Technology flexibility: conceptualization, validation, andmeasurement

Abstract

This research investigates technology flexibility, which is the technology characteristic that allows or enables adjustments and other changes to the business process. Technology flexibility has two dimensions, structural and process flexibility, encompassing both the actual technology application and the people and processes that support it. The flexibility of technology that supports business processes can greatly influence the organization's capacity for change. Existing technology can present opportunities for or barriers to business process flexibility through structural characteristics such as language, platform and design. Technology can also indirectly affect flexibility through the relationship between the technology maintenance organization and the business process owners, change request processing, and other response characteristics. These indirect effects reflect a more organizational perspective of flexibility. This paper asks the question, “what makes technology flexible?” This question is addressed by developing and validating a measurement model of technology flexibility. Constructs and definitions of technology flexibility are developed by examining the concept of flexibility in other disciplines, and the demands imposed on technology by business processes. The purpose of building a measurement model is to show validity for the constructs of technology flexibility. This paper discusses the theory of technology flexibility, develops constructs and determinants of this phenomenon, and proposes a methodology for the validation and study of the flexibility of emerging technologies

Full text

Technology Flexibility:
Conceptualization, Validation, and Measurement
Kay M. Nelson (knelson@ukans.edu) and H. James Nelson (jnelson@stat1.cc.ukans.edu)
The University of Kansas, Division of Accounting and Information Systems
Mehdi Ghods
(
sep075@boeing.com)
The Boeing Company
Abstract
This research investigates
technology
flexibility, which is
the
technology
characteristic that allows or enables
adjustments and other changes to the business process.
Technology flexibility
has two dimensions, structural and
process flexibility, encompassing both the actual
technology
application and the people and processes that
support it. The flexibility of
technology
that supports
business processes can greatly influence the organization's
capacity for change. Existing
technology
can present
opportunities for or barriers to business process flexibility
through structural characteristics such as language
,
platform,
and design.
Technology
can also indirectly
affect flexibility through the relationship between the
technology
maintenance organization and the business
process owners, change request processing, and other
response characteristics. These indirect effects reflect a
more organizational perspective of flexibility.
This
paper
asks the question, "what makes
technology
flexible?" This question is addressed by developing and
validating a measurement model of
technology
flexibility.
Constructs and definitions of
technology
flexibility are
developed by examining the concept of flexibility in other
disciplines, and the demands imposed on
technology
by
business processes. The purpose of building a
measurement model is to show validity for the constructs
of
technology
flexibility.
This paper discusses the theory
of technology flexibility, develops constructs and
determinants of this phenomenon, and proposes a
methodology for the validation and study of the flexibility
of emerging technologies.
1. Introduction
Organizations are facing an environment of ever
increasing turbulence and change. Shorter product cycle
times, global competition, an increasing regulatory
environment, and constant demands to reduce and control
costs require dynamic, flexible business processes.
Business process flexibility can be a key factor in an
organization's ability to adapt and compete [1], [2], [3].
Increased flexibility can give an organization competitive
advantage through faster response to customer needs and
environmental conditions [4]. The flexibility of technology
that supports business processes can greatly influence the
organization's capacity for change [5]. Emerging
technologies need to enable changes in the business user
environment [6]. This paper investigates technology
flexibility, which is the characteristics of technology that
allow or enable adjustments and other changes to the
business process.
One of the first places that many organizations have tried
to gain flexibility through the use of technology is in the
manufacturing area [7], [8], [9]. The primary way that this
has been accomplished is through investments in computer
integrated manufacturing (CIM). The lessons learned
from these CIM projects yield insights into the roles and
interaction of technology and people in attaining process
flexibility.
In a recent study of 61 North American paper mills,
managers rated 40% of flexibility-improvement efforts to
be unsuccessful or disappointing [10]. The primary cause
of these disappointments was the reliance on technology
alone to provide flexibility. Upton found that in reality, the
flexibility of these plants depended much more on the
people than the technology. It was the interaction and
alignment of technology and people that produced
flexibility in successful CIM projects. This example
demonstrates that the combination of technology and
people is key to achieving flexibility.
In this paper, technology is discussed in terms of the
technology application itself and the maintainers and users
of the application and the management processes they use
to support the application. In the past, technology
flexibility has been evaluated from a purely computer
science perspective [11], [12], only taking the technology
itself into consideration. Technology has often made
organizations more rigid rather than more flexible by being
too time consuming to redesign or reflecting obsolete world
views [13]. Existing technology can present opportunities
for or barriers to business process flexibility through
structural characteristics such as language and design [11].
Technology can also indirectly affect flexibility (either
positively or negatively) through the relationship between
the technology organization and the business process
owners, change request processing, and other response
characteristics. These indirect effects reflect a more
organizational perspective of flexibility [14], [15].
This paper combines the computer science and
organizational perspectives to ask the question, "What
makes a technology flexible?" This question is addressed
by developing a measurement model of technology
flexibility. First we describe the need for flexibility in the
business process environment. The literature on software
engineering and measurement is then examined for
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE

contributions to the measurement of flexibility. Constructs
and definitions of technology flexibility are then developed
by examining the concept of flexibility in other disciplines.
Measures of flexibility adapted from behavioral psychology
are used as a foundation for conceptualizing measures of
technology flexibility. A methodology for ensuring the
validity of these constructs is then discussed. Interviews
with I/S managers, system maintainers, users, and business
process owners are suggested as a means to test measures
drawn from the literature and to construct a measurement
model.
2. The business process change environment as
a driver of technology flexibility
This paper addresses emerging technologies that support
business processes. Davenport [16] defines a business
process as "a structured set of activities designed to
produce a specified output for a particular customer or
market" (p.5). Harrington [3] defines a business process as
a group of logically related tasks that use the resources of
the organization to produce results. These tasks work
interdependently to consistently produce a specified result
[1]. The business process is defined as:
a specific set of interdependent tasks
that produce a specified output for a
particular customer or market.
