

There is a difference between possessing skills and being able to use them well and consistently under difficult circumstances. To be successful, one not only must possess the required skills, but also a resilient self-belief in one’s capabilities to exercise control over events to accomplish desired goals. People with the same skills may, therefore, perform poorly, adequately, or extraordinarily, depending on whether their self-beliefs of efficacy enhance or impair their motivation and problem-solving efforts. (p. 364)

24

Self-efficacy, as explained by Bandura (1986), mediates the relationship between one’s

knowledge and actions. Knowledge and skills are needed, but insufficient alone for

successful performance. People often perform at less than optimum levels, although they

know the correct actions because their self-efficacy perceptions affect their actions.

Bandura (1988) lists four sources of perceived self-efficacy: mastery experiences,

vicarious experience, social persuasion, and physiological state. Mastery experiences,

also called success experiences, help an individual gain a sense of capability. When an

individual achieves success through sustained effort, setbacks and failures can be

managed more easily. Individuals partly judge their capabilities through comparison with

others by observing them through vicarious experiences. Self-efficacy beliefs can also be

affected by modeling – access to successful models can increase an individual’s

perception of self-efficacy. Conversely, observing others’ failures despite high efforts can

lower an individual’s perception of probable success. Social persuasion concerns the

impact of the opinions of others regarding the individual’s likelihood of successfully

completing a task. Realistic encouragement can lead to greater individual effort. The

concept of physiological state also affects an individual’s perception of self-efficacy.

Emotional arousal and tension can signal a possible poor performance. Particularly in

strength-related activities, individuals judge their possible efficacy in terms of perceived

fatigue levels, and presence/absence of pain.

Relevant to this study, the effect of social persuasion, particularly from one’s

organizational supervisor can be a key determinant of an individual’s perception of self-

efficacy as they begin a training task. Bandura’s (1988) key point on this factor is that

individuals who have a strong belief in their efficacy work, think and behave differently

25

than those who doubt their capabilities and that social persuasion, e.g. supervisory

behavior, can be a factor in an individual’s perception of self-efficacy. This view of

supervisory support as an independent variable affecting self-efficacy is explained and

elaborated upon in depth in the following section: Supervisory Support.

Self-efficacy is learner’s judgment of their capability to perform actions related to

training (Hill & Hannafin, 1997). Self-efficacy beliefs affect activities through cognitive,

motivational and decisional processes (Bandura & Locke, 2003). In his elaboration of

Kolb’s Learning Cycle model, Vince (1998) proposes that learner anxiety; fear and doubt

at the start of a learning process can either promote or discourage learning. Learner

anxiety in training may impact learning and is likely to be negatively associated with

learning (Warr & Bunce, 1995).

Bandura and Wood (1989) found that a learner’s perception of efficacy, in this

case, achievable standards of performance in operating a simulated firm, affected use of

strategically effective thinking. Results indicated both an initial higher level of strategic

thinking and subsequent increased use of strategic thinking for individuals with highest

perceived initial self-efficacy.

The positive expectation of other organizational members may result in improved

performance (the Pygmalion effect); and self-efficacy can be positively affected through

the persuasive effect of the other organizational members (Gist, 1987). Supervisors and

organizations are clear sources of support for employees and affect employee

commitment to organizational activities (Stinglhamber& Vandenberghe, 2003).

The instructional processes involved in training should increase trainee self-

efficacy and improve expectations that the training will have a positive outcome

26

(Tannenbaum & Yuki, 1992). Employees who begin training with the belief that they are

able to successfully learn the content are likely to have more successful training

experiences (Tannenbaum & Yuki). Martocchio (1994) found a significant decline in

anxiety when trainees began training with the belief that they could build on their present

abilities. A key issue, therefore, emerging from this review is whether the level of

metacognitive activity in a non-academic, learner controlled training environment is

influenced by the trainee’s perception of self-efficacy.

Age, Gender and Computer Self Efficacy

Additional variables may have an effect on self efficacy perceptions in a learner

controlled training environment: age, gender and computer self efficacy. While

demographic characteristics have been studied as variables in training studies, they most

often have been viewed as statistical control variables. The two most frequently studied

variables have been age and gender (Colquitt, LePine & Noe, 2000). In their meta

analytic path analysis of training motivation, Colquitt et al. found that older trainees

demonstrated lower motivation, learning and self efficacy. Maurer, Weiss and Barbeite

(2003) reported that older workers had lower self efficacy with regard to learning abilities

and cognitive processes. Other empirical studies have reported a negative relationship

between age and learning (Gist, Rosen & Schwoerer, 1988; Martocchio, 1994).

Age effects were noted in a study of computer attitudes (Czaja & Sharit, 1998), as

older adults reported less comfort, less competence and less control over computers than

did younger adults. Similarly, Henderson, Deane, Barrelle and Mahar (1995) found that

older users have low confidence in their ability to use computer technology. Comber,

Colley, Hargreaves and Dorn (1997) found that older employees demonstrated less

27

interest in and poor attitudes towards computer-based training. Thus, the age of the

employee may affect self efficacy in training delivered solely by computer.

Examining self efficacy with regard to gender, Choi (2004) reports that masculine

sex role traits are strongly related to independence, assertiveness and competitiveness,

while feminine sex role traits are related to dependence and interpersonal relationships.

Thus, gender may have an effect on self efficacy perceptions in training that is

accomplished on an individual basis in the absence of a live instructor. Studies of the

effect of gender on computer self efficacy have shown mixed results. Qutami and Abu-

Jaber (1997) reported that male and female college students performed equally in

computer skills training, while Comber et al. (1997) reported lower computer self

assurance for females than males. In independent tasks involving computer based Internet

use, Ford et al. (2001) found that females studied exhibited poorer performance and

lower self efficacy than males on most tasks, but no difference in overall self efficacy

related to computer use. Henry and Stone (1999) found that females had lower computer

self efficacy and lower outcome expectancy than males in using computer systems at

work. Pajares (2002), however, concludes that gender differences in self efficacy can be

eliminated or minimized when employees receive unequivocal feedback about their

capabilities as well as progress in learning.

The above suggests that age and gender could be important variables in the study

of self efficacy in learner controlled training. These were examined as moderating

variables in the research design. Since age and gender are expected to influence

metacognitive activity, self efficacy and supervisor support, the relevant research

questions appear in the next section.

28

Supervisory Support

While it would appear natural for an individual to assume responsibility for his or

her own learning, this would unreasonably dismiss the influence of the social

environment. Gagne´ (1977) explains the important effect of events in the external

environment on what and how learning takes place. From early infancy and throughout

adulthood, individuals are subject to the influences of others (parents, peers, teachers and

supervisors) on learning.

Bandura (1986) lays a theoretical foundation for the effect of external influences

on personal effort:

“People who are persuaded verbally that they posses the capabilities to master given tasks are likely to mobilize greater sustained effort than if they harbor self-doubts and dwell on personal deficiencies” (p.231).

In a study of pretraining motivation (Facteau, Dobbins, Russell, Ladd & Kudish,

1995) found that supervisory support was positively related to training motivation;

whereas, peer support, subordinate support and top management support were negatively

related to motivation.

Wood and Bandura (1989b) showed that interpretation of personal efficacy can

affect performance and that perceptions of efficacy can be affected by external factors.

Wood and Bandura’s study induced conceptions of ability among two groups of MBA

students by instructing one group that decision-making skills were acquirable through

practice (acquirable skill condition), while the second group was instructed that decision-

making reflected basic cognitive capacities already possessed (entity condition). The

sample did not differ in pretest perceived self-efficacy. The findings provided evidence

that the conception of ability has substantial impact on self-regulatory behaviors.

29

Understanding ability as an acquirable skill resulted in a highly resilient sense of

self-efficacy and high performance outcomes among that group of students. Conversely,

the group in the entity condition viewed substandard performance as due to their own

limitations and performance declined as the decision-making tasks became more complex

and difficult. Relevant to the present study was the finding that the use of analytic

strategies for decisions also varied by group. That is, the acquirable skill group developed

and successfully used strategies to improve performance, with the entity group failing to

successfully develop and utilize strategies. Thus, perceptions of efficacy affect learning

strategies as well as task performance. Jacobs, Prentice-Dunn, and Rogers (1984)

demonstrated that efficacy beliefs can be artificially altered, with subject performance

consistent with the level of efficacy imposed from the outside. Bandura and Locke (2003)

posited that competencies can be can be increased by instilling a strong sense of learning

efficacy.

The effects of supervisor behavior on subordinate attitudes and behavior was the

subject of the Ohio State Leadership Studies. Halpin and Winer (1957) identified two

independent dimensions of leader behavior: Consideration and Initiating Structure.

Consideration encompasses friendship, mutual trust and respect as aspects of supervisory

behavior towards subordinates. Initiating Structure refers to the organization and

definition of subordinate activities. Subordinate satisfaction has been found to be related

to supervisory consideration in a number of studies in the 1950’s (Fleishman, 1957;

Halpin & Winer, 1957; Halpin, 1957).

Bandura (1986) offers a social psychologist’s explanation of the relationship of

subordinate satisfaction and supervisory behavior. Bandura explains that, in human

30

development, physically rewarding events often are accompanied by expressions of

interest and approval of others, while non-rewarding events are associated with

disapproval. People choose particular actions for approval and avoid actions which elicit

disapproval. Thus, the predictive value of the social reactions of others serves as an

incentive for a person’s actions. Bandura (1986) stated:

The approval or disapproval of those who can exercise reward and punishment power has more influence on one’s actions than similar expressions by those who cannot affect one’s life....It is difficult to conceive of a society populated with people who are completely unmoved by the respect, approval and reproof of others (p. 235)

In 1961, Likert found large differences between satisfied and dissatisfied work

groups’ reporting of supervisory behaviors. For example, 61% of employees with

favorable attitudes reported that their supervisor recommends promotions, transfers and

pay increases while only 22% of employees with unfavorable attitudes reported that

particular supervisory behavior. This pattern of relationships between positive employee

attitudes and supportive supervisory behavior was consistent throughout the study. In

their meta-analysis of organizational behavior modification, Stajkovic and Luthans

(1997) found that social rewards, such as recognition and attention, were statistically

equal to financial rewards in generating increased task performance in both

manufacturing and service organizations.

Feedback from authority figures can be viewed as a form of persuasion that

affects motivation (Latham & Locke, 1991). Perceptions of a task environment can be

influenced by verbal or written persuasion from others in the social environment

(Martocchio, 1992). Supervisory cues have been found to affect employee intrinsic and

extrinsic satisfaction in a task environment (Griffin, 1983). Learners exhibit greater

31

effort and are more likely to succeed if they receive encouragement from other

organizational members (Wood & Bandura, 1989a). Perceptions of positive supervisory

support have been linked to increased trainee motivation prior to training (Cohen, 1990).

Supervisory support for training has been positively associated with successful learning

transfer (Huczynski & Lewis, 1980). Managerial knowledge of the benefits of online

training and interest in implementation fosters faster and more effective implementation

of online training designs (Newton, Hase, & Ellis, 2002).

Gist and Mitchell (1992) explain that self-efficacy is an individual’s judgment of

perceived capability to perform a specific task and that, in an organizational context,

information obtained from the individual, the task itself and others in the organizational

environment may affect the individual’s assessment of capability. The authors further

propose a model of the formation of self-efficacy that contains three broad categories of

factors: analysis of task requirements, assessment of personal and situational resources,

and attributional analysis of experience. Within attributional experience, verbal

persuasion cues may include feedback about an individual’s abilities. Gist and Mitchell

develop the concept of pure persuasion, that is, the use of emotional and cognitive

arguments to convince an individual that he or she can perform a task at a given level.

While the authors hold that this concept may result in more weakly held efficacy beliefs,

there is a clear potential for impact on efficacy beliefs.

