 

:=.7..¢..;v > w

 

. hr...
{“113 .
.. a .. .V ,
in...b.zls\m3xlvf.a. .In. 4.1.. “(0’)... I .
I U! .‘lLvl’VUo ‘5‘.
A 3.. .5 45Jv.lh\.|\
..I ' ’5?

‘dnv’
, HQLMHRHIEBMW .1
.. . .hhnbﬁlia .MﬂQHg} a; d
hiilrﬁhxg $54.9}. ; .11.. I}...
unavut‘cﬁinwnunici. ““314..- c. I .n rxunm...
.. .. \v.ﬁ.nhuu$h!»vn.azhﬂa. 91$.
3. . 91%.... t]... u .

r . V-\2.Muns.|~ '.
,‘ii...ll|.\c)t1. ..

v on «0"

an- ﬁtnh )..UNnH

.10'1 .

vtl\l).nu.r

, .uv‘7.:vfi|s.3l9
. .3111ir. ..... .
399.1-.. if! V
.. ...\....... ‘ .P. $651.3. 1

.. .. .v'L-Inlt

 

i
.5301. .
.. : . - e « _. .
xnzKunum: rt, . . .
V . ¢ . I . VI 1- .r. A a I‘ I p. n. I . V’O" ‘
icinif». .. :59!“ .

i1:§!uwur9u.»r...ll¥:hui‘ .1».

oft... 31:111.! '1 (I.
s- L |I.:55CI.UI‘

In... I!
‘ lo..!nll.!..:!|h’\x I

”my. ,. _
'g'J-‘l;

.Q
.
L

L ‘
H‘s uny,’

‘ l

9.. :5le
111.... A;

 

 

v|ru II by

.9 *99 — Illllll\llllllllllllllllll‘ill \lllﬂlll

This is to certify that the

dissertation entitled I

Behavioral Models of Advanced Manufacturing
Technology Selection

presented by

Thomas Vincent Scannell

l
l
l
I
has been accepted towards fulﬁllment ‘
of the requirements for

1

Doctor of Philosophydegree in Business Administration

 

 

City“ (ZLJCLL I] will”
Major professor Sid.»

 

Date July 29, 1999

 

MSU is an Affirmative Action/Equal Opportunity Institution 042771

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

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1m campus-nu

 

BEHAVIORAL MODELS OF
ADVANCED MANUFACTURING TECHNOLOGY SELECTION

By

Thomas Vincent Scannell

A DISSERTATION
Submitted to
Michigan State University

in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Marketing and Supply Chain Management

1 999

ABSTRACT

A BEHAVIORAL MODEL OF
ADVANCED MANUFACTURING TECHNOLOGY SELECTION

By

Thomas Vincent Scannell

The advanced manufacturing technology (AMT) adoption decision process is
examined by adapting and applying two established behavioral models. One research
model is directly developed from the Theory of Planned Behavior (TPB) and is labeled
the Advanced Manufacturing Technology Selection Model - Theory of Planned Behavior
(AMTS-TPB). A second model is adapted from the Technology Acceptance Model
(TAM) discussed by Davis et a1. (1989), and is labeled the Advanced Manufacturing
Technology Adoption Model (AMTAM). AMT for this research is restricted to those
technologies that machine, assemble, fabricate or otherwise operate on materials and
products. This included CNC/DNC machines, material working lasers and robots.

The TPB was shown to be applicable in an industrial context that focused on a
speciﬁc set of AMTs. Further, the TAM, which was designed to examine acceptance of
information systems, was shown to be applicable in a research context other than the
setting for which it was intended. Both models explained a substantial amount of
variance in behavioral intentions to adopt an AMT using a parsimonious set of factors.

Three key relationships have implications for manufacturers of AMTs. First, the
signiﬁcant inﬂuence that perceived useﬁilness of the AMT has on a decision-maker’s
intentions indicates that increased marketing of the performance beneﬁts of AMT to

potential users is warranted. Second, supplier support did not have a signiﬁcant inﬂuence

on perceived behavioral control. Manufacturers of AMTs may want to form stronger
relationships with both users of AMTs and material and components suppliers as well.
Third, subjective norms were found to signiﬁcantly inﬂuence intentions. AMT
manufacturers may want to consider broadening their marketing techniques to a cross-
functional representation of personnel at a potential adopting ﬁrm.

This research also has implications for key stakeholders within AMT adopting
ﬁrms. First, familiarity with AMT was found to have a signiﬁcant inﬂuence on perceived
behavioral control, which in turn was found to have a signiﬁcant inﬂuence on perceived
usefulness in the AMTAM. These relationships suggest that either AMT adoption
decisions are part of an overall AMT strategy or that there is a “comfort zone” in which
manufacturing managers make decisions. Second, subjective norms are key determinants
of intentions in this research context. Advocates of increased technology adoption within
a ﬁrm may wish to exploit the cross-functional impact on a decision-maker by
determining the speciﬁc attributes of the AMT that each function is most inﬂuenced by,
then marketing these attributes to each function.

Third, despite perceived behavioral control having a signiﬁcant inﬂuence on
perceived usefulness in AMTAM, PBC was found to have a non-signiﬁcant inﬂuence on
intentions to adopt an AMT in the AMTS-TPB. This suggests that users may be willing
to tolerate a difﬁcult adoption process in order to realize signiﬁcant competitive beneﬁts,
while no amount of perceived behavioral control or ease of adoption will compensate for

a system that does not provide extensive beneﬁts.

Covyright by
THOMAS VINCENT SCANNELL
1999

This dissertation is dedicated to my Mom and Dad, my ﬁrst and best teachers who
sacriﬁced so much to ensure their twelve children received a good education; to my
children Emily, Margaret and Mark, for bringing so much joy to my life; and to my wife,

Jane, who I love so very much.

ACKNOWLEDGMENTS

Many people directly and indirectly contributed to the completion of this research.
Without the support, input and encouragement of the entire “team” this dissertation could
not have been completed. My dissertation committee co-chairs, Steven Melnyk and
Roger Calantone, provided invaluable inputs and guidance. They both spent an
inordinate amount time on this research effort from the concept phase to completion. My
committee, Robert Handﬁeld and Dan Krause, contributed immensely as well. I respect
and thank all of you for your contributions.

I also thank all of the Operations and Sourcing Management Faculty at Michigan
State University for teaching me how to conduct interesting and scientiﬁc research, and
for providing me with a solid foundation in the subject matter. I particularly thank Gary
Ragatz, who critically analyzed my dissertation proposal and who I turned to for much
advice throughout my career at Michigan State University. I also especially recognize
Robert Monczka, who provided encouragement, support and wisdom, as well as
extensive insights into purchasing and supply chain management theory and practice as
they relate to my research.

I also thank the National Association of Purchasing Management. The funding
received from the Doctoral Dissertation Grant Competition was immensely appreciated.

Finally, I thank my wife Jane, for understanding and supporting me in this
endeavor. She is a strong and special person who I love and respect. I am proud to be her

husband.

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. ix
LIST OF FIGURES ............................................................................................................. x
LIST OF ABBREVIATIONS ............................................................................................ xi
Chapter 1 .............................................................................................................................. 1
INTRODUCTION ................................................................................................................ 1
1.1 Problem Statement and Research Question .............................................................. 1
1.2 Research Framework ................................................................................................ 1
1.3 Research Models ....................................................................................................... 3
1.3.1 The Theory of Planned Behavior ....................................................................... 3

1.3.2 Manufacturing Technology Selection Models ................................................... 5

1.4 Research Methodology ............................................................................................. 5
1.5 Potential Outcomes and Contributions of the Research ........................................... 8
1.6 Summary ................................................................................................................... 9
Chapter 2 ............................................................................................................................ 10
LITERATURE REVIEW ................................................................................................... 10
2.1 The Theory of Planned Behavior ............................................................................ 10
2.2 AMT Selection Criteria — Exogenous Factors ........................................................ 12
2.2.1 Attitude — Perceived Usefulness ...................................................................... 12
2.2.2 Perceived Behavioral Control .......................................................................... 14
2.2.3 Subjective Norms ............................................................................................. 17
2.2.4 Summary of AMT Selection Criteria — Exogenous Factors ............................ 17

2.3 AMT Selection Criteria — Endogenous Factors ...................................................... 18
2.3.1 Direct Extension of the TPB ............................................................................ 18
2.3.2 Adaptations of the Theory of Reasoned Action ............................................... 19
Chapter 3 ............................................................................................................................ 31
METHODOLOGY ............................................................................................................. 3 1
3.1 Unit of Analysis ...................................................................................................... 31
3.2 Data Sample ............................................................................................................ 32
3.3 Required Sample Size ............................................................................................. 32
3.4 Data Collection ....................................................................................................... 33
3.5 Measures ................................................................................................................. 33
3.6 Proposed Data Analysis .......................................................................................... 35
Chapter 4 ............................................................................................................................ 37
DATA ANALYSIS ............................................................................................................ 37
4.1 Introduction ............................................................................................................. 37
4.2 The Data .................................................................................................................. 37
4.2.1 Data Collection ................................................................................................ 37
4.2.2 Characteristics of Responding Firms ............................................................... 41

vii

4.3 Data Analysis .......................................................................................................... 49

4.3.1 Descriptive Statistics ........................................................................................ 50

4.4 Structural Analysis .................................................................................................. 52
4.4.1 Advanced Manufacturing Technology Adoption Model (AMTAM) .............. 53
4.4.2 Advanced Manufacturing Technology Selection Model (AMTS-TPB) ........... 57

4.5 Summary ................................................................................................................. 61
Chapter 5 ............................................................................................................................ 64
DISCUSSION OF RESULTS ............................................................................................ 64
5.1 Overview and Chapter Contents ............................................................................. 64
5.2 Discussion of Hypotheses Tests .............................................................................. 64
5.2.1 Hypotheses Common to the Two Models ........................................................ 64
5.2.2 Hypotheses Speciﬁc to AMTS-TPB ................................................................ 67
5.2.2 Hypotheses Speciﬁc to AMTAM ..................................................................... 68

5.3 Contributions of the Research ................................................................................. 69
5.3.1 Academic Contribution .................................................................................... 69
5.3.2 Managerial Contribution .................................................................................. 75

5.4 Limitations of the Research .................................................................................... 79
5.5 Recommendations for Future Research .................................................................. 81
5.5.1 Review of Potential Research Directions Previously Discussed ..................... 81
5.5.2 Other Potential Research Directions ................................................................ 83

5.6 Summary ................................................................................................................. 84
APPENDIX A: QUESTIONNAIRE .................................................................................. 86
BIBLIOGRAPHY .............................................................................................................. 93

viii

LIST OF TABLES

Table 1: Measurement of Selected TAM Constructs (Davis et al., 1989) ......................... 23
Table 2: External Factor Composition in ITAM (Lee, 1990) ............................................ 25
Table 3: Measurement of Selected ITAM Constructs (Lee, 1990) .................................... 26
Table 4: Results of First Wave Mailing ............................................................................. 38
Table 5: Results of Phone Contacts to Assess Non-Response Bias ................................... 39
Table 6: Results of Second Wave Mailing ........................................................................ 40
Table 7: T-tests for Equality of Means .............................................................................. 42
Table 8: SIC of Respondents ............................................................................................. 42
Table 9: Geographic Location of Respondents .................................................................. 43
Table 10: Title of Respondents .......................................................................................... 44
Table 11: Type of AMT Reported ..................................................................................... 44
Table 12: AMT Characteristics .......................................................................................... 45
Table 13: Descriptive Statistics for Measurement Variables ............................................. 51
Table 14: AMTAM Hypotheses Test Results .................................................................... 56
Table 15: AMTS-TPB Hypotheses Test Results ............................................................... 61
Table 16: Summary of Hypotheses for Both Models ......................................................... 63
Table 17: Contrast of Current Research to that of Davis et a1. (1989) .............................. 70
Table 18: AMTS-TPB — Comparison of Findings to TRA ................................................ 72
Table 19: AMTAM - Comparison of Findings to TAM .................................................... 73

ix

LIST OF FIGURES

Figure 1: Integrated I/O and AMT Selection Decision Process Framework ....................... 3
Figure 2: The Theory of Planned Behavior (Ajzen, 1991) .................................................. 4
Figure 3: Advanced Manufacturing Technology Adoption Model (AMTAM) ................... 6
Figure 4: Advanced Manufacturing Technology Selection Model (AMTS-TPB) ............... 6
Figure 5: The Technology Acceptance Model (Davis et al., 1989) ................................... 20
Figure 6: The International Technology Adoption Model (Lee, 1990) .............................. 24
Figure 7: Purchase and Installation Cost of the AMT ........................................................ 45
Figure 8: Projected ROI for the AMT Investment ............................................................. 46
Figure 9: Projected Payback Period in Months .................................................................. 46
Figure 10: Expected Useful Life in Years .......................................................................... 47
Figure 11: Installation and Debug Time in Months ........................................................... 47
Figure 12: Process Optimization Time in Months ............................................................. 48
Figure 13: Advanced Manufacturing Technology Adoption Model (AMTAM) ............... 53
Figure 14: AMTAM Measurement Model ......................................................................... 54

Figure 15:

AMTAM Fully Speciﬁed Model ...................................................................... 56

Figure 16: Advanced Manufacturing Technology Selection Model (AMTS-TPB) ........... 57
Figure 17: AMTS-TPB Measurement Model .................................................................... 59
Figure 18: AMTS-TPB Fully Speciﬁed Model ................................................................. 6O

AMT
AMTAM
AMTS-TPB

BI
CF A
CNC
DNC
EOU
I/O
ITAM
ITT
J IT
LM
LV
MV
NC
PBC
PU
RFC
SEM
SN
TAM
TPB
TRA

LIST OF ABBREVIATIONS

Advanced Manufacturing Technology
Advanced Manufacturing Technology Adoption Model
Advanced Manufacturing Technology Selection —Theory of Planned
Behavior

Behavioral Intention

Conﬁrmatory Factor Analysis

Computer Numeric Control

Direct Numeric Control

Ease of Use

Insourcing / Outsourcing

International Technology Adoption Model
International Technology Transfer

Just In Time

Lagrange Multiplier

Latent Variable

Measurement Variable

Numeric Control

Perceived Behavioral Control

Perceived Utilities

Resource Facilitating Conditions
Structural Equation Model(ing)
Subjective Norm

Technology Acceptance Model

Theory of Planned Behavior

Theory of Reasoned Action

xi

Chapter 1
INTRODUCTION
1.1 Problem Statement and Research Question
Advanced manufacturing technology (AMT) selection decisions are key to long

term business success (Hayes, Wheelright and Clark, 1989). However, research indicates
that such decisions are often made by focusing on narrowly deﬁned cost or ﬁnancial
assessment with little consideration of longer term strategic issues (Welch and Nayak,
1992). This research examines the relative importance of various criteria in the AMT
selection process using the Theory of Planned Behavior (TPB) (Ajzen, 1991) and the
Technology Acceptance Model (TAM) (Davis et al., 1989) as the underlying ﬁameworks.
The TAM is an adaptation of the Theory of Reasoned Action (TRA) (Ajzen and Fishbein,

1980). The primary research question is:

1) What inﬂuences a decision-maker’s intention to adopt an advanced manufacturing

technology?
1.2 Research Framework

Advanced manufacturing technology (AMT) is not being adopted across
industries at the high rate many researchers predicted despite its purported beneﬁts
(Canada and Sullivan, 1989; Hayes and Garvin, 1982). Part of the problem may be the
wide range of factors impacting AMT adoption decisions. Figure 1 presents an integrated
overview of the complexity of making AMT selection decisions and provides a broad
framework for this research. The framework, which was developed to guide a longer

term research effort, suggests that AMT selection decisions are made in the context of

fete—=5...— mmooeum 5392.9 gouge—um 9.52 ES 9: cog—ac“:— 9 2:99

 

wuozqasm 53 .80 a

 

 

 

 

 

oz 0386 mm; 383m

ton—ma 2392
N

98 13:38 .9. o
hum—on 35365 o
$33 0

86:9— Eon—F830

 

 

 

 

 

 

 

8:35 awe—938.9.
wﬁéoﬂsﬁwE 95
@2059 SEE Hoe—om

 
  

ugh—gm :8: 05
@2269 SEE“ 803m

     

 
  

 

 

 

 

     

   

  

 

 

     

 

was: .08 .
Esoacegam o
o 8394
00585 We ooBoésw 83850 3coEoouw<
awe—ounce... 33803209 new; 3:399
39:89:52 05 523 “239% amaze—2 \ 3:39:85 0
9 3:35qu 3825 9 mum—Banao BEBE 38:8 389 o

 

 

 

 

  

 

 

28:2:qu was 25.5

 

 

 

      

 

  
     

 

 

 

 

 

 

 

 

 

     

 

 

 

 

 
 

 

 

  

  

 

  

 

 

    
 
 

 

 

 

EB— 05 8389 EB— .20 o
beggaﬁoo 05 .«o 5&5 033—39 88m 825—8928
9 523.230 a 230m 8 E3230 98 088.809 o
E535 82:99 =Eo>o Bus—gm 808:?“ 3330 o
+ toga. 8586M o
02 - 63.9. o
322589. 3:28:52 83— conga—“ESQ a
89389. was Q2269 .cw_mo9 95 .8339 ES @2039 £389 5:80.988 980 o
20:35 2 E5350 baa—.55 ﬁESxM 9 $3596 .3335 35:9 oucotonxo «mum o
. 20:83:80 «28:0 9583.5
80““ a; A. o o O 0
mm; 8%? 2335 mm; owwuhuwﬂwwuomm .3” .
3&3 33:29:80 35 0
8558820 no 339 BEEF— 28 089. o
awe—2300.9. hwoﬁbm 33:29:80 35 .0ng o
98 .3389 .8355 $0839 93 883.80 #60 o

 

 

 

 

2.2.5 3.3.25

 

 

 

 

interrelated company strategy and insourcing/outsourcing (I/O) decisions. This
dissertation research focuses on the AMT selection decision by testing two AMT

adoption models.

1.3 Research Models

The research question will be assessed using the TPB and TAM (an adaptation of
the TRA) as the underlying theoretical frameworks. The TPB and TRA are general
behavior intention models that may be used to predict and/or explain behavior in a variety
of settings (Ajzen, 1991; Fishbein et al., 1975). Though the TPB and TRA have primarily
been used to study consumer behavior, they have also been used to examine technology
adoption decisions (Davis, Bagozzi and Warshaw, 1989; Dimnik and Johnston, 1993;
Lee, 1990). The TAM was developed by Davis et al. (1989) to speciﬁcally examine the
acceptance of information systems.
1.3.] The T hemy of Planned Behavior

Figure 2 presents the TPB general model. Behavioral Intention (BI) refers to a
decision-makers’ subjective probability that he/she will perform the speciﬁed behavior.
The TPB postulates that three conceptually independent factors determine behavioral
intention: 1) attitude (A), which reﬂects the extent to which an individual has a favorable
or unfavorable objective impression of the speciﬁed behavior; 2) subjective norms (SN),
which represents perceived social pressures to engage in or refrain from a speciﬁc
behavior; and 3) perceived behavioral control (PBC), which refers to the perceived

difﬁculty or ease of performing the speciﬁed behavior.

The TPB further postulates that attitude towards a behavior is driven by beliefs
and evaluations about the consequences of behavior. Relative or competitive advantage
is a key determinant of attitude (Davis et al., 1989; Rogers, 1983). The TPB also
postulates that SNs are determined by perceived expectations of a referent individual or
group. Referent groups in an industrial context might include executive managers or co-
workers for example. Finally, the TPB postulates that PBC is determined by control
beliefs. Control beliefs include self efﬁcacy (e.g., familiarity with the technology) and
resource facilitating conditions (e. g., supplier service and support) (Bandura, 1977;

Rogers, 1983).

 

 

Beliefs and Attitude Toward
Evaluations Behavior (A)

 

 

 

 

 

 

 

 

 

Expectations Subjective Norms
of Referent (SN)

Control ’| Perceived Behavioral

Behavioral Actual
Intention (BI) Behavior

 

 

 

 

 

 

\_.L./

Beliefs Control (PBC)

 

 

 

 

 

Figure 2: The Theory of Planned Behavior (Ajzen, 1991)

Davis et al. (1989) adapted the TRA and developed the TAM to examine
technology acceptance decisions. The TRA was a precursor to the TPB, with the primary
difference being the inclusion of perceived behavioral control in the TPB. The TAM
differed from the more generalizable TRA in the following ways.