This section explores the business process environment
in terms of incremental and revolutionary change.
Business processes are subject to ever increasing
environmental pressures to change [3]. Organizations
change both continuously and discontinuously, resulting in
different types of pressures in the business process
environment [17]. Understanding this environment is
critical to the study of technology flexibility [18] as it is a
determining condition of the nature of change required of a
technology that supports a business process.
The results of environmental pressures to change have
been categorized as long periods of relative organizational
stability followed by shorter periods of revolutionary
change [19], [20], [21]. Turbulent environments have the
potential to both reorganize organizational structure and to
add, change, or delete business processes. Once new
organizational structures and activities are in place, a cycle
of continuous improvement begins. Similarly, the
radically changed or replaced technology begins a cycle of
continuous improvement to support these new business
processes. Technologies that function well in stable
periods of incremental change are often unable to survive
revolutionary change [5]. Technologies need to be flexible
enough to function well in both types of environments.
Adaptability and Adaptation.
Turbulent changes in the
environment can result in the need for rapid, radical
changes in ways of doing business [19], [22], [17]. These
changes can take a complacent management team by
surprise, and force it to engage the unfamiliar. The
characteristic of adaptability allows organizations to
engage the unfamiliar. Huber [23] defines adaptability as
"the capacity to expand niches or to find new niches" (p.
940). Adaptability is often the result of or reaction to
turbulence in the environment. One characteristic of
adaptability is a willingness to engage an unfamiliar
environment [23]. Adaptable organizations seek
opportunities in times of environmental turbulence. They
are prepared to take on the unknown and incorporate the
unfamiliar. Technology that supports these types of
organizations must have this ability built into its structure
to support the results of organizational adaptation. The
data and functions of the adapted organization must be
readily assumed and obsolete ones easily discarded. The
capacity to change should be an integral part of the
original technology design.
Huber [23] describes a second characteristic of
adaptability is the ability to scan the external environment
for expansion into new niches. Not only is the
organization prepared to engage the unfamiliar -- it seeks
it. Scanning the environment forces organizations to
examine the current business and to discard niches that are
no longer profitable or feasible. The business is often
redefined. Unfortunately, much organizational technology
has been developed or refined expressly for existing niches.
A test of technology adaptability is its ability to withstand
this external scanning and to accommodate new business
niches that are adopted.
In the MIS literature, Allen and Boynton [24] describe
adaptable systems as dynamically stable information
systems. These systems have a stable base of capabilities
and at the same time are flexible in the long term. They
possess high levels of modularity, applicability, reusability,
re-combinability, and are open to links with other systems.
The structure of these systems provides the capability to
dramatically change, while customized modules can be
incrementally improved for increased applicability to a
specific business process. Hedberg and Jonsson [24]
maintain that information systems should incorporate the
capacity to predict the future along with the ability to cope
with unexpected developments in the current environment.
This implies that both the ability to deal with revolutionary
future changes, as well as unanticipated incremental
changes in the current environment are critical for
technology flexibility.
While adaptability characterizes revolutionary changes in
the business process environment, adaptation, as defined
by Huber [23], is the optimization of a particular niche or
business process. Business process improvement (BPI) and
total quality management (TQM) are examples of
initiatives that emphasize adaptation. These processes
stress incremental change and improvement [24]. These
requirements call for a definition of technology flexibility
that encompasses both adaptability and adaptation.
Flexibility, to be employed in a realistic manner, should
result in “little penalty in time, effort, cost, or
performance" [10 p. 73]. According to De Groote [26],
flexibility is defined in terms of yielding the best possible
performance in the face of environmental variability. In
the case of the technology, the business process
environment contains both incremental and revolutionary
variability. This research incorporates these ideas in a
definition of technology flexibility:
The ability to adapt to both incremental
and revolutionary change in the
business or business process with
minimal penalty to current time, effort,
cost, or performance.
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE

The following section examines the dimensions of
technology flexibility. To understand these dimensions,
definitions of flexibility from the MIS and other literatures
are examined for context and applicability.
3. Dimensions of technology flexibility
Definitions of flexibility vary considerably within and
between disciplines. The information systems literature
has examined flexibility in several contexts. Silver defines
flexibility as the opposite of restrictiveness [27], [28], [29].
He identifies factors that impact the design of less
restrictive and hence more flexible decision support
systems. These factors include providing a broad
repertoire of decision support tools, supporting multiple,
new, and/or changing decision-making environments, and
providing opportunities for creativity and learning. These
ideas can be extended to technologies other than decision
support systems. Does the existing system have the
breadth to support change or growth in the business
process environment? Has the ability to adapt to creative
changes or new learning been designed into the technology
product and process? Differences in flexibility can be due
to variations in design and construction of systems or
caused by changing hardware or business environments
[11]. Although both the work of Silver and Schwan and
Jones discusses characteristics of information system
flexibility, no formal dimensions of the concept are
developed. However, several examples of dimensions of
flexibility are found in the information systems, behavioral
psychology, manufacturing, and organizational literature.
Dimensions of Flexibility.
There are three conceptual
senses of flexibility in Information Technologies (IT):
flexibility in functionality, flexibility in use, and flexibility
in modification [30]. Flexibility in functionality and
modification address the response of an IT to incremental
change or variability. Flexibility in use addresses
incremental change, but also addresses the ability to
encompass new relationships and opportunities;
characteristics of revolutionary change. It implies that this
capacity for change is built into the system. These IT
dimensions of flexibility are closely linked to
environmental pressures on organizations. Flexibility is
shown as a multidimensional concept, which provides a
starting point for conceptualizing technology flexibility.
While the three dimensions proposed by Knoll and
Jarvenpaa encompass the ability of technology to adjust to
both incremental and dramatic change, these three
dimensions all address the technology itself, in this case,
the IT application. The people and process side of
technology flexibility is not specifically addressed.