Gist and Mitchell (1992) further propose that one’s judgment of self-efficacy is

composed of variable and stable components and that equal self-efficacy judgments may

result in unequal performance due to the individual differences in variable and stable

levels. In this research, therefore, it is hypothesized that a worker’s level of self-efficacy

32

is affected by the variability in social persuasion – operationalized in this study as

supervisory support.

Age and gender are expected to have a moderating effect in this study. To address

this, gender was examined in the context of both perceived supervisory support and its

effect on self efficacy and self efficacy and its effect on metacognitive activity.

The above review suggests the following research questions and relevant

hypotheses:

The first research question is: Is supervisory support related to learner control

self efficacy and computer self-efficacy in a learner controlled training environment?

Two research hypotheses were examined in addressing this question:

H1: There is a positive relationship between supervisory support and computer

self-efficacy in a learner controlled training environment.

H2: There is a positive relationship between supervisory support and learner

control self-efficacy in a learner controlled training environment.

The second research question is: Are computer self-efficacy or learner control

self-efficacy related to metacognitive activity in a learner controlled training

environment? The corresponding research hypotheses are:

H3: There is a positive relationship between computer self-efficacy and

metacognitive activity in a learner controlled training environment.

H4: There is a positive relationship between learner control self-efficacy and


33

The third research question is: Do the relationships between supervisory support,

learner control self-efficacy, and metacognitive activity vary as a function of the gender

of the learner? The corresponding research hypotheses are:

H5: The relationship between supervisory support and computer self-efficacy

varies by gender.

H6: The relationship between supervisory support and learner control self-efficacy

varies by gender.

H7: The relationship between computer self-efficacy and metacognitive activity

varies by gender.

H8: The relationship between learner control self-efficacy and metacognitive

activity varies by gender.

The fourth research question is: Do the relationships between supervisory

support, learner control self-efficacy, and metacognitive activity vary as a function of the

age of the learner? The corresponding research hypotheses are:


varies by age.

H10: The relationship between supervisory support and learner control self-

efficacy varies by age.


varies by age.


activity varies by age.

34

The fifth research question is: Does computer self-efficacy have an effect on

learner control self-efficacy which subsequently has an effect on metacognitive activity?

The research hypothesis is:

H13: Computer self-efficacy has a positive, indirect effect on metacognitive

activity through learner control self-efficacy.

35

CHAPTER III: METHODOLOGY

The literature review above identified a number of limitations of the existing

theory and empirical research on metacognitive interventions. While it is not feasible to

address all of these limitations in this study, the focus is to understand the relationships of

the variables in an organizational environment. The generalizability of findings to

organizational training is the primary limitation of prior studies. The majority of research

with metacognitive interventions has been among children and young adults in

educational settings. Indeed, the theoretical foundations of the construct by Flavell (1976,

1977, 1979 & 1987) are almost totally based on observations and research among pre-

adult populations. In this study, pre-adult is defined as individuals who are primary or

secondary school students. The present study explored the effect of variables on

metacognitive activity among adult learners in an organizationally sponsored setting.

A second limitation of prior research has been the limited consideration of the

role of self efficacy in the learner’s approach to learner controlled training. The

influences of trainee motivation and self efficacy have been virtually ignored in past

studies of learner control. This study attempted to identify the relationship between two

types of self efficacy (i.e. computer self-efficacy and learner control self-efficacy) and

metacognitive activity in a learner controlled training environment. Further, the influence

of supervisory support on self efficacy perceptions, while reasonably well researched, has

not been extensively examined in the context of training. Finally, age and gender in

learner controlled training have been almost universally viewed as control variables;

36

whereas this study examined them as moderating variables. That is, rather than merely

controlling for age and gender, their specific effects were examined.

Research Design

This was a survey-based field study designed to gather data on metacognitive

activity, learner control self efficacy, computer self-efficacy, and supervisory support in a

learner controlled training environment. Age and gender were assessed and examined as

potential moderating variables. The study is a non-experimental design; the data support

associational inferences among the variables, but not causal relationships.

The study was conducted among managers whose graduate school education is

being fully or partially sponsored by their employer. Embedded in the graduate school

curriculum is the Virtual Leader© program from Simulearn Inc. The philosophy behind

Virtual Leader is that leadership is a complex skill that can become intuitive through

practice. The program provides applied training focusing on situational awareness, group

dynamics, and managing and empowering others. The Virtual Leader program is

designed as a learner controlled training simulation with managers accessing the program

at their convenience over the 8-10 hour total training content. The Virtual Leader

program is a CD-ROM based training tool where managers utilize a traditional

desktop/laptop computer located on the organization’s premises or at the educational

institution.

The program contains three distinct sections: Leadership Fundamentals, Learning

the Principles and Applying the Principles. The final section consists of six simulated

meeting scenarios where the manager is placed in the position of a manager conducting a

meeting in a graphic interface. The manager uses the principles from the preceding

37

sections to manage the simulated participants to a successful meeting outcome. The

simulated meeting can be halted to allow the manager to review either the Fundamentals

or Principles sections preceding the meeting simulation. In this manner, managers can

alter their leadership strategies during the simulation to adapt to changing meeting

conditions.

Hacker (2003) defined metacognitive activity as “knowledge of one’s knowledge,

thought processes, and cognitive and affective states; the ability to consciously and

deliberately monitor and regulate one’s knowledge, processes, and cognitive and

affective states” (p. 6). Metacognitive activity within Virtual Leader is engendered

through on-screen progress bars detailing emotional conditions of the meeting

participants in real time, options for agreeing or disagreeing with opinions from the

simulated participants, changing the topic under discussion, and providing praise or

criticism for any participant. Feedback to the trainee’s responses (from the simulated

participants) is immediate, allowing the trainee to continue or change the strategy being

utilized.

Operationalization of Variables

A short demographic survey (Appendix A) was administered to the respondents at

the beginning of the study. Gender was assessed as male or female. Age was assessed as

a continuous variable. Ethnicity was assessed as White, Black, Hispanic, American

Indian (or Eskimo or Aleut), Asian (or Pacific Islander) or other.

Learner control self efficacy was measured using a pre-training, self-administered

questionnaire consisting of seven items to assess a learner’s confidence to perform

successfully on this task (see Appendix B). Learners rated their self efficacy on a five

38

point Likert-type scale ranging from strongly disagree to strongly agree. The instrument

is a selection of appropriate items from previous instruments used to collect self efficacy

data (Ford et al., 1998; Gist et al., 1989; Gist, 1989; Tannenbaum & Yuki, 1992). The

internal reliability of these instruments ranged from .88 to .90 (Cronbach’s α).

Computer self-efficacy was measured using a modification of the Computer Self-

Efficacy scale (Computer Self-Efficacy Survey, n.d.). The instrument (see Appendix C)

uses a five-point Likert-type scale ranging from Very Little Confidence to Quite a Lot of

Confidence. The internal reliability of the section used is .94 (Cronbach’s α).

Supervisor support was measured using a pre-training, self-administered

questionnaire consisting of five items to assess the learner’s perception of supervisory

support for the learner in a learner controlled training environment. The instrument (see

Appendix D) consisted of a selection of questions from Jiang and Klein’s (2000) scale

measuring supervisor support, internal reliability of entire instrument was α =.83. This

questionnaire was administered after the individual completed the initial self efficacy

questionnaire to minimize possible bias caused by the order of the information sought.

For example, one could reasonably speculate that asking questions about supervisory

support initially could impel the respondent to answer subsequent questions about self

efficacy in a manner artificially consistent with prior answers. Answering the questions in

an order other than intended, therefore, threatens both the reliability and the validity of

the instrument.

Metacognitive activity was measured using a post-training, self-administered

questionnaire. The training activity was administered in a learner controlled environment

which utilizes metacognitive interventions. A selection of questions appropriate for this

39

study were taken from Ford et al.’s (1998) instrument measuring metacognitive activity

(see Appendix E). Internal reliability for the entire instrument was α =.83. Learners

reported their metacognitive activities for each of ten items on a five point Likert-type

scale ranging from strongly disagree to strongly agree.

Sample

This study was conducted among managers whose graduate school business

education is being fully or partially sponsored by their employer. Questionnaires were

distributed to geographically dispersed graduate business schools across the United

States. The Virtual Leader leadership training program is used by these schools as part of

the leadership curriculum. While all students are required to participate in the training,

only those managers with organizational sponsorship were included in the data for this

study. The study ran for about three months; yielding a sample size of 120 qualified

subjects.

Procedure

The study consisted of pre and post training questionnaires administered

immediately prior to and upon completion of the leadership training session. The four

questionnaires were assembled in a set (Computer Self Efficacy, Learner Control Self

Efficacy, Supervisory Support and Metacognitive Activity), stapled and each set

numbered for identification. Oral and/or written instructions were given to the course

instructor indicating that the respondents should complete the measures in the order of

the stapled packet. All pages within a set contained the same identification number so

that the data for each individual can be identified, even if the pages become separated.

40

The course instructors distributed and collected the questionnaires during the

Virtual Leader training session. All instructors were briefed on the study and were able to

administer the questionnaires and answer any procedural questions. The instructors

mailed the completed questionnaires to the researcher.

All data were collected anonymously with age, gender and ethnicity collected on

the cover page containing instructions before the first (learner control and computer self

efficacy) questions are completed. Respondents were asked to complete the first

questionnaire (containing the demographic items, learner control self-efficacy, and

computer self-efficacy scales) before the training begins; respondents were then asked to

complete the final questionnaire (containing the supervisory support and metacognitive

activity scales) immediately upon completion of the Virtual Leader © training.

Research Questions

The current study addressed five research questions incorporating the variables of

learner control self-efficacy, computer self-efficacy, supervisory support, metacognitive

activity, age, and gender. The first four research questions are based on the model shown

in Figure 1. The first research question is: Is supervisory support related to learner

control self efficacy and computer self-efficacy in a learner controlled training

environment?

Two research hypotheses were examined in addressing this question:

H1: There is a positive relationship between supervisory support and computer

self-efficacy in a learner controlled training environment.

H2: There is a positive relationship between supervisory support and learner

control self-efficacy in a learner controlled training environment.

41

The second research question is: Are computer self-efficacy or learner control


environment? The corresponding research hypotheses are:

H3: There is a positive relationship between computer self-efficacy and


H4: There is a positive relationship between learner control self-efficacy and


The third research question is: Do the relationships between supervisory support,

learner control self-efficacy, and metacognitive activity vary as a function of the gender

of the learner? The corresponding research hypotheses are:


varies by gender.

H6: The relationship between supervisory support and learner control self-efficacy

varies by gender.


varies by gender.


activity varies by gender.



age of the learner? The corresponding research hypotheses are:


varies by age.

42

H10: The relationship between supervisory support and learner control self-

efficacy varies by age.


varies by age.


activity varies by age.

The fifth research question is based on the model shown in Figure 2. In this

model, computer self-efficacy was hypothesized to have an effect on learner control self-

efficacy, which subsequently has an effect on metacognitive activity. The research

hypothesis is:

H13: Computer self-efficacy has a positive, indirect effect on metacognitive

activity through learner control self-efficacy.

Data Analysis Plan

The current study employed both descriptive and inferential statistical techniques.

Descriptive statistics consisted of (a) a description of the sample and (b) a description of

the scores on the four primary scales of interest (i.e. supervisory support, computer self-

efficacy, learner control self-efficacy, and metacognitive activity). The description of the

sample consisted of frequencies and relative frequencies for gender, age group, and

ethnicity. The description of scores on the supervisory support, learner control self-

efficacy, computer self-efficacy, and metacognitive activity scales consisted of means

and standard deviations, as well as an assessment of reliability. Reliability was assessed

with Cronbach’s α (internal consistency).