1. Perceived usefulness is a direct determinant of behavioral intention in TAM, but
not in TRA or TPB.

2. Perceived ease of use is a direct determinant of attitude in TAM, but not in TRA
or TPB.

3. Perceived ease of use is a direct determinant of perceived usefulness in TAM, but

not in TRA or TPB.

4. Perceived ease of use does not directly determine behavioral intention in TAM,
but does in TRA and TPB.

5. Normative beliefs and subjective norms are not examined in TAM, but are in
TRA and TPB.

1.3.2 Manufacturing Technology Selection Models

Figure 3 (Advanced Manufacturing Technology Adoption Model - AMTAM) and
Figure 4 (Advanced Manufacturing Technology Selection Model -AMTS-TPB) present
the models developed for this research. Each arrow in each figure represents a
hypothesized relationship between constructs. The AMTS-TPB is developed and adapted
directly from the TPB. The AMTAM is developed and adapted directly from the TAM.

The decision to test two AMT selection models was based on the research
conducted by Davis et a1. (1989). In that research, one model labeled the “Technology
Acceptance Model (TAM)” was derived from the Theory of Reasoned Action (TRA). A
second model representing a direct application of the TRA was also examined. Chapter 2
(literature review) presents the detailed theoretical justiﬁcation for the models, while

Chapter 3 (methodology) discusses construct measurement.

1.4 Research Methodology

The unit of analysis is a speciﬁc advanced manufacturing technology selection.
Respondents will be asked to identify and discuss a recently adopted advanced
manufacturing technology that is used to fabricate, machine or assemble products at their

plant that has already been used to make an initial production run.

     
         
 

H6-AMTAM

  
  
  

     
  

F4: Attitude
Towards Adopting

   
  

F6: Intention to
Adopt AMT

     

Fl : Perceived H1 ‘AMTAM H7 ~AMTA

 

Usefulness

  

H4-AMTAM

HS-AMTAM

  

HZ-AMTAM

 

F2: Familiarity
with AMT

    
 
 

F5: Perceived
Behavioral Control

F3: Supplier
Support

   

  

H3-AMTAM

Figure 3: Advanced Manufacturing Technology Adoption Model (AMTAM)

  
 
 

   
 

F4: Attitude
Towards Adopting

   
   
 
 

F1: Perceived
Usefulness

HS-TPB

F2: Familiarity HZ-TPB

with AMT

 
       
 

  
 
 

F6: Intention to
Adopt AMT

 
 

  

F5: Perceived
Behavioral Control

     

 

F3: Supplier
Support

    
  

H7-TPB

  

H4-TPB

   

F7: Normative

Beliefs F8= SubJectwe

Norms

       

Figure 4: Advanced Manufacturing Technology Selection Model (AMTS-TPB)

Using previously established classiﬁcations of AMT (US Commerce Department,
1989; Small and Chen, 1997) such fabricating, machining or assembling technology may
include CNC/DNC machines, material working lasers and robots. Such AMT typically
includes flexible manufacturing systems and computer integrated manufacturing.
However, due to deﬁnitional issues regarding such systems, this research focuses on the
core machines. AMT excluded from this research includes design and engineering
technologies (e. g., computer aided design, computer aided process planning), automated
material handling technologies (e. g., automated storage and retrieval systems, automated
material handling systems), information technologies (e. g., material requirements
planning) and automated inspection and testing equipment.

Respondents will answer all questions with respect to the speciﬁc AMT and the
item(s) produced by this technology. Data will be gathered across a variety of industries
from for-proﬁt United States manufacturing ﬁrms that own and operate their own
production facilities. An executive or manager from manufacturing/industrial
engineering, plant/operations engineering or capital equipment purchasing will be
targeted as the respondent. The majority of questions are either eleven point Likert sealed
(0 to 10) or categorical. A copy of the survey appears in Appendix A. Chapter 3
(Methodology) speciﬁcally allocates each question (measurement variable) to a construct
(latent variable).

The overall model ﬁt and speciﬁc hypotheses for each model will be tested using
Structural Equation Modeling (SEM) techniques. SEM is a comprehensive multivariate

statistical approach to test hypotheses about relationships among measured (observed)

and latent variables. There are two stages associated with this methodology. In the ﬁrst
stage a measurement model (for each research model) will be tested using conﬁrmatory
factor analysis (CFA) to establish a satisfactory level of reliability and validity. In the
second stage, the degree to which each proposed model ﬁts the observed data will be
tested using EQS software with covariance matrices input.
1.5 Potential Outcomes and Contributions of the Research

AMT selection and investment impacts a company’s cost structure and
competitiveness and may be the most critical responsibility of manufacturing managers
(Hayes et al., 1989). Despite an increasing literature, there is still a signiﬁcant need for
theoretical and practical research to explain the AMT adoption process. This research
examines the impact of traditional factors purported to inﬂuence the AMT adoption
decision (e. g., quality, cost) as well as less frequently examined factors such as normative

beliefs and other behavioral control factors in an integrated decision model.

Companies that are potential adopters of AMT may use the decision models to
assess how AMT decisions are made and to determine which factors to weight more or
less in choosing an AMT. Proponents or opponents of AMT adoption within a company
may identify which factors to proactively manipulate in order to inﬂuence a key decision-

maker’s perceptions or evaluations of an AMT.

This research also has implications for AMT suppliers. By understanding what
most inﬂuences a buying ﬁrm’s decision to select an AMT, a supplying ﬁrm can develop
the necessary capabilities and marketing strategy to secure the buying ﬁrm’s business

(Dimnik et al., 1993).

Finally, this research extends the current body of knowledge by examining the
multiple factors purported to inﬂuence the AMT adoption decision in an integrated
model. It uses a well-established and generalizable research framework (i.e., the TPB) in
a relatively new research context to examine the impact of the decision criteria. It also
examines if a research framework designed speciﬁcally to investigate information

systems adoption decisions (i.e., the TAM) is applicable in a different context.

1.6 Summary

This research develops and tests two AMT selection models using the TPB and
TAM as the underlying frameworks. The objective is to better understand how perceived
beneﬁts, social norms and operational factors inﬂuence a decision maker’s attitude
toward AMT adoption.

The remainder of this dissertation is organized as follows. Chapter 2 reviews the
literature and formally develops the research models and hypotheses. Chapter 3 discusses
the research methodology. Data analysis is presented in Chapter 4, and a discussion of
results and conclusions is outlined in Chapter 5. Finally, all references are listed, and a

copy of the primary survey instrument is attached in the Appendix.

Chapter 2
LITERATURE REVIEW

This literature review is divided into three main sections. Section 2.1 reviews the
Theory of Flamed Behavior (TPB). The TPB provides the underlying theoretical
framework for the research. Section 2.2 examines the exogenous criteria believed to
inﬂuence AMT selection decisions. A set of what are believed to be the most salient
criteria based on the literature and interviews with academics and practitioners is
examined. Section 2.3 then discusses two research efforts that provide further theoretical
justiﬁcation and methodological direction for the research herein.

Each section develops the research hypotheses. Because two models are being
tested, the hypotheses for the model in Figure 3 (Advanced Manufacturing Technology
Adoption Model : AMTAM) will be labeled “HX-AMTAM”. Similarly, the hypotheses
for the model in Figure 4 (Advanced Manufacturing Technology Selection Model -—
Theory of Planned Behavior : AMTSTPB) will be labeled “HX-TPB”.

2.1 The Theory of Planned Behavior

Figure 2 presented the TPB general model. The model has been shown to predict
behavioral intention and actual behavior rather well in research ranging from intentions to
buy toothpaste, to intentions to conserve energy, to intentions to vote in presidential
elections (Ajzen, Timko and White, 1982; Ryan, 1982; Sheppard, Hartwick and
Warshaw, 1988; Stutzman and Green, 1982). Sheppard et al. (1988) conducted a meta-
analysis of TPB research and concluded the model has strong predictive validity and is
robust even in research situations that do not fall within the boundary conditions

originally deﬁned by the model. Though the TPB has primarily been used to study

10

consumer behavior, it has been used to examine technology adoption decisions as well
(Davis, Bagozzi and Warshaw, 1989; Dimnik and Johnston, 1993; Lee, 1990). The
application of the TPB to manufacturing technology adoption research is considered a
novel approach that shows promise for AMT adoption research (Dimnik et al., 1993).

Behavioral Intention (BI) refers to a decision-makers’ subjective probability that
he/she will perform the speciﬁed behavior. The TPB postulates that three conceptually
independent factors determine behavioral intention: 1) Attitude (A), which reﬂects the
extent to which an individual has a favorable or unfavorable objective impression of the
speciﬁed behavior; 2) Subjective Norms (SN), which represents perceived social
pressures to engage in or refrain from a speciﬁc behavior; and 3) Perceived Behavioral
Control (PBC), which refers to the perceived difﬁculty or ease of performing the
speciﬁed behavior. Speciﬁcally, Behavioral Intention is approximated by: (Behavioral
Intention = wlA + szN + w3PBC), where w1, w; and w3 are empirically determined
regression coefﬁcients.

The TPB further postulates that attitude towards a behavior is driven by beliefs (b)
and evaluations (e) about the consequences of behavior, such that (A = 2 b, e,-). Relative
or competitive advantage is a key determinant of attitude (Davis et al., 1989; Rogers,
1983). The TPB also postulates that SNs are determined by perceived expectations of a
referent individual or groups (nb) multiplied by the individual’s motivation to comply
(me) with those expectations, such that (SN = Z nb, mci). Referent groups might include
co-workers for example. Finally, the TPB postulates that PBC is determined by control
beliefs (c) multiplied by the perceived power (p) of that control factor to support or

inhibit performance of the behavior, such that (PBC = Z c.- pi). Control beliefs include

11

self efﬁcacy (e. g., familiarity with the technology) and resource facilitating conditions
(e.g., supplier service and support) (Bandura, 1977; Rogers, 1983).
2.2 AMT Selection Criteria — Exogenous Factors

Manufacturing technology and capabilities directly impact a ﬁrm’s ability to
deliver higher quality products at lower costs while providing greater responsiveness to
changing customer requirements. Thus, AMT selection requires consideration of both
strategic and operational factors (Robinson, 1988; Samli, 1985) that should reﬂect the
importance of overall company policy (Porter, 1983). Such criteria may include
perceived usefulness, control beliefs and normative beliefs (Davis et al., 1989; Dimnik et
al., 1993; Phillips, Calantone and Lee, 1994).
2. 2.1 Attitude — Perceived Usefulness

Manufacturing competitive performance criteria include innovation, cost, quality
and time (Anderson et al., 1989; F erdows et al., 1990). These performance dimensions
represent the potential relative advantages from adopting AMT. Consistent with previous
research (Davis et al., 1989; Lee, 1990) relative advantage dimensions are collectively
referred to as “perceived usefulness” in this research.
2.2.1.1 Innovation

Internal manufacturing capabilities and technologies are increasingly recognized
as key sources of product and process innovation (Skinner, 1969; Warnock, 1996). By
selecting the appropriate AMT, a company can control and accelerate technological
change and improve innovation performance (Green, Mahmoud and Zimmerer, 1994;

Stacey and Ashton, 1990). Further, AMT suppliers may drive innovations when closely

12

integrated into product development processes. These ﬁndings suggest that innovation

criteria may inﬂuence a decision-maker’s attitude toward adopting an AMT.

2.2.1.2 Cost

The increasing formal adoption of Total Cost of Ownership analysis and Activity
Based Costing (Monczka and Trent, 1997) methods suggests that companies recognize
that decisions based on purchase price alone are unlikely to yield optimal outcomes. In a
broad sense, the costs resulting from quality performance, technology capabilities, impact
on employee morale, supplier relations, etc., combined with actual purchase price
constitute total costs of a process (Raunick and Fisher, 1972).

However, it is extremely difﬁcult to develop an accurate and reliable dollar
measure of a decision’s impact on supplier relations for example. Thus, this research
examines cost in a narrow sense, viewing cost as a single factor that needs to be
considered in conjunction with other factors. Narrow costs to make a product internally
include material, direct labor, direct factory overhead, selling and administrative expense,
and other general overhead (Raunick et al., 1972). Production costs are a key determinant
of AMT selection (Herbst, 1990; Wamock, 1996) and thus may inﬂuence a decision-
maker’s attitude toward adopting an AMT.
2.2.1.3 Quality

Research suggests a link between quality performance and overall ﬁrm
performance (Garvin, 1987; Miller and Roth, 1994; Vickery et al., 1993). In many
industries, quality is a critical competitive dimension that has become an order qualifier
(Hill, 1985). Quality performance dimensions may include performance, features,

reliability, conformance, durability, serviceability, aesthetics and perceived quality

13

(Garvin, 1987). AMT selection directly impacts quality performance (Skinner, 1984;
Wamock, 1996), making quality a determinant of AMT selection (Herbst, 1990;

Wamock, 1996) that may inﬂuence a decision-maker’s attitude toward adopting an AMT.

2.2.1.4 Time

Production cycle time reduction has been identiﬁed as a key source of competitive
advantage across industries (Handﬁeld, 1995). Just in time (JIT) purchasing and
manufacturing strategies are increasingly being adopted to reduce overall cycle times
(Frazier, Spekman and O'Neal, 1988). By assessing and comparing throughput and set up
times of competing technologies, a ﬁrm can control and minimize production cycle times
to drive competitive advantages (Shingo, 1985). This suggests that cycle time criteria
may inﬂuence a decision-maker’s attitude toward adopting an AMT. This research

focuses on overall production cycle time, which includes set-up, run and tear-down time.

2.2.1.5 Perceived Usefulness Hypotheses

Innovation, cost, quality and time represent key components of perceived
usefulness. Consistent with the TAM and the TPB, the following hypotheses are
developed (note that the hypotheses for the two models in this case are identical.)

Hl-AMTAM: Perceived usefulness of the AMT directly inﬂuences a decision
maker’s attitude toward adopting the AMT.

Hl-TPB: Perceived usefulness of the AMT directly inﬂuences a decision maker’s
attitude toward adopting the AMT.

2. 2.2 Perceived Behavioral Control
Control beliefs include those factors that inﬂuence a decision-maker’s belief that a

new technology can be successfully installed and utilized. There are two components that

14

inﬂuence perceived behavioral control: self-efﬁcacy and resource facilitating conditions.
Self-efﬁcacy in this research is the judgement that the AMT is compatible with existing
operations. Resource facilitating conditions relates to external support for installing and
utilizing the AMT such as AMT supplier support.
2.2.2.1 Self-Efﬁcacy - Familiarity with AMT

A key set of factors in AMT selection decision include the availability of capable
operators and supervisors, and the extent to which the AMT is compatible with and
adaptable to current operations (Mansﬁeld, Romeo, Schwartz, Teece, Wagner and Brach,
1982). Technology diffusion occurs more effectively among potential adopters who
already have related knowledge or experience in that technology, reﬂected by the
technical experience of managers and technicians (Dickerson and Gentry, 1983; Gatignon
and Robertson, 1985). Further, the compatibility of technology with existing operations
is one of the critical determinants of successful technology adoption (Rogers et al., 1971).
Compatibility indicates the degree to which a technology is perceived to be consistent
with the current values, needs and past experiences of the recipients (Rogers, 1983;
Rogers et al., 1971). To the extent that a technology requires changes in other aspects of
operations, the slower will be the rate of technology acceptance (Robertson, 1971).

Previous experience with a technology, as well as the compatibility and
adaptability of a technology to existing conditions are therefore key factors that impact
AMT selection. A high level on any or all of the factors provides conﬁdence that a new
AMT can be successfully adopted. Thus, the following hypotheses are proposed (note

that the hypotheses for the two models in this case are identical.)

15

H2-AMTAM: Familiarity with AMT inﬂuences a decision-maker’s perceived
behavioral control over successfully adopting the AMT.

H2-TPB: Familiarity with AMT inﬂuences a decision-maker’s perceived

behavioral control over successfully adopting the AMT.
2.2.2.2 Resource Facilitating Conditions

If a company considering adopting a new AMT does not have the infrastructure
(i.e., skilled employees and compatible operations) to support the implementation and
operation of the AMT, the AMT supplier may need to provide extensive support
(Robinson, 1988; Samli, 1985). The resources committed by a supplier to the
implementation and utilization of technology directly affects the recipient’s ability to
absorb that technology (Steele, 1974). Technological resources such as tooling, provision
of spare parts, instruction manuals, on-site training and temporary use of supplier
operators may be needed to facilitate the AMT adoption (Lee, 1990).

Further, prior experience between an industrial buyer and supplier establishes a
pattern of communication and understanding that allows for more effective transactions
(Morris and Holman, 1988). Research suggests that technology transfer occurs most
effectively when buyer and seller maintain relations for a long period of time (Basche et
al., 1975), that established buyer-supplier relationships more readily facilitate technology
adoption (Wind, 1970), and that prior experiences between transfer parties positively

inﬂuences future transactions (Calantone, Lee and Gross, 1990).

Thus, the following hypotheses are proposed (note that the hypotheses for the two
models in this case are identical.)
H3-AMTAM: Familiarity with AMT inﬂuences a decision-maker’s perceived

behavioral control over successfully adopting the AMT.

16

H3-TPB: Familiarity with AMT inﬂuences a decision-maker’s perceived

behavioral control over successfully adopting the AMT.
2. 2.3 Subjective Norms

A decision-maker’s behavioral intention is inﬂuenced by strong referent groups to
the extent that the decision-maker perceives a need to behave in a manner consistent with
what the inﬂuential people think he/she should do regarding the behavior (Fishbein et al.,
1975). That is, internal normative beliefs may create pressure on decision-makers to
select an option that they believe a key referent group would favor (Bumkrant et al.,
1988). Normative beliefs inﬂuence a manufacturing manager’s promotion of adopting
advanced manufacturing technologies (Dimnik et al., 1993). In an industrial context,
upper management, co-workers and subordinates may either directly or indirectly express
their opinion on what they believe to be the best choice for a decision-maker, whether the
opinion is based on objective criteria or is politically motivated. Thus, the following

hypothesis is proposed (note that subjective norms are not part of AMTAM.)

H4-TPB: Normative beliefs towards AMT selection inﬂuence a decision-maker’s

subjective norms towards AMT adoption.
2. 2.4 Summary of AM T Selection Criteria - Exogenous Factors

The previous sections focused on perceived usefulness criteria, normative beliefs
and control beliefs that may inﬂuence a decision-maker. There are certainly many other
factors that may impact AMT selection such as: 1) government policies that promote,
hinder or support adoption of certain technologies (Calantone et al., 1990; Robinson,

1988); 2) natural environment concerns (Chatterji, 1995; Porter and van der Linde, 1996);

17

and 3) restrictions placed by a supplier on the transfer of an AMT such as tied buying
provisions (Robinson, 1988). However, sample size concerns and interviews with
industry professionals discouraged the development and test of a model that included all

of these factors.

2.3 AMT Selection Criteria — Endogenous Factors

As previously noted, two models are tested in this research. The AMTS-TPB is a
direct extension of the general TPB and thus relationships between exogenous factors in
this model are direct extensions of the general TPB. The AMTAM is adapted from the
previous research (i.e., the TAM) so that relationships between exogenous factors in
AMTAM are direct extensions of the speciﬁc TAM framework.
2. 3. 1 Direct Extension of the T PB

Consistent with the TPB, decision-makers’ attitudes inﬂuence or predict the
behavioral intention of the decision-makers in a variety of contexts (Ajzen, 1985;
F ishbein et al., 1975), and in technology adoption in particular (Davis et al., 1989; Lee,
1990). Therefore, the following are proposed (note that the hypotheses for the two
models in this case are identical.)

HS-TPB: Attitude toward adopting the AMT directly inﬂuences a decision-
maker’s behavioral intention to adopt the AMT.

H7-AMTAM: Attitude toward adopting the AMT directly inﬂuences a decision-
maker’s behavioral intention to adopt the AMT.

Also consistent with the TPB, perceived behavioral control and subjective norms
inﬂuence or predict the behavioral intention of the decision-makers in a variety of

contexts (Ajzen, 1985; Fishbein et al., 1975), and in technology adoption in particular

18

(Davis et al., 1989; Lee, 1990). Therefore the following are proposed (note that no such
relationships are posited in the AMTAM.)