In behavioral psychology, Scott [31] defines flexibility
using a two dimensional approach. The first dimension of
flexibility is the degree of response variability found in an
organism. Response variability is defined the degree of
diversity in reactions shown by a particular person under
normal, everyday conditions, which are the type of
conditions under which incremental change occurs. The
second dimension of flexibility is responsiveness to
environmental pressures to change. People are judged on
whether they have the capacity to adapt to dramatic
changes in the environment, such as moving to a new
home or responding to a new caretaker. This differs from
response variability in that it is seen as an internal capacity
to accommodate dramatic environmental change rather
than a reaction to ongoing, incremental change such as
new activities being added to a daily play schedule. This
internal ability to accommodate change can be represented
by a technology having a design that allows for rapid
additions, deletions, or restructuring in periods of dramatic
change. The ability to react to ongoing change can be
represented in a technology by processes and procedures
designed to handle change.
In the manufacturing flexibility literature, Garud and
Kotha [32] propose that the “time is right” (pg. 673) for a
broader perspective of flexibility. They suggest using a
systems approach [1] [33] for adopting social and technical
dimensions of flexibility. This perspective incorporates
both the technical aspects of the manufacturing equipment
and facility with human aspects such as job design,
management organization, work-team structure, selection
and training, and compensation and appraisal. These
aspects of flexibility are theorized as interacting in a
dynamic network, although the form of these interactions
is not specified.
The organizational literature also provides dimensions of
flexibility [15]. The conceptual framework of strategic
organizational flexibility is based on two dimensions; ex
ante and ex post. Each of these dimensions has offensive
and defensive characteristics associated with it. The ex
ante dimension anticipates change before it happens.
Agility and versatility are offensive ex ante characteristics
that provide a repertoire for dealing with novel or
unexpected situations. Robustness and hedging are
defensive ex ante characteristics that seek to deflect or
avoid the unexpected. These dimensions reflect a built-in
capacity to anticipate and deal with change. The ex post
dimension of strategic organizational flexibility has both
offensive and defensive characteristic that are incremental
in nature. Liquidity and elasticity are offensive
characteristics that allow continual change with the
environment. Corrigibility and resilience are defensive
characteristics that aid a system in recuperating or
returning to functionality after a change. This dimension
of flexibility reflects an ongoing ability to deal with
change. This two dimensional definition of flexibility is
consistent with the two types of environmental changes to
which technologies must adapt. Liquidity and elasticity
would allow technologies to incrementally change.
Corrigibility and resilience would allow technologies to
dramatically change and to recover from these changes.
While this two dimensional approach includes both types
of environmental change, it does not directly address the
components of a technology; the application itself and the
people who support it.
Huber and McDaniel [14] define organizational
flexibility as "the ease with which the organization's
structures
and
processes
can be changed" (p. 583). This
definition can be applied to technologies. The structure of
a technology can be viewed as the design and organization
of the programs and data contained in the technology
application. The processes of a technology may include
the management and technical processes and procedures
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE

used to maintain the application. These two dimensions of
flexibility can combine to impact the overall flexibility of
the technology.
The dimensions of structure and process are particularly
relevant for describing the flexibility of a technology that
includes both the technology itself and the people who
support it. The dimensions of structure and process
interact in a similar fashion to the technical and social
dimensions proposed by Garud and Kotha [32]. For
example, a technology application can be written in a way
that makes changing a data element a relatively quick and
easy task. However, the change request process that
initiates this change could be bureaucratic and time
consuming, decreasing the overall flexibility of the entire
technology. This two dimensional conceptualization of
flexibility is appropriate for technology as it captures both
the technological and social parts of the system. Each of
these individual dimensions, as well as the interaction of
the two dimensions of flexibility impact system flexibility.
Utilizing Huber and McDaniel’s [14] definition of
flexibility, this research defines the dimensions of
technology flexibility as structural and process flexibility.
Structural flexibility reflects the ability of the design of a
technology to be adapted to changes in the business process
and is pro-actively designed into the technology.
Structural flexibility is the capability of
the design and organization of a
technology to be successfully adapted to
business process changes.
The dimension of process flexibility refers to a the ability
of a technology's technical and management processes to
accommodate change.
Process flexibility is the ability of
people to make changes to the
technology using management
processes that support business process
changes.
The structure of a technology is viewed as it’s design and
organization. The processes of a technology include the
management and technical processes and procedures used
to maintain and change the application. These two
dimensions combine to impact the overall flexibility of the
technology. We propose that both of these constructs must
be present to completely capture technology flexibility.
4. Determinants of structural flexibility
What are the characteristics of technology structural
flexibility? The dimension of structural flexibility has
three determinants that represent the structural
characteristics and organization of the technology:
modularity, change acceptance, and consistency. These
determinants have been derived from determinants of
psychological structural flexibility found in the behavioral
psychology literature, and chosen for their applicability to
technologies. Garud and Kotha [32] use the brain as a
metaphor for modeling flexible production systems.
Modularity.
In behavioral psychology, the number of
different arrangements that a subject can perform by
moving an object or objects into different forms or patterns
reflects cognitive flexibility on the part of the subject [31].
Using this characteristic as guide, we examine the
modularity present in a technology.
One way to achieve technology flexibility is through
modularity [11], [34]. This approach has been used with
success by corporations such as Mrs. Field’s Cookies [35].
By structuring technology with smaller modules, changes
that involve adding or removing functions is simplified.
The technology can potentially support a greater number of
arrangements and modifications. The structuredness of a
system is often a reflection of the tools or methodologies
used to develop it. Modularity is defined as:
The degree of formal design separation
within a technology application.
Modularity can contribute to flexibility by providing
manageable units of programs or hardware that can be
modified as business processes change as well as the ability
to easily create or destroy modules [36]. For example,
software modularity is measured by the degree to which
programs are grouped to obtain a high degree of functional
relatedness [36]. However, it is possible that modularity is
not sufficient to encompass changes that require
unforeseen functionalities. More advanced types of
technologies such as object oriented, neural networks and
artificial intelligence have the potential to expand
structural flexibility past the modularity of procedural
programming [34].