43

All inferential analyses consisted of two-tailed tests and an α level of .05. Two-

tailed tests were appropriate for this analysis as the directions of the relationships were

unknown, although where one-tailed tests were required to obtain statistical significance,

the results of these tests are reported as well. Initially, the correlations between the

supervisory support, learner control self-efficacy, computer self-efficacy, and

metacognitive activity were computed for the entire sample, and then separately for

males and females and for younger and older individuals. Then, a series of path models

were examined in which perceived supervisory support is specified as a predictor of

learner control self-efficacy and computer self-efficacy, which in turn are specified as

predictors of metacognitive activity.

Path analysis is essentially a combination of several simultaneous regression

analyses where selected variables can serve as both predictors and outcomes. In the

current study, learner control self-efficacy and computer self-efficacy took on this dual

role, serving as outcomes of supervisory support and predictors of metacognitive activity,

as shown in Figure 1. The hypotheses of the current study were tested by examining the

statistical significance of the standardized regression coefficients (β) for each variable. Z

scores express value in terms of how many standard deviations it is from the mean for

that set of data. It is useful when comparing results from variables with different scales as

is the case in the current study.

The first hypothesis is: There is a positive relationship between supervisory

support and computer self-efficacy in a learner controlled training environment. The

effect marked ‘a’ in Figure 1 represents the relationship between supervisory support and

computer self-efficacy. The relationship between supervisory support and computer self-

44

efficacy was tested via the statistical significance of the standardized regression

coefficient (β) calculated for the effect marked ‘a’. The second hypothesis is: There is a

positive relationship between supervisory support and learner control self-efficacy in a

learner controlled training environment. The relationship between supervisory support

and learner control self-efficacy is represented by the effect marked ‘b’ in Figure 1, and

this relationship was tested in the same manner as the first hypothesis.

The third hypothesis is: There is a positive relationship between computer self-

efficacy and metacognitive activity in a learner controlled training environment. This

relationship is represented by the effect marked ‘c’ in Figure 1. The fourth hypothesis is:

There is a positive relationship between learner control self-efficacy and metacognitive

activity in a learner controlled training environment, and this relationship is represented

by the effect marked ‘d’ in Figure 1. To test these relationships for statistical significance,

the standardized regression coefficients (β) associated with the effects marked ‘c’ and ‘d’

were calculated.

The fifth, sixth, seventh, and eighth hypothesis relate to the effect that gender may

have on the relationships between the four primary variables in the current study (i.e.

supervisory support, computer self-efficacy, learner control self-efficacy, and

metacognitive activity). These hypotheses state that the standardized regression

coefficients labeled ‘a’ ‘b’ ‘c’ and ‘d’ in Figure 1 are not the same for males and females.

To test for differences in these relationships between males and females, a path model

was computed in which the effects marked ‘a’ ‘b’ ‘c’ and ‘d’ were computed separately

for males and females. Then, the fit of this model was compared to the fit of models in

which the effects marked ‘a’ ‘b’ ‘c’ and ‘d’ were computed to be equal for males and

45

females. Four models were computed, with the effects marked ‘a’ ‘b’ ‘c’ and ‘d’

computed to be equal between males and females one at a time. If the model in which

the effect is computed to be equal for males and females fits significantly worse than the

model in which the effects are computed separately for males and females (as determined

by the χ2 difference test), then the hypotheses related to gender effects will be supported.

The ninth, tenth, eleventh, and twelfth hypotheses relate to differences in the

relationships between the primary variables in the current study as a function of age

group (younger individuals versus older individuals). The method of testing these

hypotheses was identical to the method used to test the fifth through eighth hypotheses.

That is, models in which the effects marked ‘a’ through ‘d’ in Figure 1 will be compared

to models in which these effects are computed to be equal across the two groups, and the

statistical significance of the difference in fit between the two models will be determined

via the χ2 difference test.

The thirteenth hypothesis deals with the positive, indirect effect of computer self-

efficacy on metacognitive activity through learner control self-efficacy. The statistical

significance of the indirect effects was computed via the Sobol z test (Sobol, 1982). The

Sobol test involves computing the product of the two unstandardized regression

coefficients (the effects marked ‘a’ and ‘b’ in Figure 2) and determining if that product is

larger than would be expected if there was no indirect effect.

46

Figure 1. Path Model for Research Questions 1-4

c

b d

Figure 2. Path Model for Research Question 5

a b

Computer Self-Efficacy

Supervisory Support

Learner Control Self- Efficacy

Metacognitive Activity

a




47

CHAPTER IV: ANALYSIS OF DATA

Preliminary analyses consisted of (a) a description of the sample and (b) a

description of the scores on the four primary scales of interest (i.e. supervisory support,

computer self-efficacy, learner control self-efficacy, and metacognitive activity). The

description of scores on the supervisory support, learner control self-efficacy, computer

self-efficacy, and metacognitive activity scales consisted of means and standard

deviations, as well as an assessment of reliability. Reliability was assessed with

Cronbach’s α (internal consistency).

All inferential analyses consisted of two-tailed tests and an α level of .05. Two-

tailed tests were appropriate for this analysis as the directions of the relationships were

unknown, although where one-tailed tests would have been significant, they will be

reported along with the two-tailed test results. Initially, the correlations between the

supervisory support, learner control self-efficacy, computer self-efficacy, and

metacognitive activity were computed for the entire sample, using AMOS, and then

separately for males and females and for younger and older individuals. Then, a series of

path models were examined in which perceived supervisory support is specified as a

predictor of learner control self-efficacy and computer self-efficacy, which in turn are

specified as predictors of metacognitive activity.

Path analysis is essentially a combination of several simultaneous regression

analyses where selected variables can serve as both predictors and outcomes. In the

current study, learner control self-efficacy and computer self-efficacy occupied this dual

role, serving as outcomes of supervisory support and predictors of metacognitive activity.

48

The hypothesized relationships in the current study were tested by examining the

statistical significance of the standardized regression coefficients (β) for each effect. The

use of standardized regression coefficients is conventional in path analysis and multiple

regression. The standardized regression coefficients also have the advantage of providing

information on the effect size associated with each relationship, as they are interpretable

as the change in the criterion variable that results from an increase of one in the predictor

variable, when both variables have been converted to a common metric (i.e. z scores).

Z scores express value in terms of how many standard deviations it is from the mean for

that set of data. It is useful when comparing results from variables with different scales as

was the case in the current study.

The thirteenth hypothesis deals with the indirect effect of computer self-efficacy

on metacognitive activity through learner control self-efficacy. The statistical

significance of the indirect effects was computed via the Sobol z test (Sobol, 1982). The

Sobol test involves computing the product of the two unstandardized regression

coefficients and determining if that product is larger than would be expected if there was

no indirect effect.

Preliminary Analyses

Descriptive statistics for the sample demographic characteristics are shown in

Table 1 and Table 2. Two-thirds of the sample (66.7%) was male, and White participants

formed the largest ethnic group (78.3%). The age of the participants ranged from 21 to

62 with a mean of 30.33 years (SD=7.64 years). Table 2 shows the descriptive statistics

for the four composite measures in the current study.

49

Table 1

Descriptive Statistics for Sample Demographic Characteristic (N=120)

Frequency Percentage Gender

Female 40 33.3 Male 80 66.7

Ethnicity

White 94 78.3 Hispanic 13 10.8 Black 7 5.8 Asian 5 4.2 Refused 1 .8

Mean SD Age 30.33 7.64

50

Table 2

Descriptive Statistics for the Composite Measures (N=120)

Number

of Items

Minimum

Maximum

Mean

SD α


32 73 159 130.57 16.26 .96

Learner Self-Efficacy

7 17 35 26.09 3.96 .91

Supervisor Support

4 5 20 13.44 3.46 .89


8 19 40 33.03 4.11 .88

Cronbach’s α reliability coefficients were computed and were high for each of the

four scales (ranging from .88 for metacognitive activity to .96 for computer self-

efficacy). Table 3 shows the correlations between the four composite measures for the

total sample. Computer self-efficacy was positively correlated with both learner control

self-efficacy (r=.49, p<.001) and metacognitive activity (r=.26, p<.01) but not with

supervisory support. Learner self-efficacy was also correlated with supervisory support

(r=.22, p<.05) and metacognitive activity (r=.34, p<.001). In addition, supervisory

support was positively correlated with metacognitive activity (r=.18, p<.05).

51

Table 3

Correlations Between Composite Measures (N=120)

Computer

Self-Efficacy Learner Self-

Efficacy Supervisor

Support Metacognitive

Activity Computer Self-Efficacy 1.00 Learner Self-Efficacy .49*** 1.00 Supervisory Support .13 .22* 1.00 Metacognitive Activity .26** .34*** .18* 1.00 *p<.05, **p<.01, *** p<.001

Table 4 shows the correlations between the four composite measures, but this

time the correlations were computed separately for males and females. For females, the

only statistically significant correlation was between learner control self-efficacy and

metacognitive activity (r=.49, p<.01). Among males, the correlations between computer

self-efficacy and learner control self-efficacy (r=.68, p<.001), supervisory support

(r=.23, p<.01), and metacognitive activity (r=.31, p<.01) were all statistically

significant. In addition, learner control self-efficacy was positively correlated with

supervisory support (r=.23, p<.05) and metacognitive activity (r=.27, p<.05). The

relationship between metacognitive activity and supervisory support was not statistically

significant (but the correlation of .20 would have been statistically significant using a

one-tailed test). Although the fact that there were twice as many males as females

resulted in higher power for the statistical significance of these correlations for males, the

52

correlations for males are larger in size as well with the exception of the correlation

between learner control self-efficacy and metacognitive activity (which was larger for

females) and between supervisory support and metacognitive activity (which was

statistically non-significant and identical for both males and females).

Table 4

Correlations Between Composite Measures as a Function of Gender (N=120)

Computer


Efficacy Supervisor


Activity

Females (n=40) Computer Self-Efficacy 1.00 Learner Self-Efficacy .23 1.00 Supervisory Support .04 .18 1.00 Metacognitive Activity .16 .49** .20 1.00

Males (n=80) Computer Self-Efficacy 1.00 Learner Self-Efficacy .68*** 1.00 Supervisory Support .23* .23* 1.00 Metacognitive Activity .31** .27* .20 1.00 *p<.05, **p<.01, *** p<.001

53

In order to examine the correlations between the four composite variables as a

function of age, the sample was split into those 28 and younger (n=62, 51.7%) and those

29 or older (n=58, 48.3%). This split was chosen primarily to achieve an approximately

equal sample size for the two groups while providing a younger sample whose work

experience almost certainly includes computer use (i.e. workforce entry after 1998).

Table 5 shows the correlations between the four composite measures for the younger age

group and for the older age group.

Table 5

Correlations Between Composite Measures as a Function of Age Group (N=120)

Computer


Efficacy Supervisor


Activity

Younger Respondents (28 years old and younger, n=62) Computer Self-Efficacy 1.00 Learner Self-Efficacy .68*** 1.00 Supervisory Support .09 .30* 1.00 Metacognitive Activity .31* .37** .21 1.00

Older Respondents (29 years old and older, n=58) Computer Self-Efficacy 1.00 Learner Self-Efficacy .26* 1.00 Supervisory Support .19 .15 1.00 Metacognitive Activity .24 .33* .16 1.00

*p<.05, **p<.01, *** p<.001

54

For the younger respondents, computer self-efficacy was positively correlated

with both learner control self-efficacy (r=.68, p<.001) and metacognitive activity (r=.31,

p<.05). In addition, learner control self-efficacy was positively correlated with both

supervisory support (r=.30, p<.05) and metacognitive activity (r=.37, p<.01). Among

the older respondents, computer self-efficacy was again correlated with learner control

self-efficacy (r=.26, p<.05), but not with metacognitive activity (although the correlation

of .24 would have been statistically significant using a one-tailed test). Learner control

self-efficacy, on the other hand, was positively correlated with metacognitive activity

(r=.33, p<.05) but not with supervisory support, and supervisory support and

metacognitive activity were not correlated. Therefore, it appears that the correlations

among the four measures tended to be higher for younger respondents than for older

respondents.