H6-TPB: Perceived behavioral control over adopting the AMT directly inﬂuences
a decision-maker’s behavioral intention to adopt the AMT.

H7-TPB: Subjective norms toward adopting the AMT directly inﬂuence a
decision—maker’s behavioral intention to adopt the AMT.

2.3.2 Adaptations of the Theory of Reasoned Action
This section of the literature review presents two studies that examined the
adoption of new technologies. These two research efforts provided a more context

speciﬁc framework for the research herein.

2. 3. 2. 1 The Technology Adoption Model

Davis, Bagozzi and Warshaw (1989) both directly used and adapted TRA, a
precursor to the TPB, to model user (MBA Students) acceptance of computer systems
(WriteOne software application) in an organizational context (an MBA program.) Note
that the TPB broadened the TRA’s scope to include non-volitional behavior, or those
activities requiring resources, opportunities and speciﬁc skills by adding an exogenous
construct labeled “Perceived Behavioral Control.”

The adaptation of the TRA by Davis et al. (1989), in the form of the TAM,
examined linkages between users’ beliefs in terms of ease of use and perceived utility of a
computer system, and users’ attitude, behavioral intentions and actual computer system

adoption behavior as shown in Figure 5.

19

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Perceived
External _ Usefudpess
Variables
Attitude Behavioral Actual
Toward . .
. , Intention _ Behavror
Usrng (BI)
, (A)
Perceived
Ease of
External
Variables Use
(ECU) ~

 

 

 

 

 

 

Figure 5: The Technology Acceptance Model (Davis et al., 1989)

TAM proposes that perceived usefulness and perceived ease of use are two key
belief structures in the study of computer acceptance behavior. Perceived usefulness (U)
was deﬁned as a potential user’s subjective probability that the technology would increase
job performance within an organizational context.

Perceived ease of use (EOU) was deﬁned as the extent to which the target
system’s use or implementation would be free of effort from the prospective user’s
perspective (Davis et al., 1989). The authors note that previous research efforts and
factor analyses conducted in prior studies (Hauser and Shugan, 1980; Larcker and Lessig,
1980; Swanson, 1974) suggested that Perceived Usefulness and Perceived Ease of Use
are statistically distinct dimensions that are linked to attitudes and usage. The Perceived
Behavioral Control construct for the research herein is similar to the Perceived Ease of

Use construct of Davis et al. (1989).

20

Perceived Ease of Use is theorized to be determined by external variables such as
system features speciﬁcally designed to increase ease of use (e. g., menus, mice, touch
screens), and/or services (e. g., training and documentation.) Perceived Ease of Use in
TAM inﬂuences attitude (A) and behavior intention (BI) through two mechanisms: self-
efﬁcacy and instrumentality (Davis et al., 1989).

The Perceived Ease of Use-Perceived Usefulness relationship captures the
instrumentality of Perceived Ease of Use. Perceived Ease of Use may be instrumental in
the sense that effort saved through Perceived Ease of Use may be redeployed in other
productive efforts increasing overall efﬁciency and performance. Thus, Perceived
Usefulness is posited to be determined by external variables (e.g., objective design
characteristics), and ease of use as given by: [Perceived Usefulness = Perceived Ease of
Use + External variables] Note that no such Perceived Ease of Use-Perceived
Usefulness relationship is posited for direct extensions of the TPB. Thus, for the
AMTAM only, the following is proposed:

H4-AMTAM: Perceived Behavioral Control directly inﬂuences the Perceived

Useﬁilness of the AMT.

The direct Perceived Ease of Use-Attitude relationship is designed to capture the
inﬂuence of the self-efﬁcacy aspect of Perceived Ease of Use as previous research
suggests (Malone, 1975). Note that no such relationship is posited in the direct extension
of the TPB. Thus, for the AMTAM only, the following is proposed:

HS-AMTAM: Perceived Behavioral Control directly inﬂuences Attitude Toward
Adopting the AMT.

21

TAM also postulates that Behavioral Intention is jointly determined by Attitude
and Perceived Usefulness. The Perceived Useﬁrlness-Behavioral Intention relationship in
TAM is driven by the concept that within organizational settings, people may form
intentions towards an activity or behavior they believe will increase their performance
which may be evoked above and beyond the positive and negative feelings which may be
evoked toward the behavior per se. Thus, the Perceived Usefulness-Behavioral Intention
relationship hypothesizes that people form intentions toward computer system usage
based to a large extent on a cognitive appraisal of how it will improve their performance
(Davis et al., 1989). Previous research (Bagozzi, 1982; Brinberg, 1979) provides
theoretical and empirical justiﬁcation for this causal link. Note that no such Perceived
Usefulness-Behavioral Intention relationship is posited for direct extensions of the TPB.
Thus, for the AMTAM only, the following is proposed:

H6-AMTAM: Perceived Usefulness directly inﬂuences Intention to Adopt the

AMT.

The measurements of selected TAM constructs relevant to the research herein are
listed in Table 1. All questions can be Likert scaled using endpoints “strongly disagree”

to “strongly agree” for example.

22

Table 1: Measurement of Selected TAM Constructs (Davis et al., 1989)

 

Ease of use

I would ﬁnd it easy to get WriteOne to do what I want it to do
It would be easy for me to become skillful at using WriteOne
Ieaming to operate WriteOne would be easy for me

I would ﬁnd WriteOne easy to use

 

Usefulness

Using WriteOne would improve my performance in the MBA program
Using WriteOne in the MBA program would increase my productivity

Using WriteOne would enhance my effectiveness in the MBA program
I would ﬁnd WriteOne useful in the MBA program

 

 

Behavioral Intention
Assuming I had access to WriteOne, I intend to use it
Assuming that I had access to WriteOneﬁredict I would use it

 

Davis et a1. (1989) conducted a longitudinal study and used regression analysis to
test the relationships proposed by TAM. Major conclusions from the TAM research
include the following.

1. Computer use can be predicted reasonably well from intentions.

2. The Attitude-Intention relationship was statistically signiﬁcant at Time One, but
not at Time Two.

3. The Perceived Usefulness-Intention relationship was statistically signiﬁcant at
both Time One and Time Two. Further, the relationship was stronger at Time
Two.

4. The Ease of Use-Attitude relationship was not signiﬁcant at Time One, but was
signiﬁcant at Time Two.

5. The Perceived Usefulness-Attitude relationship was signiﬁcant at both time
periods.

6. The Base of Use-Perceived Usefulness relationship was not signiﬁcant at Time
One, but was signiﬁcant at Time Two.

In terms of the direct application of the TRA to examine intentions to use the
computer program, the following was observed.

1. The Attitude-Intention relationship was statistically signiﬁcant at both Time One
and Time Two.

23

 

2. The Subjective Norm-Intention relationship was not signiﬁcant in either time
periods.

2.3.2.2 The International Technology Adoption Model

The International Technology Adoption Model (ITAM) proposed by Lee (Lee,
1990) is shown in Figure 6. Four “internal” variables of perceived utilities (PU),
perceived ease of application (EOA), attitude toward adoption (A) and behavioral
intention (BI), and six “external” variables which theoretically inﬂuence the internal
variables either directly or indirectly were identiﬁed. The six external variables and their
associated lower order variables are listed in Table 2. Not all of these ITAM constructs
are included in the research herein, as concerns about sample size encouraged the
development and testing of more parsimonious model. Only those constructs, measures

and results most salient to the proposed research will be examined in detail.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Competitive
Environment —
Government and
International
Inﬂuences — Perceived
Utilities of
7’ Technology
. (PU)
Technological , d
Beneﬁts 1:33;: IBehavioral
Adopting ‘ nenron o
(A) Adopt
Suppliers’ -- (BI)
Inﬂuences "— Perceived
Ease of
_. Application
Prior (EOA)
Experiences H
Cultural Affmity

 

 

 

 

Figure 6: The International Technology Adoption Model (Lee, 1990)

24

Table 2: External Factor Composition in ITAM (Lee, 1990)

 

Competitive Environment

Industry concentration

Competitive price intensity

Demand uncertainty
Communication openness

Domestic availability of comparable technology
Government and International Inﬂuences
0 Economic incentives

- Restrictions on transfer practices
Technological Beneﬁts

0 Relative advantages

0 Technical compatibility

Supplier’s Inﬂuences

0 Resource support

0 Restrictions on transfer

Prior experience

0 Experience with similar technology

a Working experience with technology supplier
Cultural Afﬁnity

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The measures of selected constructs in the lTAM relevant to the research herein
are summarized in Table 3. The measures are paraphrased to save space and therefore do

not reﬂect the actual wording or grouping in Lee’s (1990) questionnaire.

25

Table 3: Measurement of Selected ITAM Constructs (Lee, 1990)

 

 

Relative Advantages

Our company’s product quality will be improved by this new technology.

Our company’s total production output quantity will be increased by this new technology.

Worker productivity will be increased by this new technology.

The new technology can be the answer to some production problems we found in our current production.
Technical Compatibility

The new technology can be utilized with other manufacturing equipment our company is currently using.
The new technology is suitable for the current raw materials our company is buying.

The new technology is compatible with our current production environment.

Supplier Resource Support

The supplier will provide the necessary equipment delivery support

The supplier will provide the necessary equipment installation support.

The supplier will provide the necessary equipment maintenance support.

The supplier will provide the necessary equipment service and repair support.

The supplier will provide in plant design support

The supplier will provide product design support

The supplier will provide on site assistance in operation

The supplier will provide the necessary on-site assistance during installation and start-up.

The supplier will provide the necessary operating manuals.

The supplier will provide the necessary spare parts.

The supplier will provide the necessary technical adjustments on equipment

The supplier will provide the necessary manager/engineer/technician training.

Prior Experience With Similar Technology

Have you ever used any similar technology before?

Has your company ever utilized any technology or equipment that is similar to the new technology?
Prior Experience with Technology Supplier

Have you ever worked with people from that particular country or company before?

As far as you know, has your company ever done business with that particular company or country?
Perceived Utilities

The new technology is good for the future of our company.

Our products will be more competitive in international markets because of this technology.

Our products will be more competitive in domestic markets because of this technology.

The overall performance of our company will be improved because of this technology.

The new technology will be helpﬁil to my job performance.

The new technology will help me manage production more effectively.

Perceived Ease of Application

It is difﬁcult to transfer that particular technology from the foreign company

It takes a long time to transfer the new technology into our company

I can foresee problems would happen when the new technology is used in our current production facilities
I have to spend a lot of time to learn to use the new technology or equipment

Attitude Toward Adopting the Technology

Adopting the new technology from the foreign company is a good idea.

I would feel very conﬁdent if our company decided to adopt the new technology from the foreign company
It is appropriate to adopt the new technology from a foreign country into our current production facilities
It will be very beneﬁcial to our company if we decide to adopt the new technology from a foreign country
My opinion about adopting the new technology is very favorable

Behavioral Intention

I intend to push for adoption of the new technology in our company

If I were asked my opinion regarding adoption of the new technology, I intend to say something favorable
I intend to recommend the adoption of the new technology to our company

If I could make the decision for our company, I would adopt the new technology

 

26

 

Only the key ﬁndings in Lee’s research that are relevant to the research herein are
discussed below. Before proceeding, it is important to note that Lee’s research sample
consisted mainly of engineers, managers and administrators of Chinese companies in the
coal mining industry in 1990. As such, the ﬁndings may not be generalizable across
nations or industries.

Perceived Ease of Application (E 0A) A technology that is readily applied can
reduce the time and effort invested in implementing it. This construct is similar to the
perceived behavioral control construct in the research herein. Lee (1990) found that:

o Perceived EOA had direct and positive effects on attitude toward adopting
(A), though the relationship was not as strong as anticipated.

0 Perceived EOA had moderate direct and positive effects on perceived utilities
of technology (PU).

Perceived Utilities of Technology (PU) PU represents the subjective probability
that adoption of the technology within the current production system will increase
organizational and personal utilities. Lee measured this construct with several lower
order constructs as reﬂected in Figure 6. There is no similar construct in the research
herein, though two of the lower order constructs in Lee’s research to be discussed below
(relative advantages and technical compatibility) form part of this construct. Lee (1990)
found that:

- PU directly and positively affected attitude (A).

o PU had an indirect effect on behavioral intention (BI) which operates through
attitude (A).

o PU directly and positively effected BI.

27

Technological Benefits: Relative Advantage Lee structured relative advantage as
a lower order construct for the higher order construct labeled perceived utilities. In the
research herein, perceived usefulness of the AMT is similar to Lee’s relative advantage
construct. Lee found that:

0 Relative advantages of a new technology were positively associated with the
perspective adopter’s perceived utilities of technology.

Technological Beneﬁts: Technical Compatibility Lee structured technical
compatibility as a lower order construct for the higher order construct perceived utilities.
In the research herein, technical compatibility is structured as a key component of self-

efﬁcacy. Lee found that:

0 Technical compatibility, as perceived by the perspective adopter, positively
affected perceived utilities of technology.

Supplier ’s Influences: Resource Support Lee (1990) structured supplier
inﬂuences as composed of both resource support and restrictions on transfer. Only
resource support is examined in the research herein, as concerns about sample size
encouraged the test of a more parsimonious model. Lee found that:

0 There was no support for the hypothesis that the more extensive resource support
from the supplier, the greater the perceived utilities of technology.

0 There was no support for the hypothesis that the more extensive resource support
from the supplier, the greater the perceived ease of application.

Attitude Toward Adopting (A) This construct is very similar to the attitude
construct in the research herein. Lee found that:

o Attitude toward adopting had a strong direct and positive effect on B1.

28

2.3.2.3 Implications for Research

The research conducted by Davis, Bagozzi and Warshaw (1989) examined user
adoption of computer software. The study examined adoption of a computer application
in a non-industrial yet “organizational” setting using MBA students. The study provided
insight into the use of TRA for technology adoption research and establishes a basic
foundation for subsequent research in the area. Because the research was conducted in a
non-industrial albeit organizational context, it may not be directly applicable to the
research herein. However, it was important for the research herein in terms of
measurement of constructs and identiﬁcation of the strength of relationships between key
criteria in a technology selection context. Other researchers have built on the work of
Davis et al. (1989) by adapting the TAM, TRA and/or TPB to examine the acceptance of

information systems.

Lee’s (1990) work, which built on the work of Davis et al. (1989), is directly
applicable to technology adoption in an industrial context. Though adaptations of the
TRA have been used many times to examine consumer choice intentions in a variety of
contexts (Sheppard et al., 1988), Lee’s research is the only application towards
manufacturing technology adoption research that this researcher could ﬁnd. Dimnik and
Johnston (1993) used general propositions from the TRA and TPB not to examine AMT
adoption per se, but the championing behaviors of manufacturing managers for AMT
adoption.

However, Lee’s research was limited to the Chinese coal mining industry. The

special economic, social and political conditions in China may have impacted the

29

ﬁndings. Regardless, the work helped further measurement of constructs and provided
further insights into the technology adoption decision in an industrial context. Lee
recommended a broader sample and possible reﬁnement of the scales. These
recommendations helped guide the development of the research models and questionnaire

for the research herein.

30

Chapter 3
METHODOLOGY

This chapter focuses on data collection and analysis. The theoretical foundation
for the research model and associated hypotheses were presented in the previous chapters
and therefore will not be repeated in this chapter.

3.1 Unit of Analysis

The unit of analysis is a speciﬁc advanced manufacturing technology selection.
Respondents were asked to identify and discuss a recently adopted advanced
manufacturing technology used to fabricate, machine or assemble products at their plant
that was already used to make an initial production run. Using the classiﬁcation of AMT
drafted by the US Department of Commerce (US Commerce Department, 1989) and
categorized by other researchers (Small et al., 1997), such technology may include
CNC/DNC machines, material working lasers, pick and place or other robots, computer
integrated manufacturing (CIM) and ﬂexible manufacturing cells/systems (FMC/F MS).
Due to deﬁnitional issues with what constitutes CIM or FMS, respondents will only be
asked to discuss CNC/DNC, material working lasers or robots.

Technologies excluded from this research include design and engineering
technologies (e. g., computer aided design, computer aided process planning), automated
material handling technologies (e. g., automated storage and retrieval systems, automated
material handling systems), information technologies (e.g., material requirements

planning) and automated inspection and testing equipment.

31

3.2 Data Sample

Data will be gathered across a variety of industries from for-proﬁt United States
manufacturing ﬁrms that own and operate their own production facilities. An executive
or manager from manufacturing/industrial engineering, plant/operations engineering or
capital equipment purchasing will be targeted as the respondent. Ideally, multiple
respondents would be used at each plant. However, the cost and time of obtaining access
to multiple respondents much less securing the required response rate discourages the use
of multiple respondents at each plant in this dissertation research.
3.3 Required Sample Size

The AMTS-TPB has the most latent variables and thus will be used to calculate
required sample size. With eight latent variables, if it is conservatively assumed that 5
manifest variables are required for each latent variable, and that 10 sample points are
required for each manifest variable, then 400 responses would be required: (8)(5)(10) =
400. Assuming a 10% response rate, this would require distribution of 4,000
questionnaires.

A less conservative rule of thumb is 5 sample points for each manifest variable.
In the case with 5 manifest variables for each latent variable, 200 responses would be
required: (8)(5)(5) = 200. Assuming a 10% response rate, this would require distribution
of 2,000 questionnaires.

However, in academic journals such as Management Science, Decision Sciences,
Journal of Operations Management, and Administrative Sciences Quarterly, articles have
been published with just a single manifest variable for latent variables. To minimize

questionnaire length (in an attempt to increase the probability of a high response rate)

32

while still testing a rather complex model, this research will use three measurement
variables for some of the constructs. Therefore, with eight latent variables, three
measurement variables for each latent variable, and ﬁve sample points for each
measurement variable, 120 responses would be required: (8)(3)(5) = 120. Assuming a
10% response rate, this would require distribution of 1,200 questionnaires. Depending
upon available funding, a minimum of 2,000 questionnaires will be distributed to secure
the required number of responses.

3.4 Data Collection

A structured questionnaire was developed for this research. The majority of
questions are eleven point Likert sealed (0 to 10). A three-step survey collection process
similar to that recommended by Dillman (1978) will be adopted. Each survey will be
mailed to the contact person. The ﬁrst mailing will contain the questionnaire and an
explanatory cover letter. Aﬂer a two-week waiting period, reminders will be sent to all
non-respondents. Assuming the required sample size is not obtained aﬁer the ﬁrst two
contact attempts, a third mailing with the questionnaire and new cover letter will again be
distributed to all non-respondents.

The cover letter will explain the beneﬁts of participation and suggest that the
questionnaire itself identiﬁes a checklist of issues for respondent to consider when
making AMT selection decisions. Further, respondents will be offered an executive
summary of the results to increase the response rate.

3.5 Measures
This section discusses the measures used to assess each AMT adoption model.

The theoretical justiﬁcations for each model and sample measures used in prior

33

technology adoption research were previously presented. A copy of the actual

questionnaire appears in the appendix.