Change Acceptance.
The acceptance of pressures to
change is another indicator of psychological
responsiveness to the environment. In human beings, this
is measured by the subject's ability to "cope" with change
[31]. Can a person accept change and to what degree
(coping) is this change integrated into the person's
makeup? The magnitude of aftereffects of change is an
indicator of coping ability. Human beings are equipped
with various amounts of ability to cope with or accept
change. Likewise, technology is written pro-actively with
various amounts of built-in restrictiveness or flexibility
[29]. This restrictiveness or flexibility may be
intentionally or unintentionally a part of the technology
design. An example of an intentional design that enhances
change acceptance is the inclusion of data control tables
that users can access. An example of unintentional
restriction of flexibility are older systems which do not
accept dates past the year 1999. Change acceptance is
defined as:
The degree to which a technology
contains built-in capacity for change.
Features such as reusable code and object libraries should
have a positive impact on change acceptance.
Consistency.
The ability of a person to integrate
different psychological regions of thinking when solving
puzzles or problems indicates cognitive flexibility [38].
Psychology assumes that different concepts will remain in
independent cognitive regions until the biological-
psychological system integrates them. The greater the
integration and number of solutions, the greater the ability
to adapt to change. The ability of data and components to
be integrated consistently across a technology application
can greatly affect flexibility. If data that is stored
independently in the application is integrated and the
result is inconsistent representations or data elements, the
ability to change data becomes very complex. This
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE

complexity will impact how much of the overall product
needs to be changed or adapted to support business process
changes. Data that is maintained in one location and
integrated consistently throughout the system can be more
readily changed. This is especially significant in the case
of reverse engineering of technology [34], where an
existing application is dismantled to accommodate major
change in structure, language, or platform integration.
Technologies that have a high consistency should be
more adaptable to changes in the business process
environment. Consistency, the third determinant of the
technology dimension, is defined as:
The degree to which data and
components are integrated consistently
across a technology.
Modularity, Change Acceptance, and Consistency are all
determinants of structural flexibility.
5. Determinants of process flexibility
The second dimension of technology flexibility is process
flexibility. The three determinants of this dimension are
rate of response, expertise, and coordination of action.
The personnel who maintain and support technology and
the management and change processes they use affect
technology flexibility. A technology’s structure may be
highly flexible, but if the management processes
supporting it are rigid, overall flexibility will be impacted.
The effectiveness of people in reacting to, implementing,
and performing change enhances or impedes technology
maintenance [39], [40], [41]. It is therefore proposed that
maintainer and user effectiveness in addressing change,
and the quality of processes that support change also
characterize the overall flexibility of the system.
Rate of Response.
Rate of reversal [42] is one way of
measuring the rate of response to change. In a human being,
this refers to both the physical and mental abilities needed to
transition from one task or perception to another. The faster
the rate of transition, the less rigid and more flexible the
individual. Similarly, the speed of transition of a technology
contributes to flexibility. This rate of response is impacted by
the people and management processes employed in changing
technology. It is possible for a change that takes two hours to
program to be caught up in a change request process for more
than a week.
Rate of response is often measured as cycle time in
technology maintenance. When this operationalization is
used, it is important to consider both the actual time it
takes to make the change and the time involved in
approving or processing the request for change. The
response of the technology group to the users is often
impacted by process characteristics such as prioritization,
limited resources, or approval processes which are
unrelated to the amount of time it takes to actually perform
a change. This is the process rate of response that is
captured in this research, which includes all of the time it
takes to make changes, not just the actual labor hours
required. Rate of response is defined as:
The degree to which changes can be
made to a technology in a timely
manner.
Expertise.
The expertise determinant of technology
flexibility is derived from the psychological concept of
facilitation of communication. Facilitation of
communication is the ability of a subject to articulate
knowledge about an object or circumstance, and to
construct inquiries leading to additional knowledge of the
area in response to variations in stimuli [42]. In humans,
this construct is measured through a range of descriptive
behaviors ranging from pre-verbal expression to complex
instructions and conversations. The ability to
communicate knowledge about a technology and to
construct inquiries that lead to new knowledge about the
system is a result of the level of expertise in the
maintainers and users of the technology. It is also a result
of the "expertness" of the documentation and management
processes of the system. Accurate version control and
complete documentation allow more rapid and accurate
response to incremental changes in the business process.
The system becomes less dependent on the availability of
expert individual maintainers and is less susceptible to
personnel turnover. Expertise is defined as:
The degree to which up-to-date
knowledge about the operation and
maintenance of a technology exists and
is communicated.
Maintainers of technologies are often dependent on the
quality of documentation about a technology application,
especially if the maintainers are not the developers of the
application. Similarly, the maintainers also impact the
quality and control of technology documentation by
performing or not performing updates to these documents.
Well documented technologies are more flexible and easy to
change because the development process and logic is more
readily traced and can be understood by someone other than
the initial developer. In the same respect, accurate version
control allows maintainers to trace changes already made in
the system and their potential impacts. While it is important
that personnel be skilled in and knowledgeable about the
technology themselves, it is equally important that they have
expertise in maintaining the standards and procedures of the
organization and in the business domain they support. This
level of expertise is often a function of the training and
assessment procedures present in the organization [34].
Coordination of Action.
Georgopoulos and Mann [44]
define coordination as the extent that interdependent parts
of an organization function according to the needs and
requirements of the other parts and of the total system.
The interdependent organizations that need to coordinate
in a technology are the users and developers and
maintainers of the application [35]. When these groups
coordinate successfully, flexibility can be enhanced.