Research Questions

The research questions were addressed using path analysis. Initially, the model

presented in Figure 1 of Chapter 3 was computed, and the resulting standardized

regression coefficients and R2 values are shown in Figure 3 (with full regression results

shown in Appendix F). The answers to the first two research questions are derived from

the coefficients of this model.

55

Figure 3. Path Model for Research Questions 1 and 2 with Regression Coefficients

*denotes statistical significance for the regression coefficient (p<.05).

Research Question 1

The first research question was: Is supervisory support related to learner control

self efficacy and computer self-efficacy in a learner controlled training environment?

Under this research question, there were two hypotheses to be examined (H1 and H2). The

first hypothesis was: There is a positive relationship between supervisory support and

computer self-efficacy in a learner controlled training environment. The regression

coefficient for the prediction of computer self-efficacy from supervisory support was .13,

which was not statistically significant. Therefore, the first hypothesis was not supported,

and we can conclude that there is no relationship between supervisory support and

computer self-efficacy in a learner controlled training environment.

The second hypothesis was: There is a positive relationship between supervisory

support and learner control self-efficacy in a learner controlled training environment.

The regression coefficient for the prediction of learner control self-efficacy from

supervisory support was .22, and this was statistically significant. Thus, the second

hypothesis is supported, and we can conclude that there is a positive relationship between


Supervisory Support



.13 .13

.22* .28*

R2=.10

R2=.05

R2=.02

56

supervisory support and learner control self-efficacy; higher levels of supervisory support

were associated with higher levels of learner control self-efficacy.

Research Question 2

The second research question was: Are computer self-efficacy or learner control


environment? Two hypotheses were stated for this question (H3 and H4). The third

hypothesis was: There is a positive relationship between computer self-efficacy and

metacognitive activity in a learner controlled training environment. The standardized

regression coefficient for the prediction of metacognitive activity from computer self-

efficacy was .13, which was not statistically significant. Therefore, the third hypothesis

was not supported, and we can conclude that there is no relationship between computer

self-efficacy and metacognitive activity.

The fourth hypothesis was: There is a positive relationship between learner

control self-efficacy and metacognitive activity in a learner controlled training

environment. The regression coefficient of .28 was statistically significant for the effect

from learner control self-efficacy and metacognitive activity, and therefore the fourth

hypothesis was supported. There is a positive relationship between learner control self-

efficacy and metacognitive activity; higher levels of learner control self-efficacy were

associated with higher levels of metacognitive activity.

Research Question 3

The third research question was: Do the relationships between supervisory


gender of the learner? There were four hypotheses included under this research question

57

(H5 to H8), and each was examined by performing a multiple-group path analysis in

which the regression coefficient in question was constrained to be equal or allowed to

vary between males and females, and the fit of the two models was compared. The path

models with all regression coefficients free to vary between males and females are shown

in Figures 4 and 5 respectively (with full regression results shown in Appendix F). This

is the baseline model, and sequentially implementing constraints that specific effects are

equivalent for males and females forms the basis for testing H5 to H8.

Figure 4. Path Model for Research Question 3 with Regression Coefficients for Males


Figure 5. Path Model for Research Question 3 with Regression Coefficients for Females



Supervisory Support



.23* .24*

.22* .12

R2=.07

R2=.05

R2=.05


Supervisory Support



.04 .05

.18 .48*

R2=.23

R2=.03

R2=.00

58

For males, the effect from supervisory support to computer self-efficacy was

statistically significant, β=.23, as was the effect from supervisory support to learner

control self-efficacy, β=.22. In addition, the effect from computer self-efficacy to meta-

cognitive activity was statistically significant, β=.24, but the effect from learner control

self-efficacy to metacognitive activity was not, β=.12. For females, the effects from

supervisory to support to computer self-efficacy, β=.04, and to learner control self-

efficacy, β=.18, were not statistically significant (although the standardized regression

coefficient from supervisory support to computer self-efficacy for females would have

been statistically significant under a one-tailed hypothesis test). The effect from

computer self-efficacy to metacognitive activity was also not statistically significant for

females, β=.05, but the effect from learner control self-efficacy to metacognitive activity

was statistically significant, β=.48.

The first hypothesis for this research question was: The relationship between

supervisory support and computer self-efficacy varies by gender. Therefore, the

regression coefficient for the effect between supervisory support and computer self-

efficacy was constrained to be equal for males and females, and the fit of the constrained

model was compared to the fit of the unconstrained model. The χ2 difference test between

the two models was statistically significant, χ2diff(1)=.72, p=.397. This indicates that

constraining the effect from supervisory support to computer self-efficacy to be the same

for males and females did not significantly worsen the fit of the model, and therefore H5

was not supported. We can conclude that the relationship between supervisory support

and computer self-efficacy was the same for males and females.

59

The next hypothesis (H6) was: The relationship between supervisory support and

learner control self-efficacy varies by gender. To test this hypothesis, the effect from

supervisory support to learner control self-efficacy was constrained to be equal for males

and females. Imposing this constraint did not significantly worsen the fit of the model,

χ2diff(1)=.00, p=.999. Therefore, we can conclude that the relationship between

supervisory support and learner control self-efficacy was the same for males and females,

and H6 was not supported.

The third hypothesis for the third research question (H7) was: The relationship

between computer self-efficacy and metacognitive activity varies by gender. To examine

this hypothesis, the effect from computer self-efficacy to metacognitive activity was

constrained to be the same for males and females. This constraint did not significantly

worsen the fit of the model, χ2diff(1)=.93, p=.336, indicating that the relationship between

computer self-efficacy and metacognitive activity was the same for males and females,

and therefore H7 was not supported.

The final hypothesis for the third research question (H8) was: The relationship

between learner control self-efficacy and metacognitive activity varies by gender. To

examine this hypothesis, the effect from learner control self-efficacy to metacognitive

activity was constrained to be the same for males and females. The resulting difference

in fit between the constrained and unconstrained models was not statistically significant,

χ2diff(1)=2.34, p=.126. Therefore, H8 was not supported, and we can conclude that the

relationship between learner control self-efficacy and metacognitive activity is the same

for males and females.

60

Given the seemingly contradictory results between the preliminary tests and the

subsequent direct tests of moderation, a regression analysis with an interaction variable

for gender was considered as a potential tool to clarify the relationship. In addition to the

issue of small sample size (Female=40), χ2diff is generally considered a more useful tool

to detect differences where sample sizes are small. Further, Russell and Bobko (1992)

found that using Likert-type scales (as in the present study) reduces the likelihood of

detecting true interaction effects.

Research Question 4



age of the learner? To address this research question, four hypotheses were tested using

the same methodology as employed for the third research question. Figures 6 and 7 show

the path model estimated separately for younger and older individuals (with full

regression results shown in Appendix F).

Figure 6. Path Model for Research Question 4 with Regression Coefficients for Younger Participants



Supervisory Support



.09 .12

.30* .29*

R2=.10

R2=.09

R2=.01

61

Figure 7. Path Model for Research Question 4 with Regression Coefficients for Older Participants


For younger individuals, the effect from supervisory support to computer self-

efficacy was not statistically significant, β=.09, but the effect from supervisory support to

learner control self-efficacy was statistically significant, β=.30. The effect from

computer self-efficacy to metacognitive activity was not statistically significant, β=.12,

but the effect from learner control self-efficacy to metacognitive activity was significant,

β=.29.

For older participants, neither the effect from supervisory to support to computer

self-efficacy, β=.19, nor the effect from supervisory support to learner control self-

efficacy, β=.15, were not statistically significant. The effect from computer self-efficacy

to metacognitive activity was also not statistically significant for older participants,

β=.17, but the effect from learner control self-efficacy to metacognitive activity was

statistically significant, β=.29.


Supervisory Support



.19 .17

.15 .29*

R2=.12

R2=.02

R2=.04

62

H9 was: The relationship between supervisory support and computer self-efficacy

varies by age. Constraining the effect from supervisory support to computer self-efficacy

did not significantly worsen the fit of the model, χ2diff (1) =.30, p=.584. This indicated

that H9 was not supported, leading to the conclusion that the relationship between

supervisory support and computer self-efficacy is the same for younger and older

individuals.

The next hypothesis (H10) was: The relationship between supervisory support and

learner control self-efficacy varies by age. Constraining the effect from supervisory

support to learner control self-efficacy resulted in a model that did not fit significantly

worse than the unconstrained model, χ2diff(1)=.65, p=.421. Thus, H10 was not supported,

and we can conclude that the relationship between supervisory support and learner

control self-efficacy is the same for younger and older individuals.

H11 was: The relationship between computer self-efficacy and metacognitive

activity varies by age. When the effect from computer self-efficacy to metacognitive

activity was constrained to be the same for older and younger participants, fit was not

significantly worse, χ2diff(1)=.08, p=.773. Therefore, H11 was not supported, and we can

conclude that the relationship between computer self-efficacy and metacognitive activity

is the same for younger and older individuals.

The final hypothesis for the fourth research question (H12) was: The relationship

between learner control self-efficacy and metacognitive activity varies by age.

Constraining the effect from learner control self-efficacy to metacognitive activity to be

the same for younger and older respondents did not significantly worsen the model fit,

χ2diff(1)=.00, p=.975. Thus, H12 was not supported, leading us to conclude that the

63

relationship between learner control self-efficacy and metacognitive activity is the same

for younger and older individuals.

Research Question 5

The fifth research question is based on the model shown in Figure 2. In this

model, computer self-efficacy is hypothesized to have a positive effect on learner control

self-efficacy, which subsequently has an effect on metacognitive activity. The hypothesis

to be tested (H13) was: Computer self-efficacy has a positive, indirect effect on

metacognitive activity through learner control self-efficacy. Figure 8 shows the results of

this model.

Figure 8. Path Model for Research Question 5 with Regression Coefficients

.49* .34*


The effect from computer self-efficacy to learner control self-efficacy was statistically

significant, β=.49, p<.001, and the effect from learner control self-efficacy to

metacognitive activity was statistically significant, β=.34, p<.001.

The indirect effect of computer self-efficacy on metacognitive activity through

learner control self-efficacy is equal to the product of the β coefficients, or .49 X .34 =

.17. The Sobol z test result for this indirect effect was statistically significant, Sobol

z=3.30, p=.001. This indicated that the hypothesis was supported, and we can conclude




R2=.24 R2=.12

64

that computer self-efficacy has an indirect effect on metacognitive activity through

learner control self-efficacy.

65

CHAPTER V: DISCUSSION AND IMPLICATIONS

This chapter presents a discussion of the findings of the study. Initially, a

summary of the findings is presented both for the preliminary analyses and those analyses

performed to address the research questions of this study. Then, the implications of these

findings are discussed. Recommendations for future research and for practice are

presented next and, finally, a brief set of conclusions from this study.

Summary of Findings

One-hundred and twenty individuals participated in this study, of which two-

thirds were male. Over three-quarters of the participants were White, and they averaged

just over 30 years of age. The reliability coefficients for the four composite measures

used in this study were high, ranging from .88 to .96. In the combined sample, computer

self-efficacy was positively correlated with both learner control self-efficacy and

metacognitive activity but not with supervisory support. Learner control self-efficacy was

also correlated with supervisory support and metacognitive activity. In addition,

supervisory support was positively correlated with metacognitive activity. When the

sample was divided into male and female groups, differences emerged. For females,

learner control self-efficacy and metacognitive activity were positively correlated, but no

other correlations were statistically significant. For males, however, computer self-

efficacy was positive correlated with all three other measures, and learner control self-

efficacy was positively correlated with supervisory support metacognitive activity. Thus,

the correlations among the four measures were generally larger and more likely to be

statistically significant for males than for females (although it should be noted that the

66

sample of males was larger, resulting in more statistical power for the male group than

the female group).