The measures for “Perceived Usefulness” are:

 

Relative to alternative technology or sourcing options, I had a high
degree of confidence that the AMT would:

Strongly Disagree
to Strongly Agree

 

Drive increased product innovations

Maximize production quality consistency
Lower fully burdened per unit production costs
Reduce overall production cycle times

 

 

 

The measures for “Familiarity with the AMT” are:

 

Relative to alternative technology or sourcing options, I had a high
degree of conﬁdence that the AMT would:

Strongly Disagree

 

Be highly compatible with existing operations
Be operated by current workers without providing additional training
Be easily adapted to existing operations

 

to Strongly Agree

 

 

The measures for “Supplier Support” are:

 

 

 

 

 

 

 

 

 

 

 

I had a high degree of conﬁdence that the recommended AMT Strongly Disagree

supplier would or could provide the desired: to Strongly Agree

AMT installation support

Training of operators, supervisors or managers

Detailed installation, operation and/or maintenance manuals

Our company and the recommended AMT supplier: Strongly Disagree
to Strorgly Agree

Had extensive experience working with each other

Were aware of each other’s competitive needs

The measures for “Normative Beliefs” are:

I believed that adoption of new AMT was favored by: Strongly Disagree

to Strongly Agree

 

Top management
Co-workers
Subordinates

 

 

34

 

The measures for “Attitude Towards Adopting” are:

 

Prior to actual adoption, I strongly believed that: Strongly Disagree
to Strongly Agree

 

Adoption of the AMT would make the plant much more competitive
The AMT would signiﬁcantly improve overall plant performance

 

 

 

The Measures for “Subjective Norms” are:

 

 

Opinions that inﬂuenced my position towards recommending the Strongly Disagree
AMT for adoption were directly or indirectly expressed by: to Strongly Agree
Top management

Co-workers

Subordinates

 

 

 

Measures for “Perceived Behavioral Control” are:

 

Prior to installation, I believed that the following resources would be Strongly Disagree
available to successfully install and/or use the AMT: to Strongly Agree

 

Technical resources and knowledge
Human resources with appropriate skills

 

 

 

 

Prior to actual adoption, I strongly believed that: Strongly Disagree

to Strongly Agree
All necessary resources would be available to successfully install and use the AMT

The AMT could be installed and operational within budget and time objectives

 

 

 

The measures for “Intention to Adopt AMT” are:

 

Based on all potential factors and overall evaluation of the AMT, I: Strongly Disagree

to Strongly Agree
Strongly recommended the AMT for adoption

Fully intended for the company to adopt the AMT
Provided well-documented support favoring adoption of the AMT

 

 

 

3.6 Proposed Data Analysis
The overall model ﬁt and speciﬁc hypotheses for each AMT adoption model will

be tested using Structural Equation Modeling (SEM) techniques. SEM is a

35

comprehensive multivariate statistical approach to test hypotheses about relationships
among measured (observed) and latent (not directly observed) variables. There are two
stages associated with this methodology. In the ﬁrst stage a measurement model will be
tested using conﬁrmatory factor analysis (CF A) to establish a satisfactory level of
reliability and validity. In the second stage, the degree to which the proposed models ﬁt

the observed data will be tested using EQS software with covariance matrices input.

36

Chapter 4
DATA ANALYSIS

4.] Introduction

This Chapter focuses on data analysis, including data collection, sample
characteristics, measurement model validation, fully speciﬁed structural model analysis
and associated hypotheses tests. Though some interpretation of the results is presented in
this Chapter, the detailed interpretation of results and recommendations for future
research are presented in Chapter 5.
4.2 The Data

This section is divided into two subsections. First, the data collection process is
described and the potential for non-response bias is assessed. Second, the characteristics
of the data and of the responding ﬁrms are discussed.
4. 2. I Data Collection

Industries most likely to adopt the AMTs focused on in this research were
identiﬁed by using inputs from the Society of Manufacturing Engineers (SME), National
Association of Purchasing Managers (N APM) and APICS (formerly the American
Production and Inventory Control Society). The industries included the research were
34XX (Fabricated Metal Products), 35XX (Industrial Machinery and Equipment), 36XX
(Electronic and Other Electrical Equipment) and 37XX (Transportation Equipment).

Data was collected in two stages, using techniques recommended by Dillman
(1978). In the ﬁrst stage, a questionnaire was sent to 2,370 contacts. After a two-week
waiting period, all non-respondents were sent a reminder postcard. These efforts resulted

in the responses presented in Table 4.

37

Table 4: Results of First Wave Mailing

 

 

 

 

 

 

 

 

 

 

 

Status/Response Frequency
Useable responses 52
Return to senderlcontact no longer at mailing address) 182
We do not respond to surveys / Please remove from mailing list 27

We are not a manufacturing ﬁrm 18

We do not invest in such technologies 6

We have not recently invested in such a technology 34
Questionnaire was too long 9

Did not have the time 5

All other non-respondents (no response/ feedback) 2037

 

 

 

Two observations from Table 4 highlight some of the challenges in conducting
this research. First, 14 (9 questionnaire was too long + 5 did not have the time) people
took the time to send back the questionnaire suggesting it was too long or would take too
much of their time. Second, 40 (34 we have not recently invested in such a technology +
6 we do not invest in such technologies) people took the time to indicate that they either
do not invest in such technologies or they have not invested in such technologies recently.
These challenges were anticipated when the research was proposed.

In total, 267 (182 return to sender + 27 we do not respond to surveys + 18 we are
not a manufacturing ﬁrm + 6 we do not invest in such technologies + 34 we have not
recently invested in such a technology) companies were not appropriate sample
candidates, 14 companies may have responded if the questionnaire were shorter, and 52
companies provided useable responses. The effective response rate for this ﬁrst wave
effort was 2.47% [52 / (2370-267)].

Besides the challenges to data collection already discussed, an attempt was made
to determine other factors impacting the response rate prior to sending out another survey

to all non-respondents. Non-respondents from the original sample were randomly

38

selected and contacted by phone until 70 people were actually contacted. Non-
respondents gave the reasons listed in Table 5 for not responding to the original survey.

Note that the total number of reasons exceeds 70, as some people cited more than one

 

 

 

 

 

 

 

 

 

 

reason.

Table 5: Results of Phone Contacts to Assess Non-Response Bias
Status/Response Frequency
Did not receive survey/ Did not recall receiving survey 9
We do not respond to surveys / Please remove from mailing list 12
We are not a manufacturing ﬁrm 1
We do not invest in such technologies 3
We have not recently invested in such a technology 16
Questionnaire was too long 27
Did not have the time 28
Receiving too many surveys 14

 

 

Table 5 suggests that 32 (12 we do not respond to surveys + 1 we are not a
manufacturing ﬁrm + 3 we do not invest in such technologies + 16 we have not recently
invested in such a technology) companies were not appropriate sample candidates.
Further, questionnaire length, the related lack of time and “survey overload” appeared to
be limiting the response rate.

Prior to a second mailing of the questionnaire to non-respondents, the
questionnaire was shortened by eliminating some demographic questions and other
questions not speciﬁc to the basic models under test. This shorter survey was sent to the
2,019 non-respondents (2,370 original sample - 52 useable responses from ﬁrst mailing -
267 invalid contacts identiﬁed in ﬁrst mailing- 32 invalid contacts identiﬁed in phone

contacts). The second mailing resulted in the responses listed in Table 6.

39

 

Table 6: Results of Second Wave Mailing

 

 

 

Status/Response Frequency
Useable responses 72
Return to sender (contact no longer at mailing address) 25

 

We do not respond to surveys / Please remove from mailing list

 

We are not a manufacturing ﬁrm

 

We do not invest in such technologies

 

 

Questionnaire was too long

 

7
0
3
We have not recently invested in such a technology 6
0
0

Did not have the time

 

 

 

 

All other non-respondents (no response/ feedback) 1906

 

Table 6 suggests that 41 (25 return to sender + 7 we do not respond to surveys + 3
we do not invest in such technologies + 6 we have not recently invested in such a
technology) companies were not appropriate sample candidates. The response rate for the
second mailing was 3.64% [72 / (2019 - 41)]. From the two mailings a total of 124
responses were received, yielding an effective response rate for the two mailings of
6.11% [(52 + 72) / (2019 - 41+ 52)].

Though the response rate was low, it was not altogether surprising given that not
all manufacturing ﬁrms will invest in the speciﬁc types of AMT focused on in this
research. Further, companies that do invest in such AMT may do so very infrequently
due to the long useful life of the AMT. These two points are somewhat supported by
Table 5, which indicates that 39.58% [(16 we have not recently invested in such a
technology + 3 we do not invest in such technologies) / (70 contacts - 9 did not receive
survey - 12 we do not respond to surveys - 1 we are not a manufacturing ﬁrm )] of the
manufacturing ﬁrms contacted by phone either do not invest in or had not recently

invested in such AMT.

40

The discussion above suggests that non-response bias may not be problematic in
this research. Regardless, a statistical test similar to that suggested by Armstrong and
Overton (1977) was conducted to further analyze potential non-response bias. The
method is premised on the assumption that late respondents typically reﬂect the attitudes
of non-respondents. Useable responses were segmented into quartiles based on date of
survey receipt. Means for 10 randomly selected questions were compared between the
ﬁrst 25% of responses and the last 25% of responses received as indicated in Table 7.
The t-tests indicated no signiﬁcant differences between responses. Coupled with the
information derived from phone contacts, the t-tests suggest that non-response bias may
be negligible in this study.

However, a conservative rule of thumb suggests that ten unique sample points
should be obtained for each manifest variable associated with each latent variable. As
will be discussed in Section 4.4, seventeen manifest variables were retained in the
AMTS-TPB after measurement model validation. This suggests that a total of 170
sample points (17 times 10) would be required under the conservative rule of thumb.
With 123 responses, there is concern of sampling bias. This issue is discussed in greater
detail in the limitations section of this analysis.

4.2.2 Characteristics of Responding Firms

Table 8 presents the standard industry classiﬁcation (SIC) code of respondents.

Note that the total exceeds 124 as many companies reported serving more than one

industry.

41

Table 7: T-tests for Equality of Means

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Question Quartile Mean Std. Sig.(2-tailed)
Dev.
Prior to actual adoption, 1 strongly believed that First (N=3 l) 8.48 1.86 Equal Var Assumed = 0.55
adoptionof the AMT would make the plant much more Last (N=3l) 8.23 1.48 Equal Var Not Assumed = 0.55
competitive
Relative to alternative technology or sourcing options, I First (N=3 l) 8.16 1.42 Equal Var Assumed = 0.26
had ? high degree F” conﬁdence "‘9‘ "’6 AMT ““0”“ Last (N=31) 8.52 1.03 Equal Var Not Assumed = 0.20
maximize production quality cons1stency
Relative to alternative technology or sourcing options. 1 First (N=3 l) 7.97 1.97 Equal Var Assumed = 0.29
“id 3 “‘14“ degree 0‘90"“969“ "’3‘ "1e AMT ““0"” be Last (N=31) 7.39 2.33 Equal Var Not Assumed = 0.30
highly compatible wrth crusting operations
Relative to alternative technology or sourcing options, I First (N=31) 8.48 1.48 Equal Var Assumed = 0.12
had a high degree of confidence that the AMT would Last (N=31) 7‘74 2.” Equal Var Not Assumed = (“2
reduce overall production cycle times
Based on all potential factors and overall evaluation of First (N=31) 8.67 1.08 Equal Var Assumed = 0.09
the AMT’ ' “mg" 'ewmmcnded "‘e AMT ‘0' Last (N=31) 8.19 1.1 1 Equal Var Not Assumed = 0.09
adoption
I believed that adoption of new AMT was favored by co- First (N=31) 8.13 1.88 Equal Var Assumed = 0.49
“Ike's Last (N=31) 7.81 1.80 Equal Var Not Assumed = 0.49
I had a high degree of confidence that the recommended First (N=3 l) 7.65 2.04 Equal Var Assumed = 0.32
AMT sul’ph‘” ””9”” 0' “09m ”0““ "’9 des'red Last (N=3l) 8.10 1.49 Equal Var Not Assumed = 0.32
detailed installation, operation and/or maintenance
manuals
Opinions that inﬂuenced my position towards First (N=31) 6.97 2.71 Equal Var Assumed = 0.73
T°°9mm°"d'"g "‘6 AMT f‘” 380mm" we"? d'rcc‘” °’ Last (N=3 I) 6.74 2.38 Equal Var Not Assumed = 0.73
indirectly expressed by subordinates
Prior to installation, I believed that the following First (N=31) 7.55 1.91 Equal Var Assumed = 0.89
resources would be available to successfully install
. . N: 7.4 .7 E 1V N tA = .89
and/or use the AMT: Human resources With appropriate L85! ( 3]) 8 1 5 qua ar 0 ssumed 0
skills
Our company and the recommended AMT supplier were First (N=31) 6.45 2.62 Equal Var Assumed = 0.20
“’8'“ 0f “Ch 0”“ 5 “0mm“ "feds Last (N=31) 7.23 2.22 Equal Var Not Assumed = 0.20
Table 8: SIC of Respondents

SIC Description Frequency
34XX Fabricated Metal Products 33
35XX Industrial Machinery and Equipment 37
36XX Electronic and Other Electrical Equipment 40
37XX Transportation Equipment 42

 

 

 

42

 

 

Table 9 presents the geographic location of respondents. Note that the total is less
than 124. Respondents were asked to attach their business card rather than having them
ﬁll in address information on the questionnaire. In some cases the card was not included.
In cases where the card was not included, the postal cancellation stamp on the return
envelope was used to determine location. Unfortunately however, the ﬁrst set of return
envelopes was improperly printed such that there was no postal cancellation code on
some of the return envelopes. Table 8 and Table 9 suggest that cross-industry and cross-
location data were obtained.

Table 9: Geographic Location of Respondents

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Location Frequency Location Frequency
Michigan 12 Delaware 2
California 9 Georgia 2
Wisconsin 9 Illinois 2
Ohio 8 Kansas 2
Arizona 5 Kentucky 2
Indiana 5 Louisiana 2
Pennsylvania 5 Missouri 2
Massachusetts 4 Nebraska 2
Minnesota 4 North Carolina 2
Texas 4 Utah 2
Iowa 3 Arkansas 1
New Jersey 3 Alabama 1
Vermont 3 Maryland 1
Virginia 3 New Mexico 1
Colorado 2 Oklahoma 1
Connecticut 2 Tennessee 1

Total 107

 

 

 

 

 

Table 10 summarizes the title of respondents. The majority of respondents were
directors or managers in operations, capital equipment planning or manufacturing

engineering. Based on ﬁeld interviews prior to distribution of the survey and inputs from

43

the SME, APICS and NAPM, people in such positions were estimated to be the key

informants and decision-makers in this research context.

Table 10: Title of Respondents

 

 

 

 

 

 

 

 

 

 

Title Frequency

Director or Manager of Manufacturing/Operations 58
Director or Manager of Capital Equipment Planning/Procurement 31
Director or Manager of Manufacturing Engineering 19
Vice President of Manufacturing/Operations 9
President, Vice President or General Manager 5
Director of Finance/Comorate Strategic Technology Management 1
Director of Quality Assurance 1
Total 1 24

 

 

 

The research focused on three speciﬁc types of AMT: CNC/DNC, robots and
material working lasers. Table 11 presents the type of AMT reported by respondents.
The classiﬁcations must be viewed with caution as in some cases the AMT identiﬁed
could have been classiﬁed in multiple categories (e. g., a “robotic laser welder” could have
been classiﬁed as either “robotics” or as a “material working laser.”) The key for this
research is that all of the AMTs assemble, machine or fabricate materials or products.

Table 11: Type of AMT Reported

 

 

 

 

 

 

Title Frequency
CNC/DNC 79
Robotics 33
Material working laser 12
Total I 24

 

 

 

Table 12 presents some data characterizing the AMT investments, while Figure 7
to Figure 12 present boxplots of each characteristic. As can be seen in the boxplots, the
mean and standard deviation of each characteristic is impacted by outliers. If a company

was identiﬁed as an outlier to the responses in Table 12, that company’s responses to a set

44

of other randomly selected questions were compared to the mean response for the

questions. This comparison suggested that there were no consistent differences between

respondents based on the characteristics identiﬁed in Table 12.

Table 12: AMT Characteristics

 

 

 

 

 

 

 

 

 

 

 

Category Mean Std. Dev. N
Purchase and installation cost of AMT $869,787 $1,072,422 120
Projected ROI for the AMT investment 56.78% 77.39% 60
ngected payback period in months for the AMT 20.38 12.06 52
Expected useﬁil life in years of the 8.47 4.51 75
Installation and debug time in months 2.97 2.63 74
Process optimization time in months 4.57 4.39 74

 

8000000

 

6000000-

Dollars 4000000

2000000

 

-2000000

*101

C31

0107

 

 

 

 

 

N:

120

Figure 7: Purchase and Installation Cost of the AMT

 

Percentage

Months

‘4
<3

nnn

 

 

 

500- *4

400-

300- area

200 l N81

100- a;

0. E-
-100 _
N = 60

Figure 8: Projected ROI for the AMT Investment

 

601

4o-

30-

20-

 

055

 

 

Figure 9: Projected Payback Period in Months

46

 

 

Years

Months

on

 

av

20 - our

 

 

 

 

 

 

 

 

 

10-
0 i
N = 75
Figure 10: Expected Useful Life in Years
15
*33
14-
12 0:2
10-
0114
8 .
6 l
4 I
2 a
0 I
N = 74

Figure 11: Installation and Debug Time in Months

47

30

 

 

 

 

 

«x91

*33
20 -

C30

all!
10 '
Months
0 .
-10 .
N = 74

 

 

Figure 12: Process Optimization Time in Months

Table 12 suggests that many companies expect remarkable ﬁnancial retums from
the AMT. Such returns would suggest that all companies should invest in such AMTs.
However, one survey question asked respondents to indicate if the decision to invest in
the AMT was based primarily on ﬁnancial analysis such as ROI estimates (end scale = 0),
balanced (mid-scale =5), or strategic factors such as competitive positioning (end-scale =
10). Based on 74 responses, the mean score was 5.28, and the median and mode were
both 5.0. This suggests that ﬁnancial return is not enough to justify investments. The

investment may also need to be aligned with technology strategy.

48

The average expected useful life of the AMT was over eight years. This provides
further indication that such AMT investments are made relatively infrequently, a factor
that may have impacted response rate as previously discussed.

“Installation and debug time” was deﬁned as:

“The time from receipt of equipment to being able to power the equipment

‘without blowing anything up’ and with the equipment ‘basically moving

as expected’ under operator control.”

“Process optimization time” was deﬁned as:

“Time after installation and debug to integrate the AMT with the

production system. This time may include integration with automated

material handling systems, operator training and a pilot run for example.

This is the time after which there was a high degree of conﬁdence that full

scale production with an “acceptable” reject rate could begin.”

These two measures of time indicate that on average it takes a total of
approximately 7.5 months aﬁer receipt of the AMT to get the AMT “production ready.”
It appears that on average, AMT adoption requires substantial time and ﬁnancial
resources.

4.3 Data Analysis

Two models were tested in this research, loosely following an approach similar to
that of Davis et al. (1989). One model was labeled “AMTAM” and was based directly on
TAM. The other model was labeled “AMTS-TPB” and was based directly on the TPB.
The CF A and fully speciﬁed model tests for each research model are presented in

separate sections below. However, the descriptive statistics are ﬁrst presented in the

following section.

49

4. 3. 1 Descriptive Statistics

Table 13 presents the descriptive statistics for each measurement variable
originally proposed. The means are shifted toward the high end of the scale. This was
expected since respondents were asked to discuss an AMT they had actually adopted.
The high end of the scale typically reﬂected a positive evaluation of the AMT decision
criteria.

Since univariate normality is necessary though not sufﬁcient evidence of
multivariate normality (West et al., 1995), it is advisable to examine the univariate
distributions. Several variables in Table 13 have values for skewness greater than 1.5
and/or values for kurtosis greater than 2.5. Rules of thumb suggest departure from
normality may be present when univariate skewness is greater than or equal to 2.0 and/or
univariate kurtosis is greater than or equal to 7.0 (Chou and Bentler, 1995).

Multivariate normality was assessed using Bentler’s (1997) EQS software which
calculates Mardia’s coefﬁcient. Mardia’s coefﬁcient for the AMTS-TPB and the
AMTAM were 65.22 and 35.77 respectively. Though there is no consensus on distinct
cut-off levels for Mardia’s coefﬁcient, ﬁgures of the magnitude calculated here suggest a
potential departure from multivariate normality. Examination of potential multivariate
outliers identiﬁed by EQS suggested that one case made a relatively substantial
contribution to multivariate kurtosis. Removing this case decreased Mardia’s coefﬁcient
by approximately 6% in the AMTS-TPB and 10% in the AMTAM. Deleting any other
cases identiﬁed by EQS made marginal contribution to reduction of Mardia’s coefﬁcient.
Thus, structural equation modeling analysis was performed on 123 out of the 124 cases

using Generalized Least Squares (GLS) estimation.