Barriers to flexibility can arise when changes to the
technology are not agreed upon or mutually understood
[45]. Standards agreed to and used by both groups can
enhance coordination of action. Another characteristic of
coordination is the ability of the user group to
communicate changes in the business process to the
information systems developers and maintainers. In this
way, coordination can lead to flexible technologies that
adapt to the business processes. Coordination of action is
defined as:
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE

The degree to which the technology
maintenance and user organizations
operate according to the requirements
of each other and the total
organization.
Rate of Response, Expertise, and Coordination of Action
are the determinants of the dimension of process flexibility.
6. Validating the measures of technology
flexibility
Measurement is often viewed as a process in which
numeric values are assigned to objects or occurrences
according to a specified set of rules. In this view, it is a
strictly empirical process, distinct from the theory being
utilized in the study. Bagozzi [47], however, proposes an
alternative view of measurement that links it holistically to
theory. In this view, measurement is seen as an
intellectual and empirical way of giving meaning to theory.
Well done measurement validation is a key to performing
successful MIS research [48]. Rigorous measurement and
validation provide the foundation for evidence of inference
[49]. In a review of MIS survey research, Newsted and
Munro [50] found that very few studies adequately
demonstrated validity and reliability.
Carmines and Zeller [51] regard measurement as the
linking of abstract concepts to empirical determinants.
These constructs are often not directly observable, but are
believed to be latent in the phenomenon to be studied [47].
Ives and Olson [52] provide a good example of this,
defining system success as a latent construct. Flexibility is
another example of a latent construct. These constructs are
not directly observable or measurable. To understand
latent variables, researchers develop measures such as Ives
and Olson's "User Information Satisfaction" scale to
capture the construct empirically.
Measurement models are used to describe how well
observed indicators serve as a measurement instrument for
latent variables [53]. Measurement models of data are
specified by the researcher based on theory. Confirmatory
factor analysis (CFA) is used when the research design
hypothesizes dimensions a priori. In CFA, precise
descriptions of latent variable structure and indicator
loadings are specified by the researcher. CFA requires the
use of analysis of covariance structure [54]. This
technique is the basis of the LISREL approach to data
analysis that we use to test the measurement model of
software system flexibility.
The testing of the validity of the measures of latent
theoretical constructs in a measurement model must be
explicitly defined and specified. Bagozzi, Yi, and Phillips
[55] provide an explicit criteria for measuring the validity
of a measurement scheme as an operationalization of a
specified theory. This approach is a holistic one, in that
the validity of the measurement instrument and the
underlying theory are tested simultaneously. The six
components of construct validity as prescribed by Bagozzi
[47] are theoretical meaningfulness of concepts,
observational meaningfulness of concepts, internal
consistency of operationalizations, convergent validity,
discriminant validity, and nomological validity.
Theoretical and observational meaningfulness of
concepts evaluate the internal consistency of language used
to describe a construct and the conceptual relationship
between a construct and its operationalization. This
evaluation does not include any specific statistical tests, but
looks at the semantic unity of the construct and
operationalization. "The
theoretical meaningfulness
of a
concept refers to the nature and internal consistency of the
language used to represent the concept" [47, p. 117]. To
be "meaningful", the terminology used to describe a
construct must describe the scope or range of the construct.
The
observational meaningfulness
of concepts refers to
the relationship between unobserved latent theoretical
variables and their observed indicators. Evaluating the
criterion of observational meaningfulness of concepts
includes assessing whether the questions used as indicators
for each construct are clear, unambiguous, and related to
the construct.
Theoretical and observational meaningfulness of
concepts assess the semantic validity and quality of the
relationship between the theoretical constructs and
observable indicators used to measure those constructs.
The
internal consistency of operationalizations
assesses
the homogeneity of indicators. This homogeneity has two
components; unidimensionality and reliability. Assessing
unidimensionality tests whether all the indicators measure
the theoretical construct of interest [55]. When measuring
a single construct, multiple indicators must lie on the same
dimension. Any multidimensionality present indicates that
more than one construct is present and violates construct
validity [47].
Reliability is defined as the extent to which measures are
free from random error components and yield consistent
results. Reliability assesses the proportion of the indicator
variance attributable to the underlying construct. The
Cronbach alpha coefficient [56] is often used to compute
reliability. This statistic computes reliability across a set of
indicators of a single theoretical construct and provides a
lower bound of reliability. There are, however, some
limitations to this method. Cronbach alpha calculations
assume that all indicators are equally important [47],
which may not be justified in cases when some indicators
measure a construct better than others. Joreskog's [57]
Analysis of Covariance Structures overcomes this problem
by testing for both unidimensionality and reliability. This
method is used for assessing reliability in this study rather
than Cronbach alpha, since it does not assume the equality
of importance across indicators.
Convergent validity
is the degree to which two or more
measures of the same theoretical construct are in
agreement.
Discriminant validity
is the degree that one
theoretical construct differs from another. Joreskog [57]
provides a procedure that simultaneously assesses
convergent and discriminant validity using confirmatory
factor analysis. Using CFA, one tests the hypothesis that
each indicator loads only onto its associated theoretical
construct by fixing the factor loadings of indicators onto
other theoretical constructs to zero.
The last type of measurement validity is
nomological
validity
. This refers to the drawing of inferences from
constructs from the fit between patterns of data, otherwise
known as the nomological net [58].
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE

7.
Testing the technology flexibility measurement
model
The technology flexibility measurement model was tested
using the example of software flexibility. The following
sections describe the testing procedure.
Discovery Phase. The purpose of the discovery phase is
to develop and perform preliminary testing of questions
and instruments to be used in the testing of technology
flexibility, using the dimensions of structure and process.
Questions can be developed in two ways. Candidate
questions are derived from the flexibility literature in other
disciplines. Operationalizations are then examined for
relevance to technology, and modified accordingly.
Secondly, a series of interviews is conducted with
maintainers and users of the technology to be studied.
These interviews are conducted to explore the theoretical
and observational meaningfulness for the questions drawn
from the literature and to generate additional questions.