The correlations between the four composite measures were also examined as a

function of age (those 28 and younger versus those 29 or older). For the younger

respondents, computer self-efficacy was positively correlated with both learner control

self-efficacy and metacognitive activity, and learner control self-efficacy was positively

correlated with both supervisory support and metacognitive activity. For the older

respondents, computer self-efficacy was positively correlated with learner control self-

efficacy, but not with metacognitive activity, and learner control self-efficacy was

positively correlated with metacognitive activity, but not with supervisory support.

Supervisory support and metacognitive activity were not correlated for the older

respondents. Overall, the correlations among the measures tended to be stronger for the

younger respondents than for older respondents.

The first research question of the current study was: Is supervisory support

related to learner control self efficacy and computer self-efficacy in a learner controlled

training environment? Results indicated that (a) there was no relationship between

supervisory support and computer self-efficacy, and (b) there was a positive relationship

between supervisory support and learner control self-efficacy, with higher levels of

supervisory support were associated with higher levels of learner control self-efficacy.

The second research question was: Are computer self-efficacy or learner control


environment? Data related to this research question showed that: (a) there was no

relationship between computer self-efficacy and metacognitive activity, and that there

67

was a positive relationship between learner control self-efficacy and metacognitive

activity such that higher levels of learner control self-efficacy were associated with

higher levels of metacognitive activity.

The third research question was: Do the relationships between supervisory


gender of the learner? Results of the analyses performed indicated that (a) the

relationship between supervisory support and computer self-efficacy was the same for

males and females, (b) the relationship between supervisory support and learner control

self-efficacy was the same for males and females, (c) the relationship between computer

self-efficacy and metacognitive activity was the same for males and females, and (d) the

relationship between learner control self-efficacy and metacognitive activity was the

same for males and females. Therefore, the answer to the third research question is that

the relationships between supervisory support, learner control self-efficacy, and

metacognitive activity do not vary as a function of gender of the learner.

The fourth research question was: Do the relationships between supervisory


age of the learner? Results indicated that (a) the relationship between supervisory

support and computer self-efficacy was the same for younger and older individuals, (b)

the relationship between supervisory support and learner control self-efficacy was the

same for younger and older individuals, (c) the relationship between computer self-

efficacy and metacognitive activity was the same for younger and older individuals, and

that (d) the relationship between learner control self-efficacy and metacognitive activity

is the same for younger and older individuals. Therefore, the answer to the fourth

68

research question is that the relationships between supervisory support, learner control

self efficacy and metacognitive activity do not vary as a function of the age of the learner.

A fifth research question was designed to examine the indirect effect of computer

self-efficacy on metacognitive activity through learner control self-efficacy. Results

indicated that computer self-efficacy did in fact have a positive indirect effect on

metacognitive activity through learner control self-efficacy. Thus, higher levels of

computer self-efficacy resulted in higher levels of learner control self-efficacy, which in

turn resulted in higher levels of metacognitive activity.

Implications

This section presents an analysis and integration of the results of this study into

existing theoretical models and empirical studies related to metacognitive activity. First,

the contribution of the current study in terms of examining metacognitive activity in

adults as opposed to children is discussed. Gender and age group differences in computer

self-efficacy are described in the next section. The next section describes supervisory

support and its relationships with the self-efficacy variables employed in the current

study. Then, the relationships between self efficacy and metacognitive activity are

addressed.

Metacognitive Activity for Children Versus Adults

One of the key contributions of this study was the inclusion of adults rather than

children as the subject group. Prior studies of metacognitive activity, especially the

theoretical foundations identified by Flavell (1976, 1977, 1979, 1987), are based almost

entirely on research among pre-adult (primary and secondary school) populations. Thus,

the demonstration that self-efficacy and metacognitive activity were positively related to

69

each other in adults is a contribution to the literature on metacognitive activity. In

addition, the current study examined differences between younger adults and older adults

and found no differences (in the direct difference tests, although some indications of

differences were found in the correlational analyses). It may be the case that these

relationships are relatively consistent once an individual reaches adulthood.

Gender, Age, and Computer Self-Efficacy

Confidence in using a personal computer has been shown to relate negatively to

age, that is, older adults evince lower confidence in operating a personal computer than

younger adults and even children (Henderson et al., 1995; Comber et al., 1997). In

addition, gender has also been identified as a variable affecting computer self-efficacy

(Comber et al.; Ford et al., 2001), with females having lower levels of computer self-

efficacy in past research. The current study did not directly test differences in computer

self-efficacy between males and females or between younger and older adults, but did

examine differences in the relationships between computer self-efficacy and other

variables between males and female and between younger and older adults. These direct

tests performed to determine if gender or age moderated the relationships between

computer self-efficacy and the other variables in this study did not reveal any differences.

Thus, it may be the case that males and females or older and younger adults differ in

levels of computer self-efficacy (as past research has indicated) but that these mean

differences do not affect the relationships between computer self-efficacy and other

variables. Alternatively, the issue of computer self-efficacy may no longer be gender or

age related given the ubiquity of these devices in the contemporary business

environment.

70

Supervisory Support, Computer Self-Efficacy and Learner Control Self-Efficacy

No relationship was found between supervisory support and computer self-

efficacy in a learner controlled training environment in the current study. This is

contradictory to past studies presented in the literature that have found support for the

hypothesis that supervisor feedback affects motivation (Cohen, 1990; Huczynski &

Lewis, 1980; Latham & Locke, 1991). This inconsistency could be explained by the

multitude of variables that can affect one’s computer self-efficacy. These include: extent

of prior computer training, access and involvement with computers both in the work and

home environment, and the likely wide range of supervisory support for employee

computer skills in general.

Supervisory support was, however, related to learner control self-efficacy for the

combined sample, as well as for males and the younger adult group separately. This

finding of a positive relationship between supervisor support and learner control self-

efficacy suggests that higher levels of supervisor support are associated with higher levels

of learner control self efficacy. Statements and attitudes expressed by supervisors

regarding positive training outcomes appears to reinforce a, employee’s feeling of

confidence. Wood and Bandura (1989a), Griffin (1989), and Martocchio (1992) found

similar relationships as positive supervisory cues increase the degree of employee

motivation and training outcomes. The fact that the relationships between these variables

was stronger for males and for younger adults is worthy of exploration in future research

as will be discussed below.

71

Metacognitive Activity and Self-Efficacy

Past research has shown that the ability of the trainee to control time on task and

sequence of learning improves metacognitive activity (Schmidt & Ford, 2003). In the

current study, metacognitive activity was positively associated with both learner self-

efficacy and computer self-efficacy. This finding in the current study supports the

findings from studies such as those by Schmidt and Ford. That is, the finding that higher

levels of self-efficacy (either learner control or computer) were associated with higher

levels of metacognitive activity is consistent with the past finding that the ability of

trainees to control time on task and sequence of learning results in higher levels of

metacognitive activity. In addition to positive relationships with metacognitive activity,

learner control self-efficacy has been shown to relate positively to other training

outcomes (e.g., Martocchio, 1994; Tannenbaum & Yuki, 1993), indicating that it has

broad importance in the field of education.

However, the results of the current study included an examination of the potential

moderating influence of gender and age on this relationship, and the findings were not as

clear. Specifically, the correlational analyses revealed some potential differences in these

relationships based on gender and age (with the relationships between the self-efficacy

variables and metacognitive activity being stronger for males and younger participants

than for females and older individuals), but the direct moderation tests indicated that

there were no differences based on gender or age. This may be due to the moderate

sample size in the current study as the correlational differences between the gender and

age groups appear substantial. Past research involving gender and computer self-efficacy

(e.g., Brosnan, 1998; Henry & Stone, 1999) suggested a lower level of computer self-

72

efficacy for females. It may be the case that the very widespread availability of desktop

computers that makes them such an obvious choice for learner controlled training, has

rendered them “gender neutral” office tools like copiers, fax machines, and so forth, and

therefore that a possible gender difference has disappeared over time. In any case, if there

are moderating effects of gender or age on the relationships between metacognitive

activity and self-efficacy, this would suggest that for certain age or gender groups, the

person factors for metacognitive activity, as explained by Flavell (1987) may override the

method of task accomplishment.

Recommendations for Future Research

Based on the results of this study, there are several recommendations that can be

made for future research in the area of supervisory support, self-efficacy and

metacognitive activity. First, the current study was performed within a single learning

environment with a single set of training goals – leadership skills. As mentioned in the

methodology chapter, organizational trainees using the Virtual Leader program from

Simulearn Inc. were the focus of this study. Because a single learning environment was

included in the current study, it is possible that some of the relationships and effects

found (and not found) in the current study are specific to that learning environment.

Therefore, it is recommended that the current study be replicated with other learning

environments to determine if the effects are specific to the Simulearn/Virtual Leader

program or if they are more generalizable.

Second, the current study used a non-experimental research design, but other

designs should be attempted. In the current study, it was not possible to experimentally

manipulate variables such as supervisory support in order to examine their effects on

73

other variables. Therefore, firm causal conclusions regarding these relationships are not

possible. While some of the variables of interest in the current study are individual

characteristics (e.g., self-efficacy) and therefore are not easily manipulated in an

experimental situation, others (such as supervisory support) could be manipulated with

relative ease. Therefore, it is recommended that alternative study designs, such as true

experiments, be attempted to advance the results of the current study.

Third, while the current study included several of the key variables related to the

theoretical model of self-efficacy, supervisory support, and metacognitive activity,

including gender and age as potential moderators, there are many other variables that

could be included in subsequent models. For example, it would be interesting to

determine if an individual’s educational level would influence the relationships among

the study variables. When modeling any complex phenomenon, any given study’s list of

included variables will be incomplete, and it is thought that the variables included in the

current study were some of the most important given this subject area. However, future

researchers should attempt to build on the results of the current study by incorporating

additional variables to examine their place within the model.

Fourth, one of the findings that is somewhat contradictory in the current study

was that there appeared to be differences between males and females and between

younger and older adults when the correlations between the key study measures were

examined in the preliminary analyses, but that there were no differences between these

groups in the direct tests of moderation. As noted above, this may be due to the somewhat

small sample sizes in the current study when subgroups were created. Specifically, the

indication in the preliminary analyses that there were potential differences between males

74

and females and between older and younger adults were based on statistically significant

correlations in some sub-groups but non-statistically significant correlations in other

groups, despite the fact that the correlations were in the same direction (positive) for all

groups even if they were not statistically significant. Thus, future research should employ

larger samples sizes to remove any ambiguity in whether or not there are differences

between males and females or between older and younger adults in the relationships

between the variables in this study.

Finally, one recommendation for future research involves re-examining the

mediation of the effect of computer self-efficacy on metacognitive activity through

learner control self-efficacy. One of the findings of the current study was that computer

self-efficacy had an indirect effect on metacognitive activity through learner control self-

efficacy (i.e. that higher levels of computer self-efficacy were associated with higher

levels of learner control self-efficacy, which in turn were associated with higher levels of

metacognitive activity). Although this finding is consistent with causality, given the

correlational nature of the current study, firm causal conclusions cannot be drawn. It

should be noted that the sample consisted of graduate school students whose computer

self-efficacy can be reasonably assumed to be higher than the general population.

Therefore, it is recommended that future researchers examine this relationship in a true

experiment. Specifically, the effect of a manipulation of computer self-efficacy (through

computer training) on learner control self-efficacy, and the effect of a manipulation of

learner-control self-efficacy on metacognitive activity should be examined in a true

experiment (involving random assignment of participants to the experimental and control

groups).

75

Recommendations for Practice

Given the findings of the current study, there are several recommendations that

can be made for business and training practice. First, the results from the current study

indicated the relationships between supervisory support, computer self-efficacy, learner

control self-efficacy, and metacognitive activity did not differ for males and females.

This indicates that when educational programs such as the one employed in the current

study are designed, one approach should work well for both males and females, and

therefore there is no need to design specific programs for each gender.

Second, the results from this study showed that there were no differences in the

relationships between supervisory support, computer self-efficacy, learner control self-

efficacy, and metacognitive activity for older versus younger participants. As was the

case with gender, this indicates that the age of the learner does not play a key role in

programs such as the one examined in the current study—older learners and younger

learners do not have to be differentiated when designing such programs.

Third, the results of the current study indicated that supervisory support was

related to learner control self-efficacy (such that higher levels of supervisory support

were associated with higher levels of learner control self-efficacy) but not to computer

self-efficacy. Thus, while the level of supervisory support provided is likely to make a

learner have stronger feelings of control, it is unlikely to assist in a learner’s confidence

in their computer activities. Therefore, it is recommended that when computer self-

efficacy is a key component to a learning program, efforts beyond additional supervisory

support will be required.

76

Fourth, the results of the current study showed that metacognitive activity could

be predicted by learner control self-efficacy (with higher levels of learner control self-

efficacy associated with higher levels of metacognitive activity) but not by computer self-

efficacy. Thus, learner control self-efficacy appears to be more important than computer

self-efficacy in terms of the possibility of increasing metacognitive activity. Therefore, it

is recommended that in learning situations where metacognitive activity is important, the

focus should be on increasing learner control self-efficacy rather than on computer self-

efficacy (but also see the next recommendation).

Fifth, one of the most interesting findings of the current study was that computer

self-efficacy had an indirect effect on metacognitive activity through learner control self-

efficacy. Specifically, higher levels of computer self-efficacy were associated with higher

levels of learner control self-efficacy, which in turn were associated with higher levels of

metacognitive activity. While the current study was correlational rather than a true

experiment, this finding is consistent with the hypothesis that manipulating computer

self-efficacy could have a downstream effect, increasing learner control self-efficacy and

consequently increasing metacognitive activity. Thus, for educational programs such as

the one in the current study, it is recommended that program administrators begin with

training focused on improving the learners’ computer skills given the effects that it is

likely to have on metacognitive activity (through learner control self-efficacy). It is

important to note that learner control self-efficacy was also directly related to

metacognitive activity, and therefore that efforts to increase learner control self-efficacy

are also likely to be worthwhile.

77

A final recommendation for practitioners relates to the size of the effects found in

the current study. Three variables served as dependent variables (predicted by other

variables) in this study: computer self-efficacy, learner control self-efficacy, and

metacognitive activity. For computer self-efficacy, the R2 values ranged from .00 to .05

across all models. For learner control self-efficacy, the R2 values ranged from .03 to .24.

For metacognitive activity, the primary dependent variable in this study, the R2 values

ranged from .10 to .23. Thus, the relatively small R2 coefficients in the current study are

important for practitioners to consider because they imply that the predictive models in

this study explain relatively small percentages of variances in the dependent variables.

There are clearly a variety of other potential predictor variables that should be considered

in attempts to understand metacognitive activity, computer self-efficacy, and learner

control self-efficacy.

Conclusions

The purpose of the current study was to examine the effects of supervisory

support, age and gender on an adult learner’s perception of self-efficacy and

metacognitive activity when metacognitive interventions are utilized in a learner

controlled training environment. The key results of this study were:

1. There was no relationship between supervisory support and computer self-

efficacy.

2. There was a positive relationship between supervisory support and learner control

self-efficacy, with higher levels of supervisory support were associated with

higher levels of learner control self-efficacy.

78

3. There was no relationship between computer self-efficacy and metacognitive

activity.

4. There was a positive relationship between learner control self-efficacy and

metacognitive activity such that higher levels of learner control self-efficacy were

associated with higher levels of metacognitive activity.

5. The relationships, or lack thereof, among the key study variables did not vary as a

function of the gender or age of the participant.

Based on these results, it was recommended that practitioners

1. Do not need to consider the age or gender of the learner in learning programs

similar to the one employed in the current study.

2. Should not rely only on supervisory support to enhance computer self-efficacy.

3. Should focus on enhancing learner control self-efficacy when attempting to

increase metacognitive activity.

4. Could also focus on computer self-efficacy as a means of increasing

metacognitive activity due to the indirect effect of computer self-efficacy on

metacognitive activity through learner control self-efficacy.

Finally, it was recommended that future researchers:

1. Replicate the current results within other learning environments to determine if

the effects are specific to the Simulearn/Virtual Leader program or if they are

more generalizable

2. Employ additional research designs such as performing true experiments

3. Include additional variables in the statistical models.

79

References

American Association of University Women, Educational Foundation Commission on Technology Gender and Teacher Education. (2000). Tech-savvy: Educating girls in the new computer Age. Washington, D.C.: American Association of University Women Educational Foundation. Retrieved Oct. 30, 2004 from: http://www.aauw. org/member_ center/publications/TechSavvy/TechSavvy.pdf

Aleven, V., Stahl, E., Schworm, S., Fischer, F. & Wallace, R. (2003). Help seeking and

help design in interactive learning environments. Review of Educational Research, 73(3), 277-306.

Armstrong, A. M. (1989). Persistence and the causal perception of failure: Modifying

cognitive attributions. Journal of Educational Psychology, 70(2), 154-166. American Society for Training and Development (2004). State of the industry report.

Retrieved July 5, 2004 from: http://www.astd.org/NR/rdonlyres/49A44CD8-2A3E-471F-B80E-42504F4A1726/0/SOIR_2004_Executive_Summary.pdf

American Society for Training and Development (2008), Industry statistics. Retrieved

January 27, 2008 from: http://www.astd.org/industrystatistics Atkinson, J. W. (1964). An introduction to motivation. New Jersey: Van Nostrand. Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change.

Psychological Review, 84, 191-215. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.

New Jersey: Prentice-Hall. Bandura, A. (1988). Organisational applications of social cognitive theory. Australian

Journal of Management 13(2), 275-301. Bandura, A., & Locke, E. A. (2003). Negative self-efficacy and goals effects revisited.

Journal of Applied Psychology, 88(1), 87-99. Bandura, A., & Wood, R. (1989). Effect of perceived controllability and performance

standards on self-regulation of complex decision making. Journal of Personality and Social Psychology, 56 (5), 805-814.

Bell, B. S., & Kozlowski, S.W. (2002). Adaptive guidance: Enhancing self-regulation,

knowledge and performance in technology based training. Personnel Psychology, 55(2), 267-307.

Birdi, K., Allan, C., & Warr, P. (1997). Correlates and perceived outcomes of four types

of employee development activity. Journal of Applied Psychology 82(6), 845-857.

80

Brosnan, M. J. (1998). The impact of psychological gender, gender-related perceptions,

significant others, and the introducer of technology upon computer anxiety in students. Journal of Educational Computing Research, 18(1), 63-78.

Brown, K.G. (2001). Using computers to deliver training: Which employees learn and

why? Personnel Psychology, 54(2), 271-297. Boekaerts, M., (1995). Self regulated learning: bridging the gap between metacognitive

and metamotivational theories. Educational Psychologist, 30(4), 195-200. Bruner, J. S., Goodnow, J. J., & Austin, G. A. (1956). A study of thinking. New York:

Wiley. Choi, N. (2004). Sex role differences in specific, academic and general self efficacy. The

Journal of Psychology, 138(2), 149-160. Cohen, D. J. (1990). What motivates trainees? Training and Development Journal,

44(11), 91-94. Collins, C. J., & Clark, K. D. Strategic human resource practices, top management team

social networks, and firm performance: the role of human resource practices in creating organizational competitive advantage. Academy of Management Journal, 46(6), 740-752.

Colquitt, J. A., LePine, J. A., & Noe, R. A. (2000). Toward an integrative theory of

training motivation: A meta-analytic path analysis of 20 years of research. Journal of Applied Psychology, 85(5), 678-707.

Comber, C., Colley, A., Hargreaves, D. J., & Dorn, L. (1997). The effects of age, gender

and computer experience upon computer attitudes. Educational Research, 39, 123-133.

Computer Self-Efficacy Survey (n.d.). Retrieved Oct. 19, 2005, from http://www.

positivepractices.com/EducationalTechnology/ComputerSelf-Efficacy Surv.html#ValidationoftheComputerSelf-EfficacyScaleCSE

Constructivist Learning Theory (n.d.). Retrieved Aug. 26, 2004, from http://www.cdli.ca

~elmurphy/emurphy/cle2b.html. Czaja, S. J. & Sharit, J., (1998). Age differences in attitudes toward computers. The

Journals of Gerontology: Series B: Psychological Sciences and Social Sciences, 53B(5), 329-341.

Despite Economy, companies spent more on training in ’02. HR Focus, 81(2), 8-10.

81

DeMar, G (2004). Behaviorism. Retrieved August 19, 2004 from http://psych.athabascau. ca/html/Behaviorism/

Eom, W., & Reiser, R. A., (2000). The effects of self-regulation and instructional control

on performance and motivation in computer-based instruction. International Journal of Instructional Media, 27(3), 247-261.

Facteau, J. D., Dobbins, G. H., Russell, J., Ladd, R. T., & Kudisch, J. D. (1995). The

influence of general perceptions of the training environment on pretraining motivation and perceived training transfer. Journal of Management, 21(1), 1-25.

Fishbein, M. (1967). Readings in attitude theory and measurement. New York: Wiley. Flavell, J. H. (1976). Metacognitive aspects of problem solving. In L. B. Resnick (Ed.),

The nature of intelligence (pp. 231-236). New Jersey: Lawrence Erlbaum Associates.

Flavell, J. H. (1977). Cognitive development. New Jersey: Prentice-Hall. Flavell, J. H. (1979). Metacognition and cognitive monitoring: a new area of cognitive-

developmental inquiry. American Psychologist, 34, 906-911. Flavell, J. H. (1987). Speculations about the nature and development of Metacognition. In

F. Weinert & R. Kluwe (Eds.), Metacognition, motivation and understanding (pp. 21-30). New Jersey: Lawrence Erlbaum Associates

Fleishman, E. A., (1957). A leader behavior description for industry. In R. M. Stogdill

and A. E. Coons (Eds.), Leader Behavior: Its Description and Measurement. Columbus: Ohio State University.

Ford, J. K., Smith, E. M., Weissbein, D. A., Gully, S. M., & Salas, E. (1998).

Relationships of goal orientation, metacognitive activity and practice strategies with learning outcomes and transfer. Journal of Applied Psychology, 83(2), 218-233.

Ford, N., Miller, D., & Moss, N. (2001). The role of individual differences in internet

searching: An empirical study. Journal of the American Society for Information Science and Technology, 52(12), 1049-1066.

Gagne´, R. M., & Briggs, L. J. (1974). Principles of instructional design. New York:

Holt, Rinehart and Winston. Gagne´, R. M. (1977). The conditions of learning. New York: Holt, Reinhart and

Winston.

82

Glaser, R., & Pellegrino, J.W. (1987). Aptitudes for learning and cognitive processes. In F. Weinert & R. Kluwe (Eds.), Metacognition, motivation, and understanding (pp. 267-288). New Jersey: Lawrence Erlbaum Associates.