50

Table 13: Descriptive Statistics for Measurement Variables

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Questions Proposed for Each Factor Std. Skew- Kur-
Mean Dev. ness tosis Label

Perceived Usefulness F1
Relative to alternative technology or sourcing options. 1 had a high degree of 6.31 2.44 -0.49 «0.24 -
conﬁdence that the AMT would drive increased moduct innovations
Relative to alternative technology or sourcing options. 1 had a high degree of 8.54 1.22 -1.19 2.85 V1
conﬁdence that the AMT would maximize production quality consistency
Relative to alternative technology or sourcing options, I had a high degree of 7.21 2.15 -0.97 0.71 V2
confidence that the AMT would lower fully burdened ger unit production costs
Relative to altemative technology or sourcing options, 1 had a high degree of 8.02 1.90 -l .19 1.23 V3
conﬁdence that the AMT would reduce overallproduction cycle times
Familiarity with AMT F2
Relative to altemative technology or sourcing options. 1 had a high degree of 7.41 2.30 -0.89 023 V4
conﬁdence that the AMT would be highly compatible with existing operations
1 had a high degree ofconﬁdeiice that the AMT would be operated by current 3.95 3.1 1 0.36 -1.l9 --
workers without providing additional training
Relative to alternative technology or sourcing options, 1 had a high degree of 6.72 2.45 -0.65 036 V5
conﬁdence that the AMT would be easily adapted to existing operations
Supplier Simport F3
1 had a high degree of conﬁdence that the recommended AMT supplier would or 7.98 1.81 -1.51 2.78 -
could provide the desired AMT installation support
1 had a high degree of confidence that the recommended AMT supplier would or 7.72 2.06 -1.64 2.60 V6
could provide the desired training of operators, supervisors or managers
1 had a high degree of conﬁdence that the recommended AMT supplier would or 7.72 2.07 -1.55 2.52 ~-
could provide the desired detailed installation, operation, maintenance manuals
Our company and the recommended AMT supplier had extensive experience 5.07 3.06 -0.05 -1.08 --
working with each other
Our company and the recommended AMT supplier were aware of each other’s 6.25 2.71 -0.71 -0.27 V7
competitive needs
Attitude towards AdoLting F4
Prior to actual adoption, 1 strongly believed that adoption of the AMT would 8.41 1.61 -1.50 3.41 V8
make the plant much more competitive
Prior to actual adoption, 1 strongly believed that adoption of the AMT would 8.28 1.50 -1.10 1.14 V9
signiﬁcantly improve overall plant performance
Perceived Behavioral Control F5
Prior to installation, I believed that the following resources would be available to 7.91 1.53 -0.97 1.36 --
successfully install and/or use the AMT: Technical resources and knowledge
Prior to installation. I believed that the following resources would be available to 7.28 2.02 -1.08 1.30 V1 I
successfully install and/or use the AMT: Human resources with appropriate skills
Prior to actual adoption, 1 strongly believed that all necessary resources would be 7.69 1.96 -1.21 1.01 V10
available to successfully install and use the AMT
Prior to actual adoption. 1 strongly believed that the AMT could be installed and 8.18 1.49 -1.18 1.96 -
operational within budget and time objectives
Intention to Adopt AMT F6
Based on all potential factors and overall evaluation of the AMT, 1 strongly 8.40 1.18 -0.65 0.56 --
recommended the AMT for adoption
Based on all potential factors and overall evaluation of the AMT, 1 fully intended 8.53 1.49 -1.36 1.75 V12
for the company to adopt the AMT
Based on all potential factors and overall evaluation of the AMT, 1 provided well- 7.73 2.12 -1.27 1.49 V13
documented support favoring adoption of the AMT
Normative Beliefs # F7
I believed that adoption of new AMT was favored by top management 8.52 1.37 -1.15 1.20 --
I believed that adoption of new AMT was favored by co-workers 7.89 1.69 -1.02 1.40 V14
I believed that adoption of new AMT was favored by subordinates 7.37 1.97 -0.51 -0.45 V15
Subjective Norms # F8
Opinions that inﬂuenced my position towards recommending the AMT for 7.93 1.98 -1.41 2.33 -—
adoption were directly or indirectly expressed by top mangement
Opinions that inﬂuenced my position towards recommending the AMT for 7.19 1.95 —0.97 1.29 V16
adoption were directly or indirectly expressed by co-workers
Opinions that inﬂuenced my position towards recommending the AMT for 6.33 2.72 -0.68 0.36 V17

adoption were directly or indirectly expressed by subordinates

 

 

 

 

 

 

Shaded areas represent latent factors
~ Indicates measure was dropped during measurement model validation
# Indicates latent factor speciﬁc to the AMTS-TPB

51

 

4.4 Structural Analysis

This research applies fairly well established underlying theoretical frameworks in
a relatively new context. The AMTS-TPB and the AMTAM were therefore developed a
priori based on the underlying frameworks. Thus, a signiﬁcance level of 0.05 was used
for statistical tests. The structural equation modeling software package EQS Version 5.7a
(Bentler, 1992) was used.

Table 13 presented the complete list of measurement variables originally
proposed. Measurement variables that are not labeled were deleted during the model
validation process. The majority of deleted variables were removed due to lack of
convergent or discriminant validity in one or both of the models. The two measurement
variables pertaining to top management “input” and “inﬂuence” were deleted because
several respondents indicated that they were “top management.” Including such measures
in their respective factors would be invalid because the two factors were intended to
measure only external inﬂuences on the respondent’s decision-making process.

The two latent variables ﬂagged with the pound sign (ii) in Table 13 are speciﬁc
to the AMTS-TPB, as are the measurement variables associated with these latent
variables. However, the measurement variables that were retained aﬁer measurement
model validation for the latent variables common to both models are identical. This is an

important step for making model comparisons.

52

4. 4.1 Advanced Manufacturing Technology Adoption Model (AM T AM)
The AMTAM (reference Figure 13) represents an application of TAM in an
industrial context that focuses on AMT adoption. The measurement model and fully

speciﬁed model for AMTAM are examined below.

  

H6-AMTAM

  
       
 

   
 

F4: Attitude
Towards Adopting

F1 : Perceived H 1 'AMTAM
Usefulness

  
   

F6: Intention to
Adopt AMT

 

 

  

H4-AMTAM

HS-AMTAM

  

H2-AMTAM

 

F2: Familiarity
with AMT

    
 
 

F5: Perceived
Behavioral Control

F 3: Supplier
Support

   

  

H3-AMTAM

Figure 13: Advanced Manufacturing Technology Adoption Model (AMTAM)

4. 4.1.] Measurement Validation for AMTAM

The ﬁnal AMTAM measurement model is shown in Figure 14. Table 13
presented the questions associated with each variable. All latent variables were free to
covary. A11 factor loadings are signiﬁcant a p = 0.05 or better. The chi-square test is not
signiﬁcant, and the ratio of chi-square value to degrees of freedom (normed chi-square) is
between the stringent range of 1.0 to 2.0. The non-normed ﬁt index (NNFI) and the

comparative ﬁt index exceed the widely held rule of thumb value of 0.90.

53

 
     
 
     
  
   
   
 

Fl: Perceived
Usefulness

F2: Familiarity
with AMT

F3: Supplier
Support

 

° Standardized solution shown
' All latent variables were free to covary
- All loadings are signiﬁcant at 0.05

 

 

 

    

F 4 Attitude
Towards Adopting

 
   

F 5: Perceived
Behavioral Control

0.67

  
   

 

0.59 0.85

V12 V_3

F6: Intention to
Adopt AMT

 

 

x2: 50.630 / 50 d.f. / p = 0.449
12:/d.f.: 1.013

NNFI: 0.987

 

C Fl: 0.992

 

 

Figure 14: AMTAM Measurement Model

As Bagozzi and Yi (1988) indicate, if either or both the chi-square test and ﬁt

indices point to a satisfactory model, the best that can be said is that the criteria tend to

support the model. Both individual parameter estimates and theory should be used in

conjunction with the analysis of overall ﬁt to determine if the data ﬁt the model.

The Lagrange Multiplier (LM) Test was used to identify potential improvements

in model ﬁt resulting from measurement model respeciﬁcation. The LM test identiﬁed

one set of error terms that could be associated to improve model ﬁt. However, no

respeciﬁcation was made because it could not be theoretically supported. Further, the

suggested association would only improve overall model ﬁt marginally (albeit

signiﬁcantly in a statistical sense.)

54

Based on the statistically signiﬁcant parameter estimates, overall model ﬁt criteria
and theory, the measurement model was judged acceptable. The following section
examines the hypothesized relationships between factors.
4.4.1.2 Fully Specified Model for AMT AM

The fully speciﬁed AMTAM is shown in Figure 15. The chi-square test is not
signiﬁcant, and the normed chi-square value is between a stringent range from 1.0 to 2.0.
The non-normed ﬁt index and the comparative ﬁt index are below the widely held rule of
thumb cut-off value of 0.90. However, as Bollen (1989) discusses, this rough guideline
must be assessed under consideration of a number of other factors including standards set
by previous work. As an example, he suggests that if existing models have ﬁt indices
below 8.0, values of 0.85 or higher can represent an important improvement over existing
work. Given that a relatively new model and measures were used in this research, the ﬁt
indices are considered acceptable though not ideal.

The LM test identiﬁed three sets of error terms that could be associated to
improve model ﬁt, but no such associations were made because none could be justiﬁed.
The LM test also suggested the association of F2 (familiarity with the technology) and F6
(intention). This association may make nomonological sense. One manufacturing
manager suggested that investments in AMT simply reﬂect a long-term strategy such that
previous AMT investments (prior behavior) in many cases establishes the pattern for new
AMT investments (current behavior intention.) However, this modiﬁcation was not made
as it only marginally improved model ﬁt. Nonetheless, the suggested relationship should

be examined in greater detail in future research.

55

   
 

   
    

    
      
    

 
    

F4: Attitude
Towards Adopting

F6: Intention to
Adopt AMT

F1: Perceived
Usefulness

 

 

 

F2: Familiarity
with AMT

 
   
  

F5: Perceived

 

  
 
  

   
 

  

 

 

Behavioral Control 7. _
F32 Supplier x . 66.544 / 57 d..f /p - 0.182
supp” -051 12:/dig 1.167
- Standardized solution shown NNFII 0.832
- * Signiﬁcant at p = 0.05 or better
° All exogenous factors allowed to covary CF11 0.877

 

 

 

 

 

Figure 15: AMTAM Fully Speciﬁed Model
Table 14 summarizes the test of the hypotheses proposed in Figure 13. Four of
the seven hypotheses are supported. Overall variance explained in the “Intention to
Adopt AMT” construct (F6) is 0.54.

Table 14: AMTAM Hypotheses Test Results

 

 

 

 

 

 

 

 

 

Hypothesis Relationship Test Result
H1 F1 — F4 Supported
H2 F2 - F5 Supported
H3 F3 — F5 Not Supported
H4 F5 — F1 Supported
H5 F5 — F4 Not Supported
H6 F1 — F6 Supported
H7 F4 — F6 Not Supported

 

 

 

 

56

A complete assessment of ﬁt must consider the stage of theory development in the
ﬁeld, parameter estimates, overall model ﬁt metrics and theory. Based on all these

criteria, it is judged that the data ﬁts the model although the ﬁt is not ideal.

4. 4.2 Advanced Manufacturing T echnology Selection Model (AM TS- T PB)
The AMTS-TPB (reference Figure 16) represents an application of the TPB in an
industrial context that focuses on AMT adoption. The measurement model and ﬁilly

speciﬁed model for AMTS-TPB are examined below.

   
     
   

F4: Attitude
Towards Adopting

F1: Perceived
Usefulness

  

    

HS-TPB

  

HZ-TPB

F2: Familiarity
with AMT

    
  

       

H6-TPB

    
 

F6: Intention to
Adopt AMT

F 5: Perceived
Behavioral Control

      

F3: Supplier
Support

   

H3-TPB

  

H7-TPB

     

F 7: Normative H4-TPB

Beliefs

   

 

F8: Subjective
Norms

       

Figure 16: Advanced Manufacturing Technology Selection Model (AMTS-TPB)

57

4.4.2.1 Measurement Validation for AM T S-T PB

The ﬁnal AMTS-TPB measurement model is shown in Figure 17. Table 13
presented the questions associated with each variable. All latent variables were free to
covary. All factor loadings are signiﬁcant a p = 0.05 or better. The chi-square test is not
signiﬁcant, and the ratio of the chi-square value to degrees of freedom (normed chi-
square) is between the stringent range of 1.0 to 2.0. The non-normed ﬁt index (NNFI)
and the comparative ﬁt index exceed the widely held rule of thumb value of 0.90.

As Bagozzi and Yi (1988) indicate, if either or both the chi-square test and ﬁt
indices point to a satisfactory model, the best that can be said is that the criteria tend to
support the model. Both individual parameter estimates and theory should be used in
conjunction with the analysis of overall ﬁt to determine if the data ﬁt the model.

The Lagrange Multiplier (LM) Test was used to identify potential improvements
in model ﬁt resulting from measurement model respeciﬁcation. The LM test identiﬁed
four sets of error terms that could be associated to improve model ﬁt. However, no
suggested respeciﬁcation was made because none could be theoretically supported.
Further, the suggested associations would only improve overall model ﬁt marginally
(albeit signiﬁcantly in a statistical sense.)

Based on the statistically signiﬁcant parameter estimates, overall model ﬁt criteria
and theory, the measurement model was judged acceptable. The following section

examines the hypothesized relationships between factors.

58

 

    

     

F4: Attitude
Towards Adopting

 
 

F1: Perceived
Usefulness

0.74 0.82

    
  
 
 
  
 

  

  
  

F5: Perceived
Behavioral Control

     
 
 

F2: Familiari
with AMTty F6: Intention to

Adopt AMT

 
   

 
 

F3: Supplier
Support

  
 
 
 

‘ 0.63

F8: Subjective

  

F7: Normative
Beliefs

  

 

     
   

 

 

 

Norms
12: 95.254 / 91 d.f. / p = 0.359
12:/d.f.: 1.047
0.49
'1 Standardized solution shown V17 NNFI: 0.936
0 All latent variables were free to covary
- All loadings are signiﬁcant at 0.05 CFI: 0.957

 

 

 

 

 

Figure 17: AMTS-TPB Measurement Model

4.4.2.2 Fully Speciﬁed Model for AM T S-T PB

The fully speciﬁed AMTS-TPB is shown in Figure 18. The chi-square test is not
signiﬁcant, and the normed chi-square value is between a stringent range from 1.0 to 2.0.
The non-normed ﬁt index and the comparative ﬁt index are below the widely held rule of
thumb cut-off value of 0.90. However, as Bollen (1989) discusses, this rough guideline
must be assessed under consideration of a number of other factors including standards set
by previous work. As an example, he suggests that if existing models have ﬁt indices

below 8.0, values of 0.85 or higher can represent an important improvement

59

   
  
    
 

  
  

F 4: Attitude
Towards Adopting

  

F1: Perceived
Usefulness

      
 
     

F2: Familiarity 1.19 * 045 0.30

with AMT

 
  
     

 
   
   

0.03

  

F5: Perceived
Behavioral Control

! 0.62 0.69 ,,,

F6: Intention to

. Adopt AMT
F3: Supplier

Support

         

 

       

 

F7: Normative 0.79 * F8: Subjective
Bends Norms 762 121.185 / 106 d.f. / p = 0.149
xzz/df: 1.143
° Standardized solution shown NNFII 0-805
0 * Signiﬁcant at p = 0.05 or better
- A11 exogenous factors allowed to covary CF13 0-848

 

 

 

 

 

 

Figure 18: AMTS-TPB Fully Specified Model

over existing work. Given that a relatively new model and measures were used in this
research, the ﬁt indices are considered acceptable though not ideal.

The LM test identiﬁed seven sets of error terms that could be associated to
improve model ﬁt, but no such associations were made because none were theoretically
justiﬁable. Further, none of the suggested relations would improve model ﬁt

substantially.

60

Table 15 summarizes the test of the hypotheses proposed in Figure 16. Five of the
seven hypotheses are supported. Overall variance explained in the intention to adopt

construct (F6) is 0.91.

Table 15: AMTS-TPB Hypotheses Test Results

 

 

 

 

 

 

 

 

 

Hypothesis Relationship Test Result
H1 F1 - F4 Supported
H2 F2 — F5 Supported
H3 F3 — F5 Not Supported
H4 F7 — F8 Supported
H5 F4 - F6 Supported
H6 F5 - F6 Not Supported
H7 F8 — F6 Supported

 

 

 

 

A complete assessment of ﬁt must consider the stage of theory development in the
ﬁeld, parameter estimates, overall model ﬁt metrics and theory. Based on all these
criteria, it is judged that the data ﬁts the model although the ﬁt is not ideal.

4.5 Summary

Two technology adoption models were developed and tested. The measurement
model for each was judged satisfactory using conservative guidelines of analysis. The
data was also judged to ﬁt each fully speciﬁed model satisfactorily given the relatively
new application of the research frameworks to the research context.

Table 16 summarizes the hypothesis tests for each model. Note that the ﬁrst four
hypotheses are identical in each model (except for labeling.) Three of these four
hypothesis tests were in agreement, while the attitude (F4) to intention (F6) results were
different. The next three hypotheses are speciﬁc to the AMTS-TPB, while the next three
hypotheses are speciﬁc to the AMTAM. Finally, totally variance explained in the

intention to adopt the AMT (F6) was 0.55 in AMTAM and 0.91 in AMTS-TPB.

61

Chapter 5 will discuss in greater detail how these ﬁndings relate to the existing
research and present potential explanations for the ﬁndings. Brieﬂy however, Table 16
suggests that the TPB may provide a useful underlying framework for future research into
AMT selection decisions, as the ﬁndings of the AMTS-TPB were generally consistent
with the TPB framework. The most notable exception is that H6-TPB (perceived
behavioral control to intention relationship) was not supported. While this is consistent
with the work of Lee (1990) that examined intentions to conduct international technology
transfer (clearly a different research context than the context herein), it is inconsistent
with the TPB in general.

Table 16 also suggests that the TAM developed by Davis et al., (1989) may be
extended into research contexts other than that for which it was intended, as the AMTAM
developed herein is generally consistent with the framework suggested by TAM. The
most notable exception is that while H7-TPB is consistent with the general TPB model, it
is inconsistent with the ﬁndings of Davis et al. (1989) in terms of TAM. This difference
is attributed to the different research contexts, as the research herein focuses on a much
less “personal” decision than the research of Davis et al. (1989). In the context herein,
one might expect subjective norms to have a signiﬁcant inﬂuence on behavioral

intentions. This and other evaluations are expanded on in the following chapter.

62

Table 16: Summary of Hypotheses for Both Models

 

 

 

 

 

 

 

 

 

 

 

 

Stated Hypothesis Label Test Result
Perceived usefulness of the AMT directly inﬂuences Hl-AMTAM Supported
a decision maker’s attitude toward adopting the Hl-TPB Supported
AMT.
Familiarity with AMT inﬂuences a decision-maker’s H2-AMTAM Supported
perceived behavioral control over successfully H2-TPB Supported
adopting the AMT
Supplier support inﬂuences a decision-maker’s H3-AMTAM Not Supported
perceived behavioral control over successfully H3-TPB Not Supported
adopting the AMT.
Attitude toward adopting the AMT directly H7-AMTAM Not Supported
inﬂuences a decision-maker’s behavioral intention to HS-TPB Supported
adopt the AMT.
Normative beliefs towards AMT selection inﬂuence H4-TPB Supported
a decision-maker’s subjective norms towards AMT
adoption.
Perceived behavioral control over adopting the AMT H6-TPB Not Supported
directly inﬂuences a decision-maker’s behavioral
intention to adopt the AMT.
Subjective norms toward adopting the AMT directly H7-TPB Supported
inﬂuence a decision-maker’s behavioral intention to
adopt the AMT.
Perceived Behavioral Control directly inﬂuences the H4—AMTAM Supported
Perceived Usefulness of the AMT.
Perceived Behavioral Control directly inﬂuences H5-AMTAM Not Supported
Attitude Toward Adopting the AMT.
Perceived Useﬁilness directly inﬂuences Intention to H6-AMTAM Supported

Adopt the AMT

 

 

 

63

 

Chapter 5
DISCUSSION OF RESULTS

5.1 Overview and Chapter Contents

This chapter interprets the data analysis in the preceding chapter. First, in Section
5.2 the hypotheses are discussed and interpreted using open-ended questions from the
survey instrument, previous research, and follow-up interviews. The section ﬁrst
examines the ﬁndings relative to existing research, then presents potential explanations
for the convergence or divergence of the ﬁndings. Next, in Section 5.3 the academic and
managerial contributions are examined. Then, in Section 5.4 the limitations of the
research are discussed. Finally, in Section 5.5 recommendations for ﬁiture research are
presented.
5.2 Discussion of Hypotheses Tests
5. 2.1 Hypotheses Common to the Two Models

The four hypotheses that were common to the two models are examined in this
section. First, both Hl-AMTAM and Hl-TPB proposed that:

Perceived usefulness of the AMT directly inﬂuences a decision-maker’s
attitude toward adopting the AMT.