Appendix A contains an instrument used to measure
software flexibility that was generated using the process
described herein.
Generation of Candidate Questions. We have shown
that flexibility has been extensively measured in the
behavioral psychology literature. Candidate questions used
in the software flexibility measurement instrument were
developed based on the work of Scott [31] and questions
from the California Psychological Inventory [59]. The
manufacturing engineering literature provided another
source of candidate questions, generated from the work of
Gerwin [7], [8] and Upton [9]. The organizational
literature was also a source of potential questions.
Structural questions are derived from Blau and
Schoenherr's [60] number of major subunits questions, and
coordination questions were suggested by Georgopoulos
and Mann [44]. Based on these questions, a discovery
interview instrument was developed.
Sample. Respondents in the discovery phase of this type
of research should represent the same organizations as the
respondents in the testing phase, but be an independent
sample. In the case of measuring software flexibility, 28
discovery interviews were completed with respondent’s
answers recorded both in written and taped forms.
Interview Methodology. The discovery interviews were
conducted using open interview techniques with probes
[61]. Respondents were asked what characteristics of
software systems most supported business processes.
Based on the response to this question, probing questions
were asked to elicit further attributes specifically
concerning structural and process flexibility. The results
of these interviews were compared to the questions derived
theoretically and were modified to reflect the data gathered
from respondents. The result of the discovery interviews
was eight to ten questions for each of the six indicators of
software system flexibility (change acceptance,
consistency, modularity, rate of response, coordination of
action, and expertise).
Pretest Sort. A pretest sort of the candidate indicators
was conducted with three Ph.D. students, two faculty
members, and fifteen managers from six participating
organizations. A list of questions was presented to subjects
with a separate list of constructs. Subjects were asked to
sort questions by construct. Based on the results of these
sorts, questions were reworded or deleted from the
candidate list. A pilot questionnaire was then developed
from the remaining list of indicators.
Pilot Instrument. Two to eight respondents representing
both users and maintainers from four of the participating
organizations filled out the pilot questionnaire. A total of
23 pilot questionnaires were collected. The pilot
questionnaire had two sections: structural flexibility and
process flexibility. Each question was rated on a seven-
point Likert scale. After completing the pilot
questionnaire, each respondent reviewed each question
with the researcher for content, clarity, and meaning.
Through these interchanges, candidate questions were
further refined and selected for inclusion in the final
research instrument. Based on the results of the pilot test,
at least four questions for each characteristic of flexibility
were retained for the testing phase of the study. The
criteria for keeping questions was clarity, meaningfulness,
ability to measure the construct, and understandability.
Testing Phase. Once instruments were developed, the
measurement model of software system flexibility was
tested by collecting data was collected from both users and
maintainers of software systems, as well as from I/S
managers and the user managers supported by the
software. The basic design of this research is a cross-
sectional field study. The sample is a heterogeneous set of
representative business processes and software systems.
This sample was drawn from twelve organizations that
showed an interest in participating in this study. These
organizations were chosen for industry diversity, ease of
data collection, and availability of metrics information,
making this a convenience sample.
Data Collection. The primary research instrument for
the collection of software system flexibility data was a
questionnaire asking about the determinants of software
system flexibility. This questionnaire contains multiple
questions for each of the six indicators of software system
flexibility, all of which were tested on the pilot
instruments. This instrument uses a seven point Likert
Scale numbered from one to seven. The instrument was
administered to groups representing 116 software systems
across 12 organizations. For each represented software
system, information was collected from expert respondents:
one or two maintainers and one or two users of the system.
These expert informants each provided knowledge of the
nature and role of the software system in use and the
purpose and nature of the business process. Indicators of
software system flexibility were tested for validity using
respondents from both the I/S and user sides of the system.
This paper describes a theoretical approach to measuring
technology flexibility, using the above described
measurement of software flexibility as an applied example.
We believe that this approach can be used to measure the
flexibility of a variety of emerging technologies, combining
the constructs developed with the validation and
construction processes described.
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE

8. Conclusion
The concept of technology flexibility as having two
dimensions has theoretical, methodological, and practical
implications. Theoretically, this two dimensional
approach gives researchers a new way of approaching the
problem of flexible technologies. The computer science
and organizational approaches can be combined to yield a
more complete view of the phenomenon. The antecedents
of both structural and process flexibility can be examined
more closely for a further understanding of technology
flexibility.
The primary methodological contribution of this study is
the development of a measurement framework for
technology flexibility. This framework gives researchers
an empirical starting point for examination of technology
flexibility and related constructs. The use of discovery
interviews, pre-test sorts and pilot tests can result in a
strong, empirically supported measurement model of
technology flexibility. The dimensions and indicators of
this model can be used as foundations for future
researchers.
From a practical standpoint, this approach provides
technology developers with a new perspective of how to
build and maintain more flexible technologies. With an
understanding of technology flexibility as a two
dimensional phenomenon, developers can design
technology to incorporate both structural and process
flexibility. This will extend the work being done in the
areas of structured programming [37} and object oriented
design.
This theory suggests that management that desires overall
technology flexibility but only considers structural flexibility
from technology may experience disappointment.
Organizations seeking flexibility through technology need to
consider it as a system containing the technology application,
the people that maintain and support the application, and the
management processes that these people use to accomplish
their work. By using the two dimensional approach to
technology flexibility, organizations can explore several
alternatives. Existing technologies can be examined to see if
they can be modified for structural flexibility or if personnel
can be given increased training to gain process flexibility.
Change request processes and coordination between
technology maintainers and users can be examined for
potential gains in process flexibility. It may be less
expensive to gain flexibility by providing training and doing
team building for maintainers and users of an existing system
than by purchasing or developing new technology. Both
types of technology flexibility can be considered when
making both cost and strategic decisions.