Gilley, J. (2002). Gender and technology awareness training in preservice teacher

education. Tech Trends, 46(6), 21-27. Gist, M. E. (1987). Self-efficacy: Implications for organizational behavior and human

resource management. Academy of Management Review, 12(3), 472-485. Gist, M. E., & Mitchell, T. R. (1992). Self-efficacy: A theoretical analysis of its

determinants and malleability. Academy of Management Review 17(2), 183-211. Gist, M. E., Rosen, B., & Schwoerer. (1988). The influence of training method and

trainee age on the acquisition of computer skills. Personnel Psychology, 41, 255-265.

Gist, M. E., Schwoerer, C., & Rosen, B. (1989). Effects of alternative training methods

on self-efficacy and performance in computer software training. Journal of Applied Psychology, 74(6), 884-891.

Gist, M. E., Stevens, C. K., & Bavetta, A. G. (1991). Effects of self efficacy and post-

training intervention on the acquisition and maintenance of complex interpersonal skills. Personnel Psychology, 44(4), 837-861.

Griffin, R. W. (1983). Objective and social sources of information in task redesign: A

field experiment. Administrative Science Quarterly, 28(2), 184-201. Hacker, D. J., (2003). Metacognition: Definitions and empirical foundations. Retrieved

Nov. 18, 2003 from http://www.psyc.memphis.edu/trg/meta.htm Halpin, A. W. (1957). The observed leader behavior and the ideal leader behavior of

aircraft commanders and school superintendents. In R. M. Stogdill and A. E. Coons (Eds.), Leader behavior: Its description and measurement. Columbus: Ohio State University.

Halpin, A. W., & Winer, B. J. (1957). A factorial study of the leader behavior

descriptions. . In R. M. Stogdill and A. E. Coons (Eds.), Leader behavior: Its description and measurement. Columbus: Ohio State University.

Hannafin, R. D., & Sullivan, H. J. (1996). Preferences and learner control over amount of

instruction. Journal of Educational Psychology, 88(1), 162-173. Henderson, R., Deane, F., Barrelle, K., & Mahar, D. (1995). Computer anxiety:

correlates, norms and problem definition in health care and banking employees using the Computer Attitude Scale. Interacting with Computers, 7(2), 181-193.

83

Henry, J. W., & Stone, R. W. (1999). The impacts of end user gender, education,

performance, and system use on computer self efficacy and outcome expectancy. Southern Business Review, 25(1), 10-17.

Hilgard, E.R., & Bower, G.H. (1966). Theories of learning. New York: Meredith

Publishing. Hill, J. R., & Hannafin. M. J. (1997). Cognitive strategies and learning from the World

Wide Web. Educational Technology, Research and development, 45(4), 37-65. Huczynski, A.A., & Lewis, J. W. (1980). An empirical study into the learning transfer

process in management training. Journal of Management Studies 17(2), 227-241. Illeris, K. (2003). Workplace learning and learning theory. Journal of Workplace

Learning, 15(4), 167-179. Jacobs, B., Prentice-Dunn, S., & Rogers, R. W. (1984). Understanding persistence: An

interface of control theory and self-efficacy theory. Basic and Applied Social Psychology, 5, 333-347.

Jegede, O., Taplin, M., Fan, R., Chan, M., & Yum, J. (1999). Differences between low

and high achieving distance learners in locus of control and metacognition. Distance Education, 20(2), 255-274.

Jeongkoo, Y. & Shane, T. (2000). Supervisor support in the workplace: Legitimacy and

positive affectivity. The Journal of Social Psychology, 140(3), 295-317. Jiang, J. J., & Klein, G. (2000). Supervisor support and career anchor impact on the

career satisfaction of the entry-level information systems professional. Journal of Management Information Systems, 16(3), 219-230.

Kinzie, M. B. (1990). Requirements and benefits of effective interactive instruction:

Learner control, self regulation, and continuing motivation. Educational Technology Research and Development, 38(1), 1-21.

Kinzie, M. B., & Sullivan, H. J. (1989). Continuing motivation, learner control, and CAI.

Educational Technology Research and Development, 37, 5-14. Kinzie, M. B., Sullivan, H. J., & Berdel, R. L. (1988). Learner control and achievement

in science computer-assisted instruction. Journal of Educational Psychology, 80 (3), 299-308.

Kraiger, K., Ford, J. K., & Salas, E. (1993). Application of cognitive, skill-based, and

affective theories of learning outcomes to new methods of training evaluation [Monograph]. Journal of Applied psychology, 78(2), 311-328.

84

Latham, G. P., & Locke, E. A. (1991). Self regulation through goal setting.

Organizational Behavior and Human Decision Processes, 50(2), 212-248. Likert, R. (1961). New patterns of management. New York: McGraw-Hill. Livingston, J. A. (1997). Metacognition: An overview. Retrieved Nov. 18, 2003 from

http://www.gse.buffalo.edu/fas/shuell/cep564/Metacog.htm. Lunts, E. (2002). What does the literature say about the effectiveness of learner control in

computer-assisted instruction? Electronic Journal for the Integration of Technology in Education, 1(2), 59-75.

Mahoney, M.J., (n.d.). What is constructivism and why is it growing? Retrieved August

29, 2004 from: http://www.constructivism123.com/What_Is/What_is_ constructivism.htm

Martocchio, J. J. (1992). Microcomputer usage as an opportunity: The influence of

context in employee training. Personnel Psychology 45(3), 529-553. Martocchio, J. J. (1994). Effects of conceptions of ability on anxiety, self-efficacy, and

learning in training. Journal of Applied Psychology, 79(6), 819-825. Mathieu, J. E., Tannenbaum, S. I., & Salas, E. (1992). Influences of individual and

situational characteristics on measures of training efficacy. Academy of Management Journal, 35(4), 828-847.

Maurer, T. J., Weiss, E. M., & Barbeite, F. G. (2003). A model of involvement in work-

related learning and development activity: The effects of individual, situational, motivational, and age variables. Journal of Applied Psychology, 88(4), 707-724.

Mevarech, Z. R. (1999). Effects of metacognitive training embedded in cooperative

settings on mathematical problem solving. Journal of Educational Research, 94 (4), 195-205.

Mevarech, Z. R., & Kramarski, B. (2003). The effects of metacognitive training versus

worked-out examples on students’ mathematical reasoning. British Journal of Educational Psychology, 73, 449-471.

Milheim, W. D., & Martin, B. L. (1991). Theoretical bases for the use of learner control:

Three different perspectives. Journal of Computer Based Instruction, 18(3), 99-105.

85

Morrison, G. R., Ross, S. M., & Baldwin, W. (1992). Learner control of context and instructional support in learning elementary school mathematics. Educational Technology Research and Development, 40(1), 5-13.

Naik, P. (2004). Behaviorism as a theory of personality: A critical look. Retrieved

Aug.18, 2004 from: http://www.personalityresearch.org/papers/naik.html Newton, D., Hase, S., & Ellis, A. (2002). Effective implementation of online learning: a

case study of the Queensland mining industry. Journal of Workplace Learning, 14(4), 156-166.

Pajares, F. (2002). Gender and perceived self efficacy in self regulated learning. Theory

into Practice, 41(2), 116-128. Paris, S. G., & Winograd, P (1990). How Metacognition can promote academic learning

and instruction. In B. F. Jones & L. Idol (Eds.), Dimensions of thinking and cognitive instruction (pp. 185-218). Hillsdale, NJ: Erlbaum.

Perez, G. C., Kester, L., & Van Merrienboer, J. G. J. (n.d.). Adaptation of education and

learner control: A model for personalized task selection. Retrieved August 31, 2004 from: http://www.iwm-kmrc.de/workshops/sim2004/pdf_files/ Corbalan_et_al.pdf.

Pollack, J. C., & Sullivan, H. J. (1990). Practice mode and learner control in computer-

based instruction. Contemporary Educational Psychology, 15, 251-260. Qutami, Y., Abu-Jaber, M., (1997). Students’ self efficacy in computer skills as a

function of gender and cognitive learning style at Sultan Qaboos University. International Journal of Instructional Media, 24(1), 63-75.

RAND Corporation (2004). The future at work: Trends and implications. Retrieved

January 22, 2008 from: www.rand.org/pubs/research_briefs/rb5070/index1.html Relan, A. (1995). Promoting better choices: Effects of strategy training on achievement

and choice behavior in learner-controlled CBI. Journal of Educational Computing Research 13(2), 129-149.

Rhoades, L., & Eisenberger, R. (2002). Perceived organizational support: A review of the

literature. Journal of Applied Psychology, 87(4), 698-714. Rodriquez, R. (2005). Meet the new learning executive. HR Magazine, 50(4), 64-70. Russell, C. G. & Bobko, P. (1992). Moderated Regression Analysis and Likert Scales,

Too Coarse for Comfort. Journal of Applied Psychology, 77(3), 336-342.

86

Schacter, J. (1999). The impact of educational technology on student achievement: What the most current research has to say. Milken exchange on Educational Technology, Santa Monica, CA. (ERIC Document Reproduction Service No. ED 430 537).

Schmidt, A. M., & Ford, K. J. (2003). Learning within a learner control training

environment: The interactive effects of goal orientation and metacognitive instruction on learning outcomes. Personnel Psychology, 56(2), 405-420.

Schroeder, E. (1994, August 29). Training takes off, using multimedia. PC Week, p.33. Semple, A. (2000). Learning theories and their influence on the development and use of

educational technologies. Australian Science Teachers Journal, 46(3), 21-28. Shore, L. M., & Shore, T. H. (1995). Perceived organizational support and organizational

justice. In R. S. Cropanzano & K. M. Kacmar (Eds.), Organizational politics, justice, and support: Managing the social climate of the workplace (pp. 149-164). Westport, Ct: Quorum.

Skinner, B. F. (1968). The technology of teaching. New York: Prentice Hall. Skinner, B. F. (1969). Contingencies of reinforcement: A theoretical analysis. New York:

Prentice Hall. Sobol, M.E. (1982). Asymptotic intervals for indirect effects in structural equation

models. In S. Leinhart (Ed.), Sociological Methodology 1982 (pp. 290-312). San Francisco: Jossey-Bass.

SPSS (2004). Frequently asked questions. Retrieved Oct. 1, 2004 from: http://www.ats.

ucla.edu/stat/spss/faq/ alpha.html Stajkovic, A. D., & Luthans, F. (1997). A meta-analysis of the effects of organizational

behavior modification on task performance, 1975-95. Academy of Management Journal, 40 (5), 1122-1149.

Stinglhamber, F., & Vandenberghe, C. (2003). Organizations and supervisors as sources

of support and targets of commitment: a longitudinal study. Journal of Organizational Behavior, 24, 251-270.

Tannenbaum, S. I., & Yuki, G. (1992). Training and development in work organizations.

Annual Review of Psychology, 43, 339-441. Tharenou, P. (2001). The relationship of training motivation to participation in training

and development. Journal of Occupational & Organizational Psychology 74(5), 599-622.

87

The New Economy Index (n.d.). Corporate expenditures on training have slightly declined. Retrieved August 31, 2004 from: http://www.neweconomyindex .org/section3_page18.html

Vince, R. (1998). Behind and beyond Kolb’s learning cycle. Journal of Management

Education, 22(3), 304-319. Vroom, V. H. (1964). Work and motivation. New York: Wiley. Warr, P., & Bunce, D. (1995). Trainee characteristics and the outcomes of open learning.

Personnel Psychology, 48(2), 347-378. Watson, J. B. (n.d.). The effect of metacognitive cues and probes on use of learner

control features in an on-line lesson for elementary students. Retrieved September 1, 2004 from: http://www.pdkintl.org/edres/ddwind2.htm.

Weinert, F. E., & Kluwe, R. H. (Eds.) (1987). Metacognition, motivation, and

understanding. New Jersey: Lawrence Erlbaum Associates. Wood, R., & Bandura, A. (1989a). Social cognitive theory of organizational

management. Academy of Management Review, 14(3), 361-384. Wood, R., & Bandura, A. (1989b). Impact of conceptions of ability on self-regulatory

mechanisms and complex decision making. Journal of Personality and Social Psychology, 56(3), 407-415.

Zimmerman, B. J., & Martinez-Pons, M., (1988). Construct validation of a strategy

model of student self regulated learning. Journal of Educational Psychology, 80 (3), 284-290.

88

Appendix A: Demographic Survey

You are asked to participate in a research study conducted by James V. Polizzi, a Ph.D.

candidate at Touro University International. You were selected as a possible participant

in this study because you are utilizing the Virtual Leader© leadership training computer

program.

Your participation in this survey is voluntary and completely anonymous.

Participation in the survey is simple. In the envelope provided, you will find two short

questionnaires. Please complete PART A before you begin the Virtual Leader ©

program.

When you have completed the Virtual Leader © program, please complete PART B.

Please return all materials to the coordinator.

First, three questions for classification purposes.

Gender: Female Male

What is your age_______________?

Are you:

White ___ Black ___ Hispanic, Latino, of Spanish Origin ___ American Indian, Eskimo, Aleut ___ Asian or Pacific Islander ___ Other ___ Don't know ___ Refused ___ Now, please complete PART A. Thank you for your participation and good luck.

89

Appendix B: Learner Control Self Efficacy Scale

1. I can meet the challenges of this training

Strongly disagree Neither agree nor disagree Strongly agree 1 2 3 4 5

2. I am confident in my understanding of how information cues are related to learning


3. I can deal with training under ambiguous conditions


4. I am certain that I can manage the requirements of this training


5. I believe I will fare well in this training as the complexity increases


6. I am confident I can cope with learning in the absence of an

instructor


7. I believe I can develop methods to handle changing aspects of this

training

90


91

Appendix C: Computer Self-Efficacy Scale

1. I feel confident working on a personal computer

Very Little Confidence Some Confidence Quite a Lot of Confidence

1 2 3 4 5

2. I feel confident getting software up and running


1 2 3 4 5

3. I feel confident logging onto the Internet


1 2 3 4 5

4. I feel confident accessing information on the Internet


1 2 3 4 5

5. I feel confident using the User’s guide when help is needed


1 2 3 4 5

6. I feel confident entering or saving data (numbers or words) into a file


1 2 3 4 5

7. I feel confident escaping/exiting from the program/software


1 2 3 4 5

8. I feel confident logging off the mainframe computer system


1 2 3 4 5

9. I feel confident calling up a data file to view on the monitor screen


1 2 3 4 5

92

10. I feel confident understanding terms/words relating to computer hardware


1 2 3 4 5

11. I feel confident understanding terms/words related to computer software


1 2 3 4 5

12. I feel confident handling a floppy disk correctly


1 2 3 4 5

13. I feel confident learning to use a variety of programs (software)


1 2 3 4 5

14. I feel confident learning advanced skills within a specific program

(software)


1 2 3 4 5

15. I feel confident making selections from an on-screen menu


1 2 3 4 5

16. I feel confident using computers to analyze number data


1 2 3 4 5

17. I feel confident using a printer to make a “hard copy” of my work


1 2 3 4 5

18. I feel confident copying a disk


1 2 3 4 5

93

19. I feel confident copying an individual file


1 2 3 4 5

20. I feel confident copying and deleting information from a data file.


1 2 3 4 5

21. I feel confident moving the cursor around the monitor screen


1 2 3 4 5

22. I feel confident writing simple programs for the computer


1 2 3 4 5

23. I feel confident using the computer to write a letter or essay


1 2 3 4 5

24. I feel confident describing the function of computer hardware (keyboard, monitor, disk drives, and computer processing unit) Very Little Confidence Some Confidence Quite a Lot of Confidence

1 2 3 4 5

25. I feel confident understanding the three stages of data processing: input, processing, and output Very Little Confidence Some Confidence Quite a Lot of Confidence

1 2 3 4 5

26. I feel confident getting help for problems in the computer system Very Little Confidence Some Confidence Quite a Lot of Confidence

1 2 3 4 5

27. I feel confident storing software correctly


1 2 3 4 5

94

28. I feel confident explaining why a program (software) will or will not run correctly on a given computer Very Little Confidence Some Confidence Quite a Lot of Confidence

1 2 3 4 5

29. I feel confident using a computer to organize information


1 2 3 4 5

30. I feel confident getting rid of files when they are no longer needed


1 2 3 4 5

31. I feel confident organizing and managing files


1 2 3 4 5

32. I feel confident troubleshooting computer problems


1 2 3 4 5

95

Appendix D: Supervisory Support Scale

1. My supervisor takes time to learn about my career goals and

aspirations


2. My supervisor cares about whether or not I achieve my career

goals


3. My supervisor keeps me informed about training opportunities in

the organization


4. My supervisor is supportive of my efforts to acquire new

knowledge and skills


96

Appendix E: Metacognitive Activity Scale

1. I revised my tactics for learning the material to deal with the different complexity levels of the material


2. While engaged in the training, I monitored how well I was learning the material


3. As I practiced the material, I evaluated how well I was learning the material Strongly disagree Neither agree nor disagree Strongly agree 1 2 3 4 5

4. When my methods were not successful, I experimented with different procedures for learning the material Strongly disagree Neither agree nor disagree Strongly agree 1 2 3 4 5

5. As I practiced applying my learning, I changed how I approached learning the material Strongly disagree Neither agree nor disagree Strongly agree 1 2 3 4 5

6. I tried to monitor closely the areas where I needed the most practice Strongly disagree Neither agree nor disagree Strongly agree 1 2 3 4 5

97

7. I noticed where I made mistakes during practice and focused on improving those areas Strongly disagree Neither agree nor disagree Strongly agree 1 2 3 4 5

8. I used my performance on the previous section to revise how I would approach learning the next section Strongly disagree Neither agree nor disagree Strongly agree 1 2 3 4 5

98

Appendix F: Full Regression Results for Path Models

Table A5.1

Regression Results for Research Questions 1 and 2

B SEB β t p Supervisory Support as a Predictor of Computer Self-Efficacy (R2 = .02, Adjusted R2 = .01, F(1, 118) = 2.11, p = .149) Constant 122.18 5.95 20.52 < .001 Supervisory Support .62 .43 .13 1.45 .149 Supervisory Support as a Predictor of Learner Control Self-Efficacy (R2 = .05, Adjusted R2 = .04, F(1, 118) = 6.06, p = .015) Constant 22.69 1.43 15.89 < .001 Supervisory Support .25 .10 .22 2.46 .015 Computer Self-Efficacy and Learner Control Self-Efficacy as Predictors of Metacognitive Activity (R2 = .13, Adjusted R2 = .11, F(1, 118) = 8.66, p < .001) Constant 21.23 3.07 6.92 < .001 Computer Self-Efficacy .03 .02 .13 1.31 .194 Learner Control Self-Efficacy .29 .10 .28 2.82 .006

99

Table A5.2

Regression Results for Research Question 3 for Males

B SEB β t p Supervisory Support as a Predictor of Computer Self-Efficacy (R2 = .05, Adjusted R2 = .04, F(1, 78) = 4.38, p = .040) Constant 119.02 6.60 18.02 < .001 Supervisory Support 1.03 .49 .23 2.09 .040 Supervisory Support as a Predictor of Learner Control Self-Efficacy (R2 = .05, Adjusted R2 = .04, F(1, 78) = 5.19, p = .044) Constant 22.73 1.57 14.44 < .001 Supervisory Support .24 .12 .22 2.05 .044 Computer Self-Efficacy and Learner Control Self-Efficacy as Predictors of Metacognitive Activity (R2 = .10, Adjusted R2 = .08, F(1, 78) = 4.44, p = .015) Constant 21.85 3.84 5.69 < .001 Computer Self-Efficacy .06 .04 .24 1.59 .116 Learner Control Self-Efficacy .12 .16 .12 .77 .442

100

Table A5.3

Regression Results for Research Question 3 for Females

B SEB β t p Supervisory Support as a Predictor of Computer Self-Efficacy (R2 = .00, Adjusted R2 = .00, F(1, 38) = .05, p = .829) Constant 124.25 12.82 9.69 < .001 Supervisory Support .19 .87 .04 .22 .829 Supervisory Support as a Predictor of Learner Control Self-Efficacy (R2 = .03, Adjusted R2 = .01, F(1, 38) = 1.26, p = .269) Constant 23.02 3.27 7.04 < .001 Supervisory Support .25 .22 .18 1.12 .269 Computer Self-Efficacy and Learner Control Self-Efficacy as Predictors of Metacognitive Activity (R2 = .23, Adjusted R2 = .20, F(2, 37) = 5.82, p = .006) Constant 19.33 5.17 3.74 <.001 Computer Self-Efficacy .01 .04 .05 .32 .748 Learner Control Self-Efficacy .45 .14 .48 3.23 .003

101

Table A5.4

Regression Results for Research Question 4 for Younger Individuals

B SEB Β t p Supervisory Support as a Predictor of Computer Self-Efficacy (R2 = .01, Adjusted R2 = .00, F(1, 60) = .47, p = .494) Constant 127.80 8.30 15.39 < .001 Supervisory Support .41 .60 .09 .69 .494 Supervisory Support as a Predictor of Learner Control Self-Efficacy (R2 = .09, Adjusted R2 = .07, F(1, 60) = 5.80, p = .019) Constant 22.02 1.94 11.34 < .001 Supervisory Support .34 .14 .30 2.41 .019 Computer Self-Efficacy and Learner Control Self-Efficacy as Predictors of Metacognitive Activity (R2 = .10, Adjusted R2 = .08, F(2, 59) = 4.87, p = .011) Constant 21.03 4.14 5.08 < .001 Computer Self-Efficacy .03 .04 .12 .70 .489 Learner Control Self-Efficacy .30 .17 .29 1.74 .043

102

Table A5.5

Regression Results for Research Question 4 for Older Respondents

B SEB β t p Supervisory Support as a Predictor of Computer Self-Efficacy (R2 = .04, Adjusted R2 = .02, F(1, 56) = 2.14, p = .149) Constant 115.73 8.39 13.80 < .001 Supervisory Support .88 .60 .19 1.46 .149 Supervisory Support as a Predictor of Learner Control Self-Efficacy (R2 = .02, Adjusted R2 = .01, F(1, 56) = 1.37, p = .248) Constant 23.24 2.09 11.10 < .001 Supervisory Support .17 .15 .15 1.17 .248 Computer Self-Efficacy and Learner Control Self-Efficacy as Predictors of Metacognitive Activity (R2 = .12, Adjusted R2 = .10, F(2, 55) = 4.42, p = .017) Constant 19.78 4.83 4.10 < .001 Computer Self-Efficacy .04 .03 .17 1.31 .197 Learner Control Self-Efficacy .31 .14 .29 2.24 .029

103

Table A5.6

Regression Results for Research Question 5

B SEB β t p Computer Self-Efficacy as a Predictor of Learner Control Self-Efficacy (R2 = .24, Adjusted R2 = .23, F(1, 118) = 36.51, p < .001) Constant 10.62 2.58 4.12 < .001 Computer Self-Efficacy .12 .02 .49 6.04 < .001 Learner Control Self-Efficacy as a Predictor of Metacognitive Activity (R2 = .12, Adjusted R2 = .11, F(1, 118) = 15.52, p < .001) Constant 23.80 2.37 10.04 < .001 Learner Control Self-Efficacy .35 .09 .34 3.94 < .001

Final 2008

Documents