The hypothesis was supported in both models. This is consistent with the general TPB
(Ajzen, 1991) which suggests that beliefs and evaluations about the consequences of
behavior inﬂuence an individual’s favorable or unfavorable objective impression of (i.e.,
attitudes towards) a speciﬁed behavior. The results are also consistent with the TAM
(Davis et al., 1989) and the IT AM (Lee, 1990), which examined technology adoption

decisions in different contexts. This suggests that the “usefulness-attitude” relationship

64

warrants continued examination, both in terms of direct and indirect inﬂuences on
intentions and in terms of further identifying measures and determinants of “usefulness.”

Second, both H2-AMTAM and H2-TPB proposed that:

Familiarity with AMT inﬂuences a decision-maker’s perceived behavioral
control over successfully adopting the AMT.

The hypothesis was supported in both models. This is consistent with the general TPB
(Ajzen, 1991) which suggests that self-efﬁcacy inﬂuences the perceived difﬁculty or ease
of performing the speciﬁed behavior. This is also consistent with technology transfer
research (Robinson, 1988; Rogers, 1983) which suggests that prior adoption of new
technology establishes a level of conﬁdence and knowledge that builds conﬁdence for
future technology adoption. No such relationship was proposed in TAM.

In Lee’s (1990) ITAM research, an insigniﬁcant relationship was found between
“prior experience with a technology” and “perceived ease of technology application”
which is contrary to the ﬁndings herein. The conﬂicting ﬁndings may be due to the
different research contexts. The research herein examined speciﬁc types of
manufacturing technology whereas ITAM examined international technology transfer in
general. Perhaps the increased number of factors (e.g., political, socio-economic) in
international technology transfer are more dynamic than the factors in AMT adoption
decisions. The ﬁndings suggest that continued investigation of the direct or indirect
impact of adaptability and compatibility of technology on intentions, as well as the
contexts under which familiarity with technology has an inﬂuence are warranted.

Third, both H3-AMTAM and H3-TPB proposed that:

Supplier support inﬂuences a decision-maker’s perceived behavioral
control over successfully adopting the AMT.

65

The hypothesis was not supported in either model. This is not consistent with the general
TPB (Ajzen, 1991) which suggests that resource facilitating conditions such as external
agency support inﬂuences the perceived difﬁculty or ease of performing the speciﬁed
behavior. However, it is consistent with the ﬁnding by Lee (1990) who also found an
insigniﬁcant relationship between supplier support and perceived ease of use.

Finally, both H7-AMTAM and HS-TPB proposed that:

Attitude toward adopting the AMT directly inﬂuences a decision-maker’s
behavioral intention to adopt the AMT.

The hypothesis was supported in the case of the AMTS-TPB. The signiﬁcant relationship
is consistent with the general TPB (Ajzen, 1991) which suggests that an individual’s
favorable or unfavorable objective impression of a speciﬁed behavior directly inﬂuences
intention to perform that behavior. It is also consistent with the ﬁndings of Davis et al.
(1989) in terms of their dim application of the TRA to examine software application
usage. The signiﬁcant relationship is also consistent with Lee’s (1990) research.

The hypothesis was not supported in the case of the AMTAM. The non-
signiﬁcant relationship is consistent with the ﬁndings of Davis et al. (1989) in terms of
their adaptation of the TRA in the form of the TAM. Davis et al. (1989) found that
initially (i.e., prior to actual adoption) attitude had a statistically signiﬁcant (albeit not
substantial) inﬂuence on behavioral intention. However, after using the information
system for a period (i.e., gaining familiarity with the technology), attitude had a

statistically non-signiﬁcant relationship with behavioral intention.

66

5.2.2 Hypotheses Specific to AM T S- T PB
The three hypotheses speciﬁc to the AMTS-TPB are examined in this section.
The research found support for the following hypothesis speciﬁc to the AMTS-TPB:

H4-TPB: Normative beliefs towards AMT selection inﬂuence a decision-
maker’s subjective norms towards AMT adoption.

This ﬁnding is consistent with the general TPB (Ajzen, 1991) which suggests that
perceived expectations of a referent individual or group inﬂuences perceived
social pressures to engage in or refrain from a speciﬁc behavior. Neither Davis et
al. (1989) nor Lee (1990) explicitly examined the Normative Belief — Subjective
Norm relationship in their research.

The research found support for the following hypothesis speciﬁc to the AMTS-
TPB.

H7-TPB: Subjective norms toward adopting the AMT directly inﬂuence a
decision-maker’s behavioral intention to adopt the AMT.

This ﬁnding is consistent with the general TPB (Ajzen, 1991) which suggests that
perceived social pressures to engage in or refrain from a speciﬁc behavior
inﬂuence intentions to perform that behavior. However, this ﬁnding is not
consistent with the ﬁndings of Davis et al. (1989) in terms of their d_irgc_t
application of the TRA to examine software application usage.

Perhaps the difference in ﬁndings is due to the cross-functional nature of
the AMT decision-making process. One optional question in the survey asked
respondents to rate extent of agreement (0 = strongly disagree, 10 = strongly
agree) with the following statement: “Extensive cross-functional input was used to

evaluate AMT options”. Based on 79 responses, the mean response was 6.9,

67

while both the median and mode were 8.0, suggesting potentially strong cross-
functional inﬂuence on a decision-maker. A second question asked respondents
to rate the extent of agreement to this statement: “Extensive cross-functional
decision-making was used to recommend the AMT.” The mean was 6.5, the
median 7.0 and the mode 8.0, again suggesting cross-functional inﬂuence on a key
decision-maker.

The research found no support for the following hypothesis speciﬁc to the
AMTS-TPB.

H6-TPB: Perceived behavioral control over adopting the AMT directly
inﬂuences a decision-maker’s behavioral intention to adopt the AMT.

This ﬁnding is not consistent with the general TPB (Ajzen, 1991) which suggests that the
perceived difﬁculty or ease of performing the speciﬁed behavior directly inﬂuences
intentions. Neither Davis et al. (1989) nor Lee (1990) tested the direct “Ease of
Application” to “Behavioral Intention” relationship, so discussing the lack of support for
this relationship in terms of existing research is not possible.
5. 2.2 Hypotheses Specific to AM T AM

The three hypotheses speciﬁc to the AMTAM are examined in this section. The
research found no support for the following hypothesis speciﬁc to the AMTAM.

HS-AMTAM: Perceived Behavioral Control directly inﬂuences Attitude
Toward Adopting the AMT.

The ﬁnding is basically consistent with the ﬁnding of Davis et al. (1989) in terms
of their development and test of TAM. In their longitudinal study, they found no

signiﬁcant relationship in the ﬁrst time period, but found a signiﬁcant albeit not

68

strong relationship in the second time period. Lee (1990) also found a statistically
signiﬁcant but not strong relationship between the factors.
The research found support for the following hypothesis speciﬁc to the AMTAM.

H4-AMTAM: Perceived Behavioral Control directly inﬂuences the
Perceived Useﬁilness of the AMT.

The ﬁnding is generally not consistent with the ﬁnding of Davis et al. (1989) in
terms of their development and test of TAM. In their longitudinal study, they
found no signiﬁcant relationship in the ﬁrst time period, but found a signiﬁcant
albeit not strong relationship in the second time period. Lee (1990) also found a
statistically signiﬁcant but not strong relationship between the factors.
The research found support for the following hypothesis speciﬁc to the AMTAM.

H6-AMTAM: Perceived Usefulness directly inﬂuences Intention to Adopt
the AMT

The ﬁnding is consistent with both the ﬁnding of Davis et al. (1989) who found a
signiﬁcant relationship in both time periods of their longitudinal study, and the ﬁndings
of Lee (1990).
5.3 Contributions of the Research

The primary focus of this research was to examine the inﬂuence of speciﬁc factors
on a key decision-maker’s intentions to adopt a speciﬁc type of AMT. A research plan
loosely based on the approach used by Davis et al. (1989) was followed. The following
sections examine both the academic and managerial implications of this research.
5.3.1 Academic Contribution

Davis et al. (1989) used both the TRA and an adaptation of the TRA in the form

of the TAM to study user acceptance of information technologies. Research conducted

69

by Lee (1990) extended the TAM into international technology transfer research and
found general support for use of the TAM in such a context.

Though Davis et al. (1989) state that the “TAM is considerably less general than
TRA” and that the TAM is “designed to apply only to computer usage behavior” this
research attempted to determine if the TAM might also be applicable in examining
adoption of AMTs. Table 17 contrasts the research approach and framework used by
Davis et al. (1989) to the research herein. Clearly there are substantial differences in the
research approaches despite both efforts using similar underlying frameworks and
focusing on the adoption of “technology.” Given these differences, a direct comparison
of results is not feasible. However, the following sections contrast the research ﬁndings

using a subjective approach.

Table 17 : Contrast of Current Research to that of Davis et a1. (1989)

 

 

 

 

 

 

 

 

 

 

 

Category Davis et al., 1989 Research Herein
Research Technique Experimental Survey based
Research Method Longitudinal Cross-sectional
Unit of Analysis Adoption of a computer Adoption of AMT
software application (CNC/DNC, material
working lasers, robots)
Setting MBA program Manufacturing ﬁrm
Research Participants MBA students Operations/Capital
Equipment Managers
Number of Participants 107 123
Analysis Technique Regression Structural equation
modeling
General Behavioral Model Theory of Reasoned Action Theory of Planned Behavior
(TRA) (TPB) derived from the
TRA
Specific model adapted Technology Adoption Advanced Manufacturing
from underlying theoretical Model (TAM) Technology Adoption
framework Model (AMTAM) derived
from TAM

 

 

 

70

 

5.3.1.1 Comparison of Direct Applications T RA and T PB Models

Given the differences in methodology and contexts, direct comparisons of current
ﬁndings to the ﬁndings of Davis et al. (1989) are not appropriate. However, this section
compares ﬁndings of the @1391 application of the underlying general behavioral models
(the TRA in the case of Davis et al., and the TPB in the research herein.) Table 18
summarizes and compares the ﬁndings of Davis et al. (1989) to the ﬁndings herein.

Both research efforts determined that behavioral intentions to adopt a
“technology” can be explained reasonably well using a parsimonious set of factors. This
does not suggest that there are not other important factors (e. g., organizational, political,
strategic, etc.), but it does suggest that in these speciﬁc contexts a signiﬁcant amount of
variance can be explained using the factors identiﬁed.

Both research efforts found a signiﬁcant and substantial inﬂuence of attitude on
intentions (in the ﬂee; extension of the general models.) Subjective norms were not
found to have a signiﬁcant inﬂuence on behavioral intentions in the research context of
Davis et al. (1989), but were found to have a signiﬁcant inﬂuence in the research herein.
As was discussed in the hypotheses test section of this report, this difference may be
explained by the signiﬁcant cross-functional nature of the AMT adoption process. As
Davis et al. (1989) pointed out, adoption of a word processing program is a fairly
personal and individual decision, and may be driven less by social inﬂuences than more
cross-ﬁmctional applications such as electronic mail or group decision support systems.
Perceived behavioral control was not explicitly examined by Davis et al. (1989) so a

comparison of the inﬂuence of this factor is not feasible.

71

Table 18: AMTS-TPB — Comparison of Findings to TRA

 

 

 

 

 

 

 

 

 

 

 

 

 

Davis et al. (1989) Davis et al. (I 989) Current Research
at Time One at Time Two Herein
(TRA) (TRA) (AMTS-TPB)
Equation Rz Beta R2 Beta R2 Estimate
B1 = A + SN 0.32 0.26 —- -
A 0.55 * 0.48 *
SN 0.07 0.10
B1 = A + SN + PBC -- -- -- -- 0.91
A 0.50 *
SN 0.69 *
PBC 0.03
* Signiﬁcant at p s 0.05
B1 = Behavioral Intention
A = Attitude

SN = Subjective norms
PBC = Perceived behavioral control

5.3.1.2 Comparison of Adapted Applications the T RA

Given the differences in methodology and contexts, direct comparisons of current
ﬁndings to the ﬁndings of Davis et al. are not appropriate. However, this section
compares the TAM test results of Davis et a1. (1989) to the AMTAM test results herein.

Table 19 summarizes and compares the ﬁndings. Both research efforts
determined that behavioral intentions to adopt a “technology” can be explained
reasonably well using a parsimonious set of factors. This does not suggest that there are
not other important factors (e. g., organizational, political, strategic, etc.), but it does
suggest that in these speciﬁc contexts a signiﬁcant amount of variance can be explained
using the factors identiﬁed.

Both research efforts suggest that perceived usefulness of a technology is the key

determinant of intention, while attitude may have minimal inﬂuence on intentions.

72

Table 19: AMTAM - Comparison of Findings to TAM

 

 

 

 

 

 

 

 

 

 

 

 

Davis et al. (1989) Davis et al. (1989) Current Research
at Time One at Time Two Herein (AM TAM)
(TAM) (TAM)
Equation R2 Beta Rz Beta R2 Estimate
BI = A + U 0.47 0.51 0.55
A 0.27 * 0.16 - 0.45
U 0.48 * 0.61 * 0.90 *
A = U + EOU 0.37 0.36 0.34
U 0.6] * 0.50 "' 0.62 *
EOU 0.02 0.24 * -0.09
U = EOU 0.01 0.05 0.17
EOU 0.10 0.23 * 0.42 *

 

 

* Signiﬁcant at p S 0.05

B1 = Behavioral Intention

A = Attitude

U = Perceived Usefulness

EOU = Perceived ease of use in TAM; z Perceived behavioral control in AMTAM

Similarly, both research efforts suggest that perceived usefulness is a key determinant of
attitude, while ease of use may have minimal inﬂuence on attitude.

Finally, the research efforts appear to suggest different levels of inﬂuence of ease
of use (perceived behavioral control in the research herein) on perceived usefulness. The
research herein suggests a stronger relationship between these factors than does research
by Davis et al. (1989). Davis et al. (1989) originally proposed this relationship based on
the premise that improvements in EOU may be “instrumental,” contributing to increased
performance. That is, decision-makers may believe that any effort saved due to relative
ease of use may be redeployed, enabling a person (or organization) to accomplish more
work for the same or less effort.

In the research herein, this proposition may be especially relevant. Follow-up
interviews with manufacturing managers shed some light on this relationship. One

manager in an electronics ﬁrm indicated that the having conﬁdence that the necessary

73

resources would be available to successfully install and utilize the AMT increases the
perceived utility of the AMT for two primary reasons. First, it suggests that resources
have been planned and allocated in the organization, which in turn suggests the speciﬁc
AMT adoption plans are more likely to be stable and accurately forecasted allowing for
increased efﬁciency and effectiveness of overall resource planning and management.
Second, conﬁdence in successful installation and/or utilization suggests that actual AMT
performance objectives could be more readily and quickly realized. The measures of
AMT performance objectives for this research ultimately included increased quality,
reduced production costs and reduced cycle times. By more quickly realizing all or any
of these AMT performance objectives, resources (people, money or time) would become
available for allocation to other activities.
5.3.1.3 Summary of Academic Contributions
This research was motivated by many factors, including the suggestion that the
application of the TPB to manufacturing technology adoption research is a novel
approach that shows promise for AMT adoption research (Dimnik et al., 1993). Although
the data herein found less than ideal overall model ﬁt for the two speciﬁc fully speciﬁed
theoretical models tested, the following key academic contributions have been made.
1. The Theory of Planned Behavior (TPB) was shown to be applicable in an
industrial context that focused on technologies other than “information systems”
2. The Technology Adoption Model (TAM) developed by Davis et al. (1989)
speciﬁcally to examine adoption of information systems/technologies was shown
to be applicable in an industrial context that focused on speciﬁc types of advanced
manufacturing technologies (i.e., the TAM was shown to be applicable in a
research context other than the setting for which it was intended.)
3. A substantial amount of variance in behavioral intentions to adopt AMT can be
explained using a parsimonious set of factors.

4. Subjective norms are a major determinant of a key decision-maker’s intentions to
adopt an AMT.

74

These contributions are not presented without some caveats, as the fully speciﬁed
model ﬁt indices were not ideal. However, the ﬁndings suggest that future research into
the measures used and application of both the TPB and TAM to AMT adoption decisions
is warranted.

5.3.2 Managerial Contribution

This research has implications for managers in both potential adopting ﬁrms users
and in actual AMT manufacturing ﬁrms. The sections below examine these implications.
5.3.2.1 Implications for AMT Manufacturers

Three key relationships have speciﬁc implications for manufacturers of AMTs.
First and most obvious is the signiﬁcant inﬂuence that perceived usefulness of the AMT
has on a decision-maker’s intentions. Clearly, continued or increased marketing of the
performance beneﬁts of AMT to potential users is warranted.

Second, supplier support did not have a signiﬁcant inﬂuence on perceived
behavioral control. This was somewhat surprising given that prior research suggests that
technology supplier support and prior experience with a supplier encourage increased and
support faster adoption of technologies (Rogers, 1983). Interviews with manufacturing
managers perhaps will shed light on the lack of a signiﬁcant relationship.

One operations manager for a furniture manufacturer remarked that it is often the
supplier or developer of raw materials or component technologies who has the most
extensive knowledge on the installation and use of the AMTs. Thus, it may be the
manufacturers of material or component inputs that are asked to support AMT adoption

decisions and implementation.

75

This comment suggests a closer working relationship continues to evolve between
material “suppliers” and “customers”. It might further suggest that AMT investments are
no longer viewed as company-speciﬁc investments. That, is perhaps as speciﬁc case
studies have suggested, capital equipment decisions are increasingly (albeit slowly
increasingly) being made from a “supply chain management” perspective such that the
company with the capabilities or competencies to best utilize the technology is jointly
identiﬁed by speciﬁc supply chain members. This company may be “required” to make
the investment to globally optimize capital equipment resources throughout the supply
chain rather than attempting to locally optimize the resources. Further, perhaps this
suggests that manufacturing technologies need to become increasingly “mobile” in the
sense that they can be allocated and deployed within the supply chain to provide the
greatest value for the current program. This would have signiﬁcant implications for
risk/reward sharing, ﬁnancing alternatives and the development of technical
competencies throughout the supply chain.

Another manager with an electronics ﬁrm suggested that there are very few
“breakthrough” AMTs being adopted, such that AMT supplier support is not a key
criterion in the decision process. That is, prior internal experience with the technology
mitigates the need for external support.

Based on these ﬁndings, manufacturers of AMTs may want to form stronger
relationships not only with end-users of AMTs, but material and components suppliers as
well. Given that perceived behavioral control was found to signiﬁcantly inﬂuence
perceived usefulness in AMTAM, AMT manufacturers may be able to impact decision-

makers by increasing the amount of perceived support they can provide to users.

76

Third, subjective norms were found to signiﬁcantly inﬂuence intentions. AMT
manufacturers may want to consider broadening their marketing techniques to a cross-
functional representation of personnel at a potential adopting ﬁrm, rather than focusing
solely on the person perceived to be the key decision-maker.
5.3.2.2 Implications for Potential AMT Adapters

This research has implications for key decision-makers, cross-functional
personnel that may be part of an AMT evaluation team, and technology strategists within
potential AMT adopting ﬁrms. First, familiarity with AMT was found to have a
signiﬁcant inﬂuence on perceived behavioral control, which in turn was found to have a
signiﬁcant inﬂuence on perceived usefulness in the AMTAM. Such relationships may
have positive or negative consequences for an organization depending on the perspective.

On one hand, one manufacturing manager suggested that these relationships
indicate that AMT adoption decisions are part of an overall AMT strategy, such that
compatibility and adaptability of the technology are simply extensions of that plan. The
relationship simply reﬂects long term technology planning. On the other hand, another
manufacturing manager suggested that the relationships suggest that there is a “comfort
zone” in which manufacturing managers make decisions such that concerns with
adaptability and compatibility slow the adoption of truly “advanced” technologies. A
technology strategist or advocate for increased adoption of AMTs within their ﬁrm may
wish to use “change management” strategies or processes to help move key decision-
makers out of their comfort zones.

Second, subjective norms are key determinants of intentions in this research

context. Advocates of increased technology adoption may clearly wish to exploit the

77

cross—functional impact on a decision-maker by determining the speciﬁc attributes of the
AMT that each function is most inﬂuenced by, then marketing these attributes to each
function. The strength of the subjective norms - intention relationship also begs the
question, is there such a thing as a “key decision-maker” in such a context? All
subsequent interviews with manufacturing managers suggest that clearly there is a team
leader on any AMT evaluation team, and that soliciting inputs from a variety of functions
and developing consensus for a recommendation are simply part of that job. Most
managers also indicated that they realize the signiﬁcant importance of cross-functional
inputs and that they try to proactively manipulate the perceptions of personnel in other
functions.

This points to an interesting “chicken or the egg” question for future research. Is
the attitude of a key decision-maker already formed based on such factors as perceived
utilities and perceived behavioral control independent of subjective norms? That is, does
a decision-maker basically already have his/her mind made up but proactively inﬂuences
the position of other personnel providing input into the decision? This would require a
series of questions capturing the timing of events in the decision process. The
relationship has some theoretical grounding in the sense that promising new technologies
reviewed for adoption are often brought under review by technology leaders or advocates
that champion their adoption (Dimnik et al., 1993).

Third, despite perceived behavioral control having a signiﬁcant inﬂuence on
perceived useﬁilness in AMTAM, PBC was found to have a non-signiﬁcant inﬂuence on
intentions to adopt an AMT in the AMTS-TPB. This suggests that users may be willing

to tolerate a difﬁcult adoption process in order to realize signiﬁcant competitive beneﬁts,

78

while no amount of perceived behavioral control or ease of adoption will compensate for
a system that does not provide extensive beneﬁts.
5.4 Limitations of the Research

The results generally support the fully speciﬁed models, though speciﬁc
relationships within each model were found to be insigniﬁcant and overall model ﬁt was
not ideal across all model ﬁt indices. This section examines research limitations and their
impact on interpretation of the results.

First, structural equation modeling does not “prove” a model is the right model.
SEM only serves to disconﬁrrn a model. Even when data is determined to ﬁt a model,
there may be alternative models that the data will also ﬁt. This point is exempliﬁed by
this research as well as the research of Davis et al., (1989). Both efforts examined two
models that explained a substantial amount of variance in the intention to perform a
speciﬁed behavior. However, SEM does simultaneously analyze multiple relationships
while providing statistical efﬁciency. Further, it allows for analysis of unobserved
concepts and accounts for measurement error during estimation.

Second, self-assessment measures were used. Though self-assessment is the most
commonly used approach in empirical research, such measures may be biased due to
“halo effects” or “socially desirable” responses for example. However, by targeting the
most informed respondent and pre-screening the survey instrument to minimize confusing
and/or potentially sensitive questions, an attempt was made to mitigate such bias.

Third, a single respondent was used raising the potential for bias. However, based
on ﬁeld interviews prior to the survey being distributed and inputs from SME, NAPM

and APICS, it was determined that the respondents targeted in this research were likely

79

the key informants and played a key role in the decision-making process. Regardless,
perhaps researchers that can harness the resources can take a more “cross-functional”
approach and/or use multiple respondents in future manufacturing technology adoption
research.

Fourth, this study focused on decisions to adopt speciﬁc types of manufacturing
technology, limiting its generalizability. However, the TPB is a general model that is
intended for adaptation to speciﬁc contexts. By limiting the research to speciﬁc AMT,
deﬁnitional problems regarding what really constitutes a manufacturing technology were
mitigated (e. g., the ambiguity of the term “computer integrated manufacturing system”
was avoided.) Further, the focus on speciﬁc AMT allowed for consistent targeting of
informed respondents.

Finally, the response rate was lower than anticipated (but not altogether
unexpected) despite targeting those industries most likely to adopt such technologies by
using information provided by the SME, NAPM and APICS. Major issues that
apparently impacted response rate were the length of the original questionnaire, the fact
that not all manufacturing ﬁrms use the AMTs examined in this study, and that such
investments may be made very infrequently due to the relatively long useful life of the
AMT. Perhaps future research can screen targeted respondents to ensure they have
recently invested in such technologies prior to sending them a survey. This would at least
ensure the targeted respondents were viable participants, though it would not guarantee
their participation.

As was discussed in Section 4.2.1, the number of usable responses (123) was

below the “rule of thumb calculation” number (170). A bootstrap analysis was conducted

80

to check for potential sample bias. Bootstrapping is a re-sampling procedure that can be
used to estimate the sampling distribution of any statistic whose theoretical sampling
distribution is unknown. A bootstrap using ten replications and a sample size of 123 for
each replication was performed. Two of the ten replications converged with no condition
code. Results from the converged iterations indicated that the root mean squared error of
approximation (RMSEA) was 0.078, with lower and upper bounds of 0.056 and 0.097
respectively. A rule of thumb suggests that the RMSEA should be less than 0.050, with
lower and upper bounds of 0.000 to 0.050 respectively. The bootstrap analysis suggests
that the relatively small sample size biased the results. Clearly, attempts to increase
sample size and to further reﬁne measures are warranted in future research.
5.5 Recommendations for Future Research

Much of the discussion in Chapter 5 already presented many potential directions
for future research. This section ﬁrst summarizes the directions already identiﬁed herein,
then discusses other potential directions.
5. 5. 1 Review of Potential Research Directions Previously Discussed

The following methodological and theoretical directions or questions for future
research were already discussed in this report. First, in an attempt to secure a higher
response rate it may be advisable to pre-screen the list of potential participating ﬁrms to
ensure all contacts have adopted the technology under investigation. This would require
phone numbers for the majority of potential contacts. Phone numbers for speciﬁc
contacts are not typically available in mailing lists.

Second, the use of multiple respondents at each ﬁrm could mitigate bias concerns.

This would require extensive resources and would likely negatively impact response rate.

81

It also would suggest that there is more than one key decision-maker in terms of the
evaluation of the AMT. This is contrary to the suggestions made by SME, APICS and
NAPM, as well as the comments from managers interviewed in this research.

Third, the temporal precedence or the potential for recursive relationships of
factors inﬂuencing the key decision-maker could be examined. For example, perhaps as
one manufacturing manager suggested, the decision-maker’s intention to adopt a
technology is formed prior to cross-functional input, and that through marketing of the
AMT by the key person, cross-functional perspectives are brought in-line with the key
decision-maker’s or team leader’s perspective in advance of the actual evaluation process.

Fourth, the role of material suppliers in the decision making process could be
examined. As one manager suggested, many material suppliers may be at least as
informed and perhaps more knowledgeable of the most effective and efﬁcient AMT
installation and utilization processes. Further, over time material suppliers are likely to
develop closer and more integrated relationships with AMT adopters than are AMT
manufacturers, due to a higher frequency of contact between the ﬁrms.

Fifth, the application of TAM, TRA or TPB could be expanded into research of
adoption decision processes for the other categories of advanced manufacturing
technology classiﬁed by the US Department of Commerce. This would include design
and engineering technologies (i.e., CAD, CAM, CAPP) automated material handling
technologies, and manufacturing information technologies (i.e., MRP/MRPH, ERP) for
example. However, the key decision-maker for these distinct categories of AMT may be
in very different functional positions in each case. Research that includes all of these

technology categories into a single research effort may produce confounding results.

82

5. 5.2 Other Potential Research Directions

The following directions for future research were not previously discussed in
Chapter 5. First, this research examined decision-making processes in cases of actual
AMT adoption. In doing so, there is limited insight into what is preventing companies
from adopting AMTs. Future research may want to examine the decision process from a
more hypothetical perspective (e. g., if this condition were true, what would your attitude
be?). For this early stage in the research stream however, it was decided to take an
approach similar to that of Davis et al. (1989) and examine actual adoption cases. Also,
the hypothetical perspective might capture espoused decision processes that may be
signiﬁcantly different than decision-making processes in practice.

Second, future research may wish to explicitly examine the inﬂuence of long term
business or technology strategy on the decision process. This research only implicitly
examined the role of this factor through previous experience with the technology.

Third, the measures used for certain factors, particularly supplier support, need
further reﬁnement. Despite the lack of a signiﬁcant relationship between supplier support
and perceived behavioral control in this and in Lee’s (1990) research, there is enough
case study evidence to continue to examine the role of AMT supplier support. Perhaps
measures in future research should more clearly distinguish between the tacit aspects of
supplier support and the tactical aspects.

Fourth, perhaps “familiarity with AMT” has a direct inﬂuence on intentions to
adopt an AMT as one manufacturing manager suggested, rather than an inﬂuence that is

mediated by perceived behavioral control. Alternative models with this direct

83

relationship were explored, but none were found to signiﬁcantly improve the amount of
variance explained in the intention to adopt an AMT.

Finally, a statistically signiﬁcant and strong relationship was found between
subjective norms and a key decision-maker’s intention to adopt an AMT. Future research
that investigates the inﬂuences of various factors on the AMT evaluation process of
personnel other than the team leader or key decision-maker may provide valuable insights
for AMT manufacturers as well as AMT adopters. The application of TAM, TRA or
TPB to the study of supporting team members’ evaluation or decision-making processes
appears to be warranted.

5.6 Summary

This research developed and tested two models of manufacturing technology
adoption. The speciﬁc limitations and contributions of the research were discussed
throughout Chapter 5 and will not be repeated here.

Broadly however, the research contributed to the body of knowledge by ﬁlling a
void in technology adoption research. While many case studies and surveys of AMT
analysis techniques have been presented in the literature, they tend to be solely
descriptive and only make a limited attempt to integrate the ﬁndings into a decision
process model and to advance theory. Clearly the previous research has ﬁlled the critical
role of identifying the factors that may inﬂuence the AMT adoption decision and in
developing appropriate measures. This research built on those efforts by developing
integrated decision models based on the widely applied and generalizable Theory of

Planned Behavior framework, and on the more context speciﬁc Technology Acceptance

84

Model. The research conﬁrms the suggestion that the TPB is a novel approach that
shows promise for AMT adoption research (Dimnik et al., 1993).

The research also suggests that research into technology adoption decision
processes should be careful when selecting a unit of analysis. That is, if this research had
addressed all categories of advanced manufacturing technology identiﬁed by the US
Department of Commerce (e.g., design systems such as CAD/CAM, planning
technologies such as MRPH) perhaps the relationships would have been different. For
example, though not directly comparable, the comparison of this research to the research
of Davis et al. (1989) into information systems adoption found that subjective norms have
substantially different inﬂuences in the respective research contexts. In an attempt to
increase the generalizability of research, a determination must ﬁrst be made if the attempt
to generalize ﬁndings is valid.

Numerous extensions of the research were identiﬁed. Perhaps the most
interesting and forward-looking research will examine how (if) capital equipment
decisions will be made in a supply chain context. Assuming such decisions are
increasingly made in a supply chain context, research that examines successful ways
supply chains are structured, relationships managed, equipment utilized, risks and
rewards shared, and total resources deployed throughout a supply chain will make
substantial practical and theoretical contributions by extending management and behavior

sciences into capital equipment/ supply chain management research.

85

APPENDIX A: QUESTIONNAIRE

The questionnaire size was reduced to meet dissertation formatting requirements.

SECTION I: TECHNOLOGY IDENTIFICA TION

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1. Please Identify a manufacturing technology that meets the criteria listed on the cover page.
NOTE: From this point forward “AMT” refers to the specific manufacturing technology identified below.
1. Please fill in each of the blanks below. If not known exactly, please provide your best estimate.
a) Date (month / year) the AMT was installed at your plant > I
b) Annual sales the year the AMT was installed at your plant > S
c) Purchase and installation cost for the AMT (at time ofinstallation in US dollars) > S
d) Installation and debug time (in months) Refer to previous deﬁnition if necessary > months
e) Process optimization time (in months) Refer to previous deﬁnition if necessary > months
1) Expected useful life (in years) ofthe AMT without major re-tooling or re-investment > years
g) Projected payback period (in months) for the AMT > months
h) Projected ROI for the AMT investment > %
SECTION 2: AMT SELECTION PROCESS
Questions in this section refer to evaluations of the AMT prior to actual AM 1 adm' Please respond
without letting hindsight impact your responses. Remember that all responses will be kept strictly
conﬁdential
2. Prior to actual adoption, I strongly believed that: Strongly Strongly
Dismee Neutral Agree

8) All necessary resources would be available to successfully install and use the AMT 0 | 2 3 4 5 6 7 8 9 10
b) Adoption of the AMT would make the plant much more competitive 0 l 2 3 4 5 6 7 8 9 10
c) The AMT would signiﬁcantly improve overall plant performance 0 l 2 3 4 5 6 7 8 9 10
d) The AMT could be installed and operational within budget and time objectives 0 l 2 3 4 5 6 7 8 9 10
3. Relative to alternative technology or sourcing options, I had a high Strongly Strongly

dggree of confidence that the AMT worle: Dlsggree Neutral Agree
a) Minimize defect rate 0 l 2 3 4 5 6 7 8 9 10
b) Maximize production quality consistency 0 l 2 3 4 5 6 7 8 9 10
c) Reduce labor costs 0 l 2 3 4 5 6 7 8 9 10
d) Lower raw material and purchased component costs 0 l 2 3 4 5 6 7 8 9 10
e) Lower fully burdened per unit production costs 0 1 2 3 4 5 6 7 8 9 10
f) Minimize equipment setup and tear down times 0 1 2 3 4 5 6 7 8 9 10
g) Reduce overall production cycle times 0 1 2 3 4 5 6 7 8 9 10
h) Provide greater volume ﬂexibility 0 l 2 3 4 5 6 7 8 9 10
1) Provide responsiveness to technological changes 0 1 2 3 4 5 6 7 8 9 10
j) Allow for a wider variety of products to be produced 0 1 2 3 4 S 6 7 8 9 10
k) Drive increased product innovations 0 l 2 3 4 5 6 7 8 9 10
1) Drive increased process innovations 0 l 2 3 4 5 6 7 8 9 10
m) Be highly compatible with existing operations 0 l 2 3 4 5 6 7 8 9 10
11) Be operated by current workers without providing additional training 0 l 2 3 4 5 6 7 8 9 10
0) Be easily adapted to existing operations 0 I 2 3 4 5 6 7 8 9 10
5. My attitude towards Investment In the AMT was favorable based on Strongly Strongly

projected performance In terms of: Disagree Neutral Agree
a) Cost 0 l 2 3 4 5 6 7 8 9 10
b) Quality 0 l 2 3 4 5 6 7 8 9 10
c) Flexibility 0 1 2 3 4 5 6 7 8 9 10
d) Production cycle time 0 1 2 3 4 5 6 7 8 9 10
e) lnnovativeness 0 l 2 3 4 S 6 7 8 9 10

 

 

86

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6. I had a high degree of confidence that the recommended AMT Strongly Strongly
supplier would or could provide the desired: Disagree Neutral Agree
a) AMT installation support 0 l 2 3 4 5 6 7 8 9 l0
b) Training of operators. supervisors or managers 0 l 2 3 4 5 6 7 8 9 10
c) Detailed installation. operation and/or maintenance manuals 0 l 2 3 4 S 6 7 8 9 l0
7. I believed that adoption of new AMT was favored by: Strongly Strongly
Disaﬂee Neutral Agree
a) Top management 0 l 2 3 4 5 6 7 8 9 l0
b) Co-workers 0 l 2 3 4 5 6 7 8 9 10
c) Subordinates 0 l 2 3 4 5 6 7 8 9 IO
8. Please rate extent of agreement with the following statements. Strongly Strongly
Dis_ag_ree Neutral Agree
a) Extensive cross-functional MWas used to evaluate AMT options 0 l 2 3 4 5 6 7 8 9 10
b) Extensive cross-functional dieision-making was used to recommend the AMT 0 l 2 3 4 5 6 7 8 9 10
9. Our company and the recommended AMT supplier: Strongly Strongly
Disagree Neutral Agree
a) Had extensive experience working with each other 0 l 2 3 4 5 6 7 8 9 10
b) Were familiar with each other‘s operations and capabilities 0 l 2 3 4 5 6 7 8 9 l0
c) Were aware of each other's competitive needs 0 l 2 3 4 5 6 7 8 9 10
10. I believed that adopting the AMT would make the plant more Strongly Strongly
competitive in terms of performance in: Disagree Neutral Agree
a) Cost 0 l 2 3 4 5 6 7 8 9 l0
b) Quality 0 l 2 3 4 5 6 7 8 9 10
c) Flexibility 0 l 2 3 4 S 6 7 8 9 l0
d) Production cycle time 0 l 2 3 4 S 6 7 8 9 10
e) lnnovativeness 0 l 2 3 4 5 6 7 8 9 10
ll. Prior to installation, I believed that the following resources would be Strongly Strongly
available to successfully install and/or use the AMT: Disagree Neutral Agree
8) Supporting technology (e.g.. material handling, MRP system) 0 l 2 3 4 5 6 7 8 9 l0
b) Technical resources and knowledge 0 l 2 3 4 5 6 7 8 9 10
c) Hunmn resources with appropriate skills 0 l 2 3 4 5 6 7 8 9 10
d) Financial resources 0 l 2 3 4 5 6 7 8 9 10
12. Opinions that W my position towards recommending the Strongly Strongly
AMT for adoption were directly or indirectly expressed by: Disagree Neutral Agree
a) Top nunagement 0 l 2 3 4 5 6 7 8 9 10
b) Co-workers 0 l 2 3 4 5 6 7 8 9 10
c) Subordinates 0 l 2 3 4 5 6 7 8 9 10
13. Based on all potential factors and overall evaluation of the AMT, I: Strongly Strongly
Diggee Neutra_l Agree
a) Strongly recommended the AMT for adoption 0 l 2 3 4 5 6 7 8 9 l0
b) Fully intended for the company to adopt the AMT 0 l 2 3 4 5 6 7 8 9 10
c) Provided well-documented support favoring adoption of the AMT 0 l 2 3 4 5 6 7 8 9 IO

 

87

 

 

 

 

 

 

SECTION 3: ITEM lNSOURClNG/OUTSOURCING DECISION

Questions in this section refer to the insourcing/outsourcing
decision process that led to the decision to insource, ultimately driving the

need for investment in new manufacturing technology.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I4. Please Identify the ITEM produced by the AMT. If a variety of items are produced, indicate as such and identify a
representative item(s). Please discuss unique product and/or process characteristics.
NOTE: From this point forward “ITEM” refers to the product identified below.
IS. Please ﬁll in each of the blanks below. If not known exactly, please provide your best estimate.
a) Number of units ofthe ITEM produced annually >
b) Approximate fully burdened per unit production cost for the ITEM > S
I6 Is this ITEM typically shipped as part of a higher level assembly to your Yes D No 0
external customer?
17. People involved in the decision to make the item internally (insource) Strongly Strongly
projected that relative to outsourcing options we could or needed to: Disagree Neutral Agree
a) Minimize defect rates 0 l 2 3 4 5 6 7 8 9 IO
b) Maximize production quality consistency O I 2 3 4 5 6 7 8 9 IO
c) Reduce labor costs 0 I 2 3 4 S 6 7 8 9 I0
(I) Lower raw material and purchased component costs 0 l 2 3 4 5 6 7 8 9 l0
e) Lower fully burdened per unit production costs 0 I 2 3 4 5 6 7 8 9 IO
0 Minimize equipment setup and tear down times 0 I 2 3 4 5 6 7 8 9 IO
g) Reduce overall production cycle times 0 I 2 3 4 S 6 7 8 9 IO
h) Provide greater volume ﬂexibility O I 2 3 4 5 6 7 8 9 10
i) Provide responsiveness to technological changes 0 i 2 3 4 5 6 7 8 9 l0
j) Allow for a wider variety of products to be produced 0 I 2 3 4 5 6 7 8 9 l0
k) Drive increased product innovations 0 I 2 3 4 5 6 7 8 9 IO
l) Drive increased process innovations 0 I 2 3 4 5 6 7 8 9 IO
18. People involved In the decision to make the item internally (Insource) Strongly Strongly
believed that successful insourcing would require us to select an: Disagree Neutral Agree
a) AMT supplier that provided signiﬁcant AMT installation support 0 I 2 3 4 5 6 7 8 9 I0
b) AMT supplier that provided extensive training for our personnel 0 l 2 3 4 5 6 7 8 9 IO
c) AMT supplier that provided installation. operation and mintenance manuals O I 2 3 4 5 6 7 8 9 l0
d) AMT that was highly compatible with existing operations 0 I 2 3 4 5 6 7 8 9 IO
e) AMT that was easily adapted to our existing operations 0 I 2 3 4 5 6 7 8 9 IO
l) AMT that could be run by current workers without providing added training 0 I 2 3 4 5 6 7 8 9 l0

 

88

 

 

 

 

 

SECTION 4: RESUL TS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I9. Please rate actual AMT process optimization performance results Far worse Met Much better
relative to goals for each measure below. than goals goals than goals
a) Time required to optimize process 0 I 2 3 4 5 6 7 8 9 IO
b) Budget required to optimize process 0 I 2 3 4 5 6 7 8 9 IO
c) Number or complexity of AMT technical installation/optimization problems 0 l 2 3 4 5 6 7 8 9 IO
d) Disruption to ongoing production operations during installation/optimization O I 2 3 4 5 6 7 8 9 IO
20. For thg ﬁrst 6 months of progugtign afte: pmess optimizatigg, please Far worse Met Much better
rate average production perform_ance relative to goals for: than goals ggals than goals
a) Defect rate 0 I 2 3 4 5 6 7 8 9 IO
b) Production quality consistency 0 l 2 3 4 5 6 7 8 9 IO
c) Labor costs 0 I 2 3 4 5 6 7 8 9 l0
d) Raw material and purchased component costs 0 l 2 3 4 S 6 7 8 9 IO
e) Fully burdened per unit production costs 0 I 2 3 4 5 6 7 8 9 IO
0 Equipment setup and tear down times 0 I 2 3 4 5 6 7 8 9 IO
g) Overall production cycle times 0 l 2 3 4 5 6 7 8 9 I0
h) Volume ﬂexibility 0 l 2 3 4 S 6 7 8 9 I0
i) Responsiveness to technological changes 0 I 2 3 4 5 6 7 8 9 IO
j) Variety of products produced 0 I 2 3 4 5 6 7 8 9 IO
k) Product innovations O I 2 3 4 5 6 7 8 9 I0
I) Process innovations O I 2 3 4 5 6 7 8 9 IO
m) Overall plant competitiveness O l 2 3 4 5 6 7 8 9 IO
21- For WWW» Please Far worse Met Much better
rate the averggercontriljution of ITEMS Jm_ade on the AMT to: than goals M than goals
a) Proﬁt on the sale ofthe ITEM or end product the ITEM is installed in 0 I 2 3 4 5 6 7 8 9 IO
b) Sales volume of the ITEM or end product the ITEM is installed in 0 I 2 3 4 5 6 7 8 9 l0
c) Customer satisfaction with the ITEM or end product the ITEM is installed in O I 2 3 4 5 6 7 8 9 IO
22. For MWWMQM identify actual average ITEM quality using either of
the metrics listed: PPM defective - OR — % defect rate.
23. Please rate extent of agreement with the following statements. The Strongly Strongly
AMT supplier provided: Disagree Neutral Agree
a) Valuable inputs to improve product design/manufacturability O I 2 3 4 5 6 7 8 9 10
b) Technology improvements in the AMT miﬁe to our needs 0 l 2 3 4 5 6 7 8 9 IO
c) Signiﬁcant AMT installation support 0 I 2 3 4 5 6 7 8 9 I0

 

 

This concludes the “required” section.
However, we encourage you to complete the following “optional” pages.
Thank you.

89

 

SECTION 5: OPTIONAL
To provide a useable response, the questions below do not need to be answered. However, we encourage
you to respond to these questions so that we can provide you with additional insights.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

24. To what extent was investment in the AMT justified on financial Financial Strategic
estimates (e.g., ROI, payback) and/or strategic factors (e.g., Only Balanced Only
competitive positioning, support of core competencies) 0 I 2 3 4 5 6 7 8 9 10

25. Please rate extent of agreement with the following statements. Strongly Strongly

Diggree Neutral Agree

a) The use of state-of~the-art AMT by competitors pressures us to adopt new AMT 0 I 2 3 4 5 6 7 8 9 IO

b) Internal organizational pressures inﬂuenced the decision to adopt new AMT O I 2 3 4 5 6 7 8 9 10

c) We need to keep pace with AMT investments made by our competitors 0 I 2 3 4 5 6 7 8 9 I0

(I) External competitive pressures inﬂuenced the decision to adopt new AMT O I 2 3 4 5 6 7 8 9 IO

26. At the time of selection, the AMT was: Strongly Strongly

Disagree Neutral Agree

a) A technology already proven effective and reliable in industry 0 I 2 3 4 5 6 7 8 9 10

b) Selected to provide new technical capabilities for new product development 0 I 2 3 4 5 6 7 8 9 l0

c) Chosen to replace existing equipment (e.g., to improve cost, quality) 0 I 2 3 4 5 6 7 8 9 l0

d) A technology that directly supported a core product. process or capability O I 2 3 4 5 6 7 8 9 IO

27. For the ms; recggt 9 months 2f produgtigg, please rate average Far worse Met Much better
production performance agile AMT rel_a_t_ive to £2!!! for: t_l_ian goals gals than goals

a) Defect rate 0 I 2 3 4 5 6 7 8 9 IO

b) Fully burdened per unit production costs 0 I 2 3 4 S 6 7 8 9 IO

c) Overall production cycle times 0 l 2 3 4 5 6 7 8 9 I0

d) Volume ﬂexibility 0 I 2 3 4 5 6 7 8 9 IO

e) Responsiveness to technological changes 0 l 2 3 4 5 6 7 8 9 IO

0 Product innovations 0 I 2 3 4 5 6 7 8 9 IO

g) Process innovations 0 l 2 3 4 5 6 7 8 9 IO

h) Overall plant competitiveness O I 2 3 4 5 6 7 8 9 IO

28. For thg mg mm! 6 maths of productiog, identify actual average ITEM quality using either of the metrics listed:

PPM defective - OR - % defect rate.

29. For WW, please rate the average Far worse Met Much better
contribution of ITEMS made on the AMT to: than goals goals than goals

a) Proﬁt on the sale of the ITEM or end product the ITEM is installed in O I 2 3 4 5 6 7 8 9 l0

b) Sales volume of the ITEM or end product the ITEM is installed in 0 I 2 3 4 5 6 7 8 9 I0

c) Customer satisfaction with the ITEM or end product the ITEM is installed in 0 l 2 3 4 5 6 7 8 9 IO

30. Please rate extent of agreement with the following statements. Strongly Strongly

Diggree Neutrﬂ A_g[ee

a) Manufacturing strategy is closely aligned with overall business strategy 0 I 2 3 4 5 6 7 8 9 IO

b) Manufacturing is highly integrated into new product development 0 I 2 3 4 5 6 7 8 9 IO

c) Lessons from the AMT selection and optimization process were documented 0 I 2 3 4 5 6 7 8 9 IO

d) A formal project plan guided the AMT selection and installation process 0 I 2 3 4 5 6 7 8 9 IO

e) AMT strategy/selection criteria are aligned with insourcing/outsourcing strategy 0 I 2 3 4 5 6 7 8 9 IO

 

 

9O

 

SECTION 5: OPTIONAL (Continued)

To provide a useable response, the questions below do not need to be answered. However, we encourage
you to respond to these questions so that we can provide you with additional insights.

 

 

 

 

 

 

 

 

 

 

 

 

 

36.

technth stratggies across different plants or business units?

Yes

 

31. Without any commitment to purchase, prior to AMT selection our Strongly Strongly
company: Disagree Neutral Agree

a) Viewed full scale operation ofthe AMT O l 2 3 4 5 6 7 8 9 10

b) Ran a pilot project in our facility or at the AMT supplier's facility 0 I 2 3 4 5 6 7 8 9 IO

32. Prior to adoption of the AMT, I believed that: Strongly Strongly

Disagree Neutral Agree

a) Our competitors were using the same or similar AMT O I 2 3 4 5 6 7 8 9 l0

b) Primary customers expected us to invest in new manufacturing technologies 0 I 2 3 4 5 6 7 8 9 10

c) Our competitors recently made a signiﬁcant upgrade to their AMT capabilities O l 2 3 4 S 6 7 8 9 IO

33. Please classify the role that each of the following analysis Not Minor Moderate Significant
techniques had in evaluating the AMT investment Used Use Use Use

a) Payback period CI 0 l 2 3 4 5 6 7 8 9 IO

b) Return on invesnnent (ROI) D 0 I 2 3 4 S 6 7 8 9 l0

c) Computer simulation of AMT performance CI 0 I 2 3 4 5 6 7 8 9 I0

(I) Formal risk analysis U 0 I 2 3 4 5 6 7 8 9 IO

34. Please discuss the most significant barriers to justification, Implementation and/or utilization of AMT investments and
how they are managed or overcome.

35. Do AMT investment decisions at your plant need to be coordinated with business or E] E]

No

 

 

If you answered “yes” please discuss the process and how this helps or hinders AMT justification and adoption.

 

 

 

91

 

 

 

 

 

SECTION 5: OPTIONAL (Continued)
To provide a useable response, the questions below do not need to be answered. However. we encourage
you to respond to these questions so that we can provide you widt additional insights.

 

 

 

 

 

37. Based strictly on initial financial analysis (e.g., ROI, payback) was investment in this AMT y“ D No
‘ your company’s highest rated option ﬂncluding outsourcing options)?
38. Were any numbers “massaged” or “manipulated” after initial analysis to make financial Yes E] No
analysis more supportive of the investment? Remember, conﬁdentiality is guaranteed

 

39- When a strictly financial analysis indicates that a specific AMT Investment Is not the highest rated option, please

discuss how that AMT Investment is justiﬁed over other investments or outsourcing options.

 

 

 

40. Do insourcing/outsourcing decisions at your plant need to be coordinated with business or y“ D No
sourcing stratggies across different plants or business units?

 

 

 

 

41. If you answered “yes” please discuss if the process typically favors one option over the other and why. Please also
discuss if this coordination helps or hinders insourcing/outsourcing decisions at your firm.

 

 

 

 

42. Please discuss the process at your firm for aligning insourcing/outsourcing strategy with manufacturing technology
strategy. If there is no alignment, please discuss If alignment would improve both insourcing/outsourcing decisions

and AMT investment decisions, and how such alignment could be achieved.

 

 

 

 

Thank you. Please return the questionnaire in the enclosed envelope or via fax (616) 432-2094.

92

BIBLIOGRAPHY

Ajzen, I. (1985). From Intentions to Actions: A Theory in Planned Behavior. New York:
Springer-Verlag.

Ajzen, I. (1991). The Theory of Planned Behavior. Organizational Behavior and Human
Decision Processes, 50(2), 179-211.

Ajzen, I., C. Timko and J. White (1982). Self-Monitoring and the Attitude-Behavior
Relations. Journal of Personality and Social Psychology, 42, 426-435.

Anderson, J., G. Cleveland and R. Schroeder (1989). Operations Strategy: A Literature
Review. Journal of Operations Management, 8(2), 133-159.

Bagozzi, R. (1982). A Field Investigation of Causal Relations among Cognitions, Affect,
Intentions and Behavior. Journal of Marketing Research, I 9, 562-5 84.

Bandura, A. (1977). Self-Efﬁcacy: Toward a Unifying Theory of Behavioral Change.
Psychological review, 84(2), 191-215.

Basche, J. and M. Duerr (1975). International Transfer of Technology: A Worldwide
Survey of Chief Executives, A Report from the Conference Board.

Brinberg, D. (1979). An Examination of the Determinants of Intention and Behavior: A
Comparison of Two Models. Journal of Applied Social Psychology, 9, 560-575.

Bumkrant, R. and T. Page (1988). The Structure and Antecedents of the Normative and
Attitudinal Components of Fishbein’s theory of Reasoned Action. Journal of
Experimental and Social Psychology, 24, 66-87.

Calantone, R., M. Lee and A. Gross (1990). Evaluating International Technology Transfer
in a Comparative Marketing Framework. Journal of Global Business Marketing, 3(3), 23-
46.

Canada, J. and W. Sullivan (1989). Economic and Multiattribute Evaluation of Advanced
Manufacturing Systems. Englewood Cliffs, NJ: Prentice Hall.

Chatterji, D. (1995). Achieving Leadership in Environmental R&D. Research Technology
Management, 38(2), 3742.

Davis, F ., R. Bagozzi and P. Warshaw (1989). User Acceptance of Computer
Technology: A Comparison of Two Theoretical Models. Management Science, 35(8),
982-1003.

93

 

Dickerson, M. and J. Gentry (1983). Characteristics of Adopters and Non-Adopters of
Home Computers. Journal of Consumer Research, 10, 225-235.

Dimnik, T. and D. Johnston (1993). Manufacturing Managers and the Adoption of
Advanced Manufacturing Technology. Omega, 21(2), 155-162.

Farley, J ., B. Kahn, D. Lehmann and W. Moore (1987). Modeling the Choice to
Automate. Sloan Management Review, 28(2), 5-15.

Ferdows, K. and A. De Meyer (1990). Lasting Improvements in Manufacturing
Performance. Journal of Operations Management, 9(2), 168-184.

Fishbein, M. and I. Ajzen (1975). Belief Attitude, Intention and Behavior: An
Introduction to Theory and Research. Reading, MA: Addison-Wesley.

Frazier, G., R. Spekman and C. O'Neal (1988). J ust-in-time Exchange Relationships in
Industrial Markets. Journal of Marketing, 52(4), 52-67.

Garvin, D. (1987). Competing on the Eight Dimensions of Quality. Harvard Business
Review, 65(6), 101-109.

Gatignon, H. and T. Robertson (1985). A Propositional Inventory for New Difﬁision
Research. Journal of Consumer Research, I, 35-49.

Gerwin, D. (1993). Manufacturing Flexibility: A Strategic Perspective. Management
Science, 39(9), 395-410.

Goldhar, J. and M. J elinek (1983). Plan for Economies of Scope. Harvard Business
Review, 61(6), 141-148.

Green, R., Y. Mahmoud and T. Zimmerer (1994). Manufacturing Technology: Strategic
Opportunities for Success. Information Strategy, 10(2), 38-43.

Handﬁeld, R. (1995). Re—engineering for T ime-based Competition: Benchmark and Best
Practices for Production, Purchasing and R&D. Quorum Press.

Hauser, J. and S. Shugan (1980). Intensity Measures of Consumer Preference. Operations
Research, 28, 279-320.

Hayes, R. and D. Garvin (1982). Managing as if Tomorrow Mattered. Harvard Business
Review.

Hayes, R., S. Wheelright and K. Clark (1989). Dynamic Manufacturing: Creating the
Learning Organization. New York: Free Press.

94

Herbst, A. (1990). The Handbook of Capital Investing. El Paso, TX: Harper Business.
Hill, T. (1985). Manufacturing Strategy. London: MacMillan.

Hines, P. (1996). Purchasing for lean production: The new strategic agenda. International
Journal of Purchasing & Materials Management, 32(1), 2-10.

Kenny, G. and A. Dunk (1989). The Utility of Performance Measures: Production
Managers' Perceptions. IEEE Transactions on Engineering Management, 36(1), 47-50.

Klammer, T. (1994). Managing Strategic and Capital Investment Decisions: Going
Beyond Numbers to Improve Decision Making. New York: Irwin.

Larcker, D. and V. Lessig (1980). Perceived Usefulness of Information: A Psychometric
Examination. Decision Sciences, 1 1, 121 -1 34.

Lee, M. (1990). International Technology Adoption: Conceptual Model and Empirical
Test and China. Department of Marketing. Lexington, Kentucky, The University of
Kentucky: 21 8.

Malone, T. (1975). Toward a Theory of Intrinsically Motivating Behavior. Cognitive
Sciences, 4, 333-369.

Mansﬁeld, E., A. Romeo, M. Schwartz, D. Teece, S. Wagner and P. Brach (1982).
Technology Transfer, Productivity, and Economic Policy. New York: Norton.

Meredith, J ., N. Hyer, D. Gerwin and S. Rosenthal (1986). Research Needs in Managing
Factory Automation. Journal of Operations Management, 6(2), 208-228.

Miller, J. and A. Roth (1994). A Taxonomy of Manufacturing Strategies. Management
Science, 40(3), 285-304.

Monczka, R. and R. Trent (1997). Purchasing and Sourcing 1997: Trends and
Implications. Tempe, AZ, NAPM/CAPS.

Morris, M. and J. Holman (1988). Source Loyalty in Organizational Markets: A Dyadic
Perspective. Journal of Business Research, 16(2), 1 17-131 .

Nabseth, L. and G. Ray (1974). The Diffusion of New Industrial Processes: An
International Study. Cambridge: University Press.

95

Phillips, L., R. Calantone and M. Lee (1994). International Technology Adoption:
Behavior Structure, Demand Certainty and Culture. Journal of Business and Industrial
Marketing, 9(2), 16-28.

Porter, M. (1983). The Technological Dimension of Competitive Strategy. In Research on
Technological Innovation, Management and Policy, R. Rosenbloom (eds.), J AI Press,
Greenwhich, CT.

Porter, M. and C. van der Linde (1996). Green and Competitive: Ending the Stalemate.
Harvard Business Review, 73(5), 120-123.

Primrose, P. (1991). Investment in Manufacturing Technology. London: Chapman &
Hall.

Raunick, D. and A. Fisher (1972). A Probabilistic Make-Buy Model. Journal of
Purchasing, 8(1), 63-80.

Robertson, T. (1971). Innovative Behavior and Communication. New York: Holt.

Robinson, R. (1988). The International Transfer of Technology: Theory, Issues, and
Practice. Cambridge: Ballinger.

Rogers, E. (1983). Diffusion of Innovations. New York: The Free Press.

Rogers, E. and F. Shoemaker (1971). Communication of Innovations: A Cross-Cultural
Approach. New York: The Free Press.

Ryan, M. (1982). Behavioral Intention Formation: A Structural Equation Analysis of

Attitudinal and Social Inﬂuence Interdependency. Journal of Consumer Research, 9, 263-
278.

Samli, A. C., Ed. (1985). Technolog Transfer: Geogphic. Economic. Cultural and
Techngl Dimensions. Westport, Quorum Books.

Sheppard, B., J. Hartwick and P. Warshaw (1988). The Theory of Reasoned Action: A
Meta-Analysis of Past Research with Recommendations for Modifications and Future
Research. Journal of Consumer Research, 15, 325-343.

Shingo, S. (1985). A Revolution in Manufacturing: The SMED System. Stanford,
Massachusetts: Productivity Press.

Skinner, W. (1969). Manufacturing -- Missing Link in Corporate Strategy. Harvard
Business Review.

96

Skinner, W. (1984). Operations Technology: Blind Spot in Strategic Management.
Interfaces, 116-125.

Slack, N. (1989). Focus on Flexibility. In International Handbook of Production and
Operations Management, R. Wild (eds.), Cassell, London, 50-73.

Small, M. and 1. Chen (1997). Economic and Strategic Justiﬁcation of AMT: Inferences
from Industrial Practices. International Journal of Production Economics, 49(1), 65-75.

Stacey, G. and W. Ashton (1990). A Structured Approach to Corporate Technology
Strategy. International Journal of Technology Management, 5(4), 389-407.

Steele, L. (1974). Barriers to International Technology Transfer. Research Management,
17-21.

Stutzman, T. and S. Green (1982). Factors Affecting Energy Consumption: Two Field
Tests of the Fishbein-Ajzen Model. Journal of Social Psychology, 11 7, 183-201.

Swanson, E. (1974). Information Channel Disposition and Use. Decision Sciences, 18,
13 1- 145.

US Commerce Department, U. (1989). Current Industrial Reports: Manufacturing
Technology. Washington, DC, Bureau of the Census.

Vickery, S., C. Droge and R. Markland (1993). Production Competence and Business
Strategy: Do They Affect Business Performance? Decision Sciences, 24(2), 435-455.

Wamock, I. (1996). Manufacturing and Business Excellence: Strategies, Techniques and
Technologies. New York: Prentice Hall.

Welch, J. and P. Nayak (1992). Strategic Sourcing: A Progressive Approach to the Make-
or-Buy Decision. Academy of Management Executive, 6(1), 23-31.

Williamson, 0. (1979). Transaction-Cost Economics: The Governance of Contractual
Relations. Journal of Law and Economics, 22, 233-261.

Wind, Y. (1970). Industrial Source Loyalty. Journal of Marketing Research, 7, 450-45 7.

97

HICHIGnN srnr: UNIV. LIBRRRIES
llHIWillillllllllllllllllllllllllllllllllllHI
31293018345698