It is possible that many of the disappointments with
emerging technologies may be a result of a failure to consider
process flexibility. Upton [10], in the case of computer
integrated manufacturing, states that, "Not only is computer
integration not the panacea for flexibility problems but it also
comforts managers with the thought that they are doing
something, when all along they should have been doing
something else" (pg. 83). This something else , in the case
of emerging technologies, is process flexibility.
Emerging technologies must be able to change both
incrementally and dramatically to keep up with changes in
the business environment. By taking into account the
interaction and alignment of structural and process
flexibility, designers, maintainers, and users of emerging
technologies can use these technologies to contribute to
overall organizational success.
References
[1] Thompson, J. D., Organizations in Action, McGraw
Hill, Inc., New York, N.Y., 1967
[2] Davenport, T. H. and Short, J. E., "The New Industrial
Engineering: Information Technology and Business
Process Redesign", Sloan Management Review, Summer
1990, pp. 11-27
[3] Harrington, H. James, Business Process Improvement,
McGraw-Hill, Inc., New York, NY, 1991
[4] Fiegenbaum, A. and Karnani, A., "Output Flexibility -
A Compettitve Advantage for Small Firms", Strategic
Management Journal, Vol. 12, No. 2, February 1991, pp.
101-115
[5] Hammer, M. and Champy, J., Reengineering the
Corporation: A Manifesto for Business Revolution, Harper
Collins Publishing, May 1993
[6] Ahituv, N. and Neumann, S., The Principles of
Information Systems Management, Third Edition, William
C. Brown Publishers,Dubuque, Iowa,1990
[7] Gerwin, D., "An Agenda for Research on the
Flexibility of Manufacturing Processes," International
Journal of Operations and Production Management, Vol.
7, No. 1, 1987
[8] Gerwin, D., "Manufacturing Flexibility: A Strategic
Perspective", Management Science, Vol. 39, No. 4, April
1993, pp. 395-410
[9] Upton, D. M., "The Management of Manufacturing
Flexibility", California Management Review, Vol. 36,, No.
2, Winter 1994, pp. 72-89
[10] Upton, D. M., "What Makes Factories Really
Flexible?", Harvard Business Review, Vol. 73, No. 4, July-
August 1995, pp. 74-86
[11] Schwan, K. and Jones, A. K., "Flexible Software
Development for Multiple Computer Systems," IEEE
Transactions on Software Engineering, Vol SE-12, No. 3,
March 1986, pp.385-401.
[12] Gibson, V. R. and Senn, J. A., "System Structure and
Software Maintenance Performance", Communications of
the ACM, Vol. 32, No. 3, March 1989, pp.347-358
[13] Hedberg, B., and J., Sten, "Designing Semiconfusing
Information Systems for Changing Organizations," Data
Base, Winter-Spring 1982, pp. 12-25.
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE

[14] Huber, G. P. and McDaniel, R. R., "The Decision-
Making Paradigm of Organizational Design", Management
Science, Vol. 32, No. 5, May 1986, pp. 572-589.
[15] Evans, J. S., "Strategic Flexibility for High
Technology Manoeuvres: A Conceptual Framework",
Journal of Management Studies, Vol. 28, No. 1, January
1991, pp. 69-89
[16] Davenport, T. H., Process Innovation: Reengineering
Work through Information Technology, Harvard Business
School Press, Boston MA, 1992
[17] Meyer, A. D., Goes, J. B., and Brooks, G. R.,
"Organizations Reacting to Hyperturbulence", pages 66-
111, in Huber, George P. and Glick, William H.,
Organizational Change and Redesign, Oxford University
Press, New York, 1993
[18] Emery, F.E. and Trist E.L., "The Causal Texture of
Organizational Environments", Human Relations, Vol. 18,
1965, Pages 21-32
[19] Tushman, M. and Romanelli, E., "Organizational
Evolution: A Metamorphosis Model of Convergence and
Reorientation", in Cummings, L.L. and Staw, B.M. (Eds.),
Research in Organizational Behavior, vol. 7, 1985, JAI
Press, Greenwich Conn., pp. 171-222
[20] Gersick, C. J. G., "Revolutionary Change Theories: A
Multilevel Exploration of the Punctuated Equilibrium
Paradigm", Academy of Management Review, vol. 16, no.
1, 1991, pp. 10-36
[21] Romanelli, E. and Tushman, M. L., “Organizational
Transformation as Punctuated Equilibrium: An Empirical
Test,” The Academy of Management Journal, Vol. 37, No.
5, October 1994, pp. 1141-1166
[22] Miles, R. E. and Snow, C. C., "Organizations: New
Concepts for New Forms," California Management
Review, Vol. 18, No. 3, Spring 1986, pp. 62-73
[23] Huber, G.P.,, "The Nature and Design of Post-
Industrial Organizations", Management Science, vol. 30,
no. 8, 1984, pp. 928-951.
[24] Allen, B. R. and Boynton, A. C., "Information
Architecture: In Search of Efficient Flexibility", MIS
Quarterly, Vol. 15, No. 4, December 1991, pp. 435-446
[25] Robson, G. D., Continuous Process Improvement, The
Free Press, New York, NY., 1991
[26] De Groote, X., “The Flexibility of Production
Processes”, Management Science, Vol. 40, No. 7, July
1994, pp. 933-945
[27] Silver, M. S. (1), "On the Restrictiveness of Decision
Support Systems", Organizational Decision Support
Systems, 1988, pp. 259-270
[28] Silver, M. S., "User Perceptions of Decision Support
System Restrictiveness: An Experiment", Journal of
Management Information Systems, Vol. 5, No. 1, Summer
1988, pp.51-65
[29]Silver, M. S., "Decisional Guidance for Computer-
Based Decision Support", MIS Quarterly, March 1991,
pp.105-122
[30] Knoll, K. and Jarvenpaa, S. L., "Information
Technology Alignment of "Fit" in Highly Turbulent
Environments: The Concept of Flexibility", Proceedings of
the 1994 SIGCPR Conference, Alexandria, VA, March 24-
26, 1994, pp. 1-14
[31] Scott, W. A., "Flexibility, Rigidity, and Adaptation:
Toward Clarification of Concepts" in Harvey, O.J. (ed.),
Expeience, Structure and Adaptability, Springer
Publishing Company, Inc., New York, 1966.
[32] Garud, R. and Kotha, S., “Using the Brain as a
Metaphor to Model Flexible Production Systems,” The
Academy ofManagement Review, Vol. 19, No. 4, October
1994, pp. 671-698
[33] Van de Ven, A. H. and Drazin, R., "The Concept of
Fit in Contingency Theory", Research in Organizational
Behavior, Vol. 7, 1985, pp.333-365
[34] Premelani, W. J. and Blaha, M. R., “An Approach for
Reverse Engineering of Relational Databases”,
Communications of the ACM, Vol. 37, No. 5, May 1994,
pp. 42-53
[35] Haeckel, S. H. and Nolan, R. L., “Managing by
Wire,” Harvard Business Review, Vol. 71, No. 5,
September-October 1993, pp. 122-133
[36] Yourdon, E., Managing the Structured Techniques:
Strategies for Software Development in the 1990's,
Yourdan Press/Prentice Hall, New York, 1986
[37] Sommerville, I., Software Engineering, Addison-
Wesley Publishing Company, Wokingham, England, 1989
[38] Kounin, J. S., "Experimental Studies of Rigidity",
Character and Personality, Vol. 9, 1941, pp. 251-282
[39] Banker, R. D., Datar, S., and Kemerer, C., "A Model
to Evaluate Variables Impacting the Productivity of
Software Maintenance Projects", Management Science,
Vol. 37, No. 1, January 1991.
[40] Banker, R. D., Datar, S., and Kemerer, C., "Factors
Affecting Software Maintenance Productivity: An
Exploratory Study", Proceedings of the Eighth ICIS,
Pittsburgh, PA, December 6-9, 1987, pp. 160-175.
[41] Swanson, E. B. and Beath, C. M.,
"Departmentalization in Software Development and
Maintenance", Communications of the ACM, Vol. 33, No.
6, June 1990, pp. 658-667
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE

[42] Wertheimer, M. and Wertheimer, N., "A metabolic
interpretation of Individual Differences in Figural
Aftereffects", Psychological Review, Vol. 61, 1954, pp.
279-280
[43] Coulter, W. A. and Morrow, H. W., Adaptive
Behavior: Concepts and Measurement, Grune and Stratton,
New York, 1978
[44] Georgopoulos, B. S. and Mann, F. C., The
Community General Hospital, Macmillian, New York,
1962
[45] Cooprider, J. G. and Victor, K. M., "The
Contributions of Shared Knowledge to I/S Group
Performance,” Proceedings of the Fourteenth ICIS,
Orlando, FL, December 15-18, 1993
[46] Bagozzi, R. P., Causal Modelling in Marketing, Wiley
& Sons, New York, NY, 1980.
[47] Lucas, H. C. Jr., "Methodological Issues in
Information Systems Survey Research", in The Information
Systems Research Challenge: Survey Research Methods,
K. L. Kraemer, (ed.), Harvard Business School Research
Colloquium, Boston, MA, 1991
[48] Straub, D. W., "Validiating Instruments in MIS
Research," MIS Quarterly, Vol. 13, No. 2, June 1989, pp.
147-165
[49] Newsted, P. R., and Munro, M. C., "Data Acquisition
Instruments in Management Information Systems
Research", in The Information Systems Research
Challenge: Survey Research Methods, K. L. Kraemer,
(ed.), Harvard Business School Research Colloquium,
Boston, MA, 1991
[50] Carmines, E. G. and Zeller, R. A., Reliability and
Validity Assessment, Sage University Press, Beverly Hills,
CA, 1979
[51] Bagozzi, R. P., "The Role of Measurement in Theory
Construction and Hypothesis Testing: Toward a Holistic
Model," in Conceptual and Theoretical Developments in
Marketing, American Marketing Association, Chicago, IL,
1979
[52] Ives, B., and Olson, M., "User Involvement and MIS
Success: A Review of Research," Management Science,
Vol. 30, No. 5, May 1984, pp. 586-603.
[53] Joreskog, K. and Sorblom, D., LISREL VI: Analysis
of Linear Structure Relationships by Method of Maximum
Likelihood, Scientific Software, Mooresville, IN, 1984
[54] Byrne, B. M., A Primer of LISREL: Basic
Applications and Programming for Confirmatory Factor
Models, Springer-Verlag, New York, 1989
[55] Bagozzi, R. P., Yi, Y. and Phillips, L.W., "Assessing
Construct Validity in Organizational Research,"
Administrative Science Quarterly, Vol. 36, 1991, pp. 421-
458.
[56] Cronbach, L. J., "Coefficient Alpha and the Internal
Structure of Tests," Psychometrika, Vol. 16, September
1951, pp. 297-334
[57] Joreskog, K., "Analyzing Psychological Data by
Structural Analysis of Covariance Matrices," in
Contemporary Developments in Mathematical Psychology,
Vol. 11, R.C. Atkinson et al., eds., San Francisco, CA
1974 pp. 1-56
[58] Cook, T. D., and Campbell, D. T., Quasi-
Experimentation: Design & Analysis Issues for Field
Settings, Houghton Mifflin Company , Boston, MA, 1979
[59] Megargee, E. I., The California Psychological
Inventory Handbook, Jossey-Bass Inc., San Francisco,
1972
[60] Blau, P. M., and Schoenherr, R. A., The Structure of
Organizations, New York, Basic Books, 1971
[61] Rossi, P. H., Wright, J. D., and Anderson, A. B.,
Handbook of Research Methods, Academic Press Inc.,
Orlando FL, 1983
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE