Intelligent Information Management, 2013, 5, 150-161
http://dx.doi.org/10.4236/iim.2013.55016 Published Online September 2013 (http://www.scirp.org/journal/iim)
Sensing Attributes of an Agile Information System
Pankaj*, Micki Hyde, Arkalgud Ramaprasad
Management Information Systems and Decision Sciences, Eberly College of Business & IT,
Indiana University of Pennsylvania, Indiana, USA
Received June 26, 2013; revised July 28, 2013; accepted August 9, 2013
Copyright © 2013 Pankaj et al. This is an open access article distributed under the Creative Commons Attribution License, which
permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ABSTRACT
Information Systems (IS) agility is a current topic of interest in the IS industry. The study follows up on the work on the
definition of the construct of IS agility wh ich is defined as the ability of IS to sense a ch ange in real time; diagn ose it in
real time; and select and execute an action in real time. It explores the attributes for sensing in an Agile Information
System. A set of attributes was initially derived using the practitioner literature and then refined using interviews with
practitioners. Their importance and validity were estab lished using a survey of the indu stry. All attributes derived from
this study were deemed pertinent for sensing in an agile information system. Dimensions underlying these attributes
were identified using Exp loratory Factor Analysis. This list of attributes can form the basis for assessing and establish-
ing sensing mechanisms to increase IS agility.
Keywords: Information Systems Agility; Agile Information Systems; Sensing; Sensing Attributes
1. Introduction
Change is the rule of the game in the current business
environment. The rate of change has been continuously
increasing due to factors like globalization and opportu-
nities presented by the development and evolution of
technologies. Not only are the changes occurring at an
increasing rate, but they are becoming increasingly un-
predictable. This unpredictability can involve when a
known change will occur, what an unknown change will
look like, or the combination of these. The rapid rate of
change implies that an organization needs to become an
expert at changing and morphing itself rapidly in re-
sponse to a change. Retention of leadership positions re-
quires that an organization should be able to change at
will in any direction, without significan t cost and time, to
counter a threat or capitalize on an opportunity. Such an
organization may be characterized as an agile organiza-
tion. For most organizations the survival and/or retention
of market share demands that the organization should be
able to change faster than, or as fast as, new entrants and
rivals.
Information Systems (IS) pervade all aspects of mod-
ern organizational functioning and play an integral role
in information processing activities of an organization.
Information Systems are needed for organizational agil-
ity on account of their ability to provide shared, distrib-
uted and integ rated, current, and fast flowing info rmation
[1-7].
Modern business processes in organizations use IS as a
core resource or component. In many cases (like ecom-
merce), IS may completely embed a business process
(e.g., Internet banking). The pivotal role of IS in modern
organizational business processes means that an agile
organization cannot change its business processes unless
the IS changes as well. Thus an agile organization would
need an agile IS. What Brandt and Boynton [8] indicated
in 1993 still holds true—that current IS are not easy to
change. Markets change but IS do not. Though many
more organizations are experimenting with Agile Devel-
opment methods, most IS have been developed and are
still being made to cover a closed set of requirements
using the waterfall development methodology. The per-
formance of an IS is also optimized for these require-
ments. The result of this optimization is that IS changes
are often arduous and complicated, especially in cases
where the requirements were not explicitly foreseen by
the designers. But such requirements are frequent in to-
day’s environment. Problems in changing an IS are fur-
ther aggravated by other factors such as outsourcing,
where the knowledge of the IS primarily resides outside
the organization. The inability of IS to change fast im-
pedes organizationa l agility. Th e challenge for an organi-
zation is to structure its IS to meet a variety of changing
*Corresponding author.
C
opyright © 2013 SciRes. IIM

PANKAJ ET AL. 151
requirements, many of which are not even known when
the information systems are built.
2. The Construct of Information Systems
Agility
Agility in general is defined [9,10] as a formative con-
struct comprised of the ability of an information to sense
a change, diagnose a change, select a response, and exe-
cute the response in real-time:
1) Sense: Ability to sense the stimuli for change (as
they occur) in real-time;
2) Diagnose: Ability to interpret or analyze stimuli in
real-time to determine the nature, cause, and impact of
change;
3) Respond: Ability to respond to a change in real-
time, further disaggregated into select and execute:
a) Select: Ability to select a response in real-time
(very short planning time) needed to capitalize on the
opportunity or coun ter the threat.
b) Execute: Ability to Execute the response in real-
time.
Real-time is defined as the span of time in which the
correctness of the task performed not only depends upon
the logical correctness of the task performed but also
upon the time at which the result is produ ced. If the tim-
ing constraints of the system are not met, system failure
is said to have occurred [11]. Thus an Agile IS may be
defined as one that has the ability to sense a change in
real-time, diagnose the change in real-time, select a re-
sp on se in r e al -ti me , and execute the response in real-time.
Due to the formative nature of the construct several, or
some, of these abilities might exist.
3. Sources of Change in Information Systems
IS needs to change continuously. There are several mo-
tivators for an IS change. IS may change on account of
internal changes, organizational changes, and/or environ-
mental changes. IS forms the core of information proc-
essing in modern organizations. IS has to continuously
improve its performance to improve organizational per-
formance, thus requiring continuous changes. Such chan-
ges to the IS are termed as internal changes in this study.
In modern organizations, IS and business processes are
tightly coupled [12]. Business processes have to change
in response to changes in the environment or change in
organizational strategy. A change in the business-process
will require a change in the IS. Conversely, an IS’s abil-
ity to change may permit or prohibit changes in busi-
nessprocesses. An IS has to change to enable changes in
business processes. Such changes are termed as organ-
izational changes. Environmental changes may impact
an IS directly and warrant change. Issues such as techni-
cal compatibility with suppliers, customers, and other
partners; termination of support for obsolete technolo-
gies; upgrade of the hardware and software products;
increase in viruses and cyber-attacks; changes in licens-
ing terms; etc. pose a requirement for changes in IS.
Such changes are termed as e nvironmental changes.
4. Information Systems
IS in this study are restricted to computer-based informa-
tion systems and we categorize IS components into hu-
man and IT components [13] operating within an organ-
izational context (refer to Figure 1). This is a simplistic
definition of an IS and is used in this research to put a
boundary around the domain of interest, since IS in their
holistic view can be all encompassing. Several alterna-
tive ways of classifying components like having a data-
base component and/or including storage in hardware is
possible. Each has its own merits and advantages.
The organizational context for an IS may be described
in different ways depending upon the purpose of the re-
search question. This research borrows from the work of
Ein-Dor and Segev [14] to further elaborate on the issue
of organizational context. Organizational context may be
viewed as a combination of uncontrollable, partially con-
trollable, and controllable factors. Controllable factors
are those whose status is given with respect to IS and
include organizational size, organizational structure, or-
ganizational time-frames, and extra-organizational situa-
tions. The partially controllable factors are those whose
change can be affected by the IS in many situations like
organizational resources, organizational maturity, and the
psychological climate in the organization. Controllable
factors are those for which changes can be affected by
the IS like rank and location of the top IS execu tives, an d
existence of a steering committee. Organizational poli-
cies and rules also play an important part as they control
and guide all organizatio nal actions and imperatives.
The human component of IS consists of the IS staff
responsible for planning, development, operation, and
maintenance of the IS. The human IS component should
have four types of knowledge and skills. These are tech-
nology management knowledge and skills; business and
functional knowledge and skills; interpersonal and man-
agement skills; and technical knowledge and skills [15].
The IT component consists of the computing, storage,
Information Technology
Component
Hardware
Networking
Storage
Software
Human Component
IS Personnel
Figure 1. A model of information systems.

PANKAJ ET AL. 152
and networking equipment, and software that runs on this
equipment. The absence of data as an explicit component
may be seen as an issue. The rationale for not providing
data as an explicit component is that data is a contextual
component rather than an independent component and
can be subsumed in software and storage. The logic of
the business processes is embedded into the configura-
tion of IT components. The human component of the IS
is responsible for embedding or programming this logic
and associated data into the IT components.
5. Research Approach and Objectives
Though there is a lot of discussion of agility in the cur-
rent practitioner literature, theories in the area o f IS agil-
ity are in early formative stages (see [16] for theory
building). IS agility is also an area where practitioners
have taken the lead. In the practitioner literatu re, IS agil-
ity is equated to a set of technologies that enable seam-
less interconnection and collaboration between the IT
components to achieve rapid configuration changes. The
conceptualization of IS agility as proposed in this study
is much broader and more comprehensive in scope. This
study therefore aims to fulfill the following objectives:
1) Arrival at a conceptualized set of attributes for
sensing in an agile IS. (Similar work for diagnosis and
execution will be presented in subsequent manuscripts
due to considerations of length).
2) Refine the conceptualized set of attributes for sens-
ing in an agile IS based on the feedback from practitio-
ners.
3) Verify if the attributes for sensing in an agile IS as
conceptualized in this study provide a comprehensive set
of attributes for sensing for IS agility through a survey.
6. Sensing in an Agile Information System
The human and IT components of an agile IS have sev-
eral attributes which allows the IS to sense, diagnose, and
respond to internal, organizational, and environmental
changes in real time. These changes may be in either of
two components of an IS. A comprehensive list of attrib-
utes of an agile IS will include attributes relevant to
sensing, diagnosing, and implementing a change in the
human and IT components of IS due to organizational,
external, and internal changes.
We use sensing in the same meaning as understood in
control theory. Sensing means that relevant signals are
received and the information on the level/measure of
parameter(s) with which the signal is concerned, is re-
corded. Signals come from everywhere [17]. The ability
to receive signals is contingent upon several factors. Phy-
sical limitations of the receiver may limit what signals
are received [18]. Also, the perceptual limitations of the
living and social entities [17,19] may pose various limi-
tations. As such, an IS should employ a wide variety of
sensors that include machines, living entities, and social
entities to sense the stimuli. Sensing a change is an im-
portant ability. Ability to correctly sense the relevant
parameters is a prerequisite to recognizing a shift in them.
Accurate and timely sensing is especially important in
cases where the state/position of the IS has to shift con-
stantly to meet the stated perfor mance goals.
Sensing requires a plan for measurement/monitoring
of relevant parameters to avoid information overload and
a waste of resources. This set of relevant parameters will
be different for different organizations. Objective meas-
urement of the parameters is highly desirable, e.g., the
percentage of fragmentation for a disk, etc. However, in
many instances, subjective sensing by humans is helpful
and should be used. Though sensing by humans and oth-
er social entities may be subject to various biases [20], it
has its benefits. The sensing in social entities can be ex-
panded (hypothetically in an infinite fashion) through
addition of members with different and additional capa-
bilities [20]. Stimuli in subjective sensing may be user-
reported problems, observations by experts in the various
organizational functions, customers, suppliers, and/or
other external parties. Some signals may be sensed both
objectively and subjectively. Sensing may happen in a
proactive mode or in a reactive mode. In the proactive
mode stimuli that precede the change are sought so that a
change can be anticipated, while in the reactive mode,
stimuli that result from the change are sensed. An agile
IS may have both objective and subjective mechanisms
for proactive and reactive sensing.
7. Attributes for Sensing in an Agile IS
Due to paucity of the peer reviewed academic literature
in the area of sensing for IS agility, most attributes have
been derived from a survey of practitioner literature
(about 250 in number) and the consulting experiences of
the authors. Some academic articles in the area of IS
flexibility (e.g. [21,22 ]) were also referenced.
Since stimuli may come from various sources, the IS
staff needs to be cognizant of the possible sources of
stimuli to set up sensing mechanisms to sense change s in
IT and human (themselves) components. Since the IS
function has limited resources and cannot possibly moni-
tor all sources of stimuli, only some sources may be
monitored continuously while others may be monitored
on an as-needed basis and/or based on the judgment of
the IS staff, and rules, and policies of the organization.
To sense a change in the business processes, good in-
terfaces between IS and the different business units are
needed. These interfaces allow the IS to sense the stimuli
from anticipated and currently occurring changes in
business processes, and rules and policies. Select IS staff
should have an in-depth knowledge of the business and

PANKAJ ET AL. 153
its operations. All IS staff should have a basic level of
kno wl ed ge of the business and its operations. Thi s k no wl -
edge can enable the IS staff to sense a change in the
business as a matter of routine observation. While some
changes may be sensed through observation, other chan-
ges may need more detailed information of the business
operations and may be sensed by the IS staff only when
they are up to date on the current state of affairs in the
various business units. This may be achieved through
open and regular communication between the IS function
and the business units. Open communication and busi-
ness knowledge enables the IS staff to sense change on a
continuing basis. Communication and work with other
business units requires an adequate lev el of interpersonal
skills in the IS staff and their ability to work in cross-
functional teams. All business process changes occur in
the context of the overall business plan and strategic di-
rection of business units and the organization. Knowl-
edge of this context and how it is evolving is an impor-
tant input to sensing. The IS staff should therefore be
involved in the business planning at some level, at least
as observers.
The stimuli for changes are also provided through di-
rect input from end users, buyers, suppliers, and other
partners that may use the IS. All these sources are outside
the IS function and hence can provide signals on aspects
of IS functioning that may not be discernible from within
the IS function. There are three sources of these inputs.
They are change-requests for applications, queries to the
helpdesk for support, and requests for ad-hoc reports.
The business-process logic is implemented through IT
components like software programs and applications.
Requests to alter application functionality, add function-
ality, and fix bugs in the application software are made
on a regular basis. A formal change-management proce-
dure logs these requ ests. This log of the requests captur es
the stimuli that may be employed for further analysis.
Users of an IS need support o n the use of IS. These su p-
port queries are answered by a helpdesk. These helpdesk
queries may indicate a need for change in the IS. A for-
mal log of these he lpdesk queries captures the stimuli for
further analyses. The final direct input from users that
may serve as stimuli is requests for ad-hoc reports. Ad-
hoc reports provide users with information that is not
produced by the IS on a regular basis. The number and
pattern of such reports may indicate a need for change in
the IS. A formal log of these requests may serve as a sti-
mulus for change. Data warehouses and data marts allow
users to pose ad-hoc queries to retrieve information
which may not be obtain ed from the regular transaction al
reports. These queries akin to the ad-hoc report requests
may contain signals of possible changes that are needed.
A formal log of these queries may be maintained.
Various IS, like the online transaction processing sys-
tems (OLTP), enterprise resource planning systems
(ERP), and business intelligence systems (BI), produce
reports on the operations conducted by the organization.
These reports provide a detailed picture of the operations
of the organization, using moving averages, past summa-
ries, thresholds, variances and other figures that may in-
dicate changes in the business operations. The applica-
tion portfolio in an IS should provide such management
reports that serve as stimuli for changes in business that
would trickle down to the IS.
The IT components in an IS should be continuously
monitored and appropriate logs generated. These logs
may point to possible changes occurring in the IT com-
ponents. Most IT components have a predefined set of
parameters that are recommended to be monitored and
logged. At a minimum, an IS should log these parame-
ters. Most IT components in an organization are sourced
from vendors. A good liaison and open communication
with the vendors ensures that detailed information on
possible changes to the IT components and associated
services is passed on to the IS staff. In complex modern
IS this information may constitute the primary stimuli
indicating major and complex changes in the IT compo-
nents not attributable to internal changes.
For the human component of IS, a periodic as well as
ad-hoc assessment of technology management knowl-
edge and skills; business and functional knowledge and
skills; interpersonal and management skills; and techni-
cal knowledge and skills, may be undertaken either in-
formally or formally to sense a change. Such an assess-
ment may be undertaken internally or externally. Peri-
odic internal assessment may be done as part of the for-
mal annual IS strategic planning process.
Environmental developments may directly impact the
IS. IS staff should actively scan the environment for new
technical developments in IS products and processes; the
use of existing and new technologies to solve business
problems; and ways in which IS are being used in peer
organizations. These environmental developments pro-
vide signals for changes in business processes and IT,
leading to changes in IS. The attributes relevant to sens-
ing changes in IS are described in Table 1.
Refining Attributes for Sensing in an Agile IS
Attribute refinement was done through interviews with
ten IS executives. Interviews were conducted to obtain
their opinion on, and validation of, various aspects of IS
agility including sensing. The IS executives targeted
were involved with both business and technological as-
pects of IS and could give a balanced perspective on dif-
ferent aspects of an agile IS ranging from organizational
rules and policies to IT components. As per Bonoma’s
verification guidelines [23], multiple interviews were
proposed fo r purposes of literal replication. The selectio n

PANKAJ ET AL. 154
Table 1. Attributes for sensing a change in an agile infor-
mation system.
1. IS staff skills
1.1. All staff with basic k n o w l edge of business and its
operations
1.2. Some staff with in depth knowled ge of b usi nes s a nd its
operations
1.3. Good interpersonal skills in IS staff.
1.4. Ability of the IS staff to work in cross-functional teams.
1.5. Involvement of IS staff in all bu s i n e ss level planning.
1.6. Periodic and/or ad-hoc assessment of the skills of the IS
staff either internally or externally.
2. Direct input from users
2.1. Existence of a formal change management process that
logs and organizes the request for chang es.
2.2. Formal log of the helpdes k’s communication with the
users of IS.
2.3. Formal log of re quests for ad-hoc reports.
3. Input from the IS
3.1. Reports from the I S t hat provide an overview of the
operations together with comparison data.
3.2. Logs of queries on the data warehouse and data marts.
3.3. Monitoring/measureme nt of performance of IT
components through performance logs.
4. IS interface
4.1. Open communication a nd a go od l iai son between the IS
and business functions.
4.2. Open communication a nd g ood liaison between the IS
staff and the vendors of IT components.
4.3. Active scanning of the IS and proce ss d evel opments, and
innovations in the IS are a .
of the interviewees was done from a list of organizations
that were enlisted in a discussion round table of a re-
search center in the business school of a major university.
The only criterion for selection of the organizations was
to select those organizations that were expected to have a
need for IS agility based on an informal assessment of
the antecedents of IS agility for those organizations [9].
The interviewees demonstrated diversity in organiza-
tions in terms of industry and size; diversity in role and
responsibilities; and diversity in organizations in terms of
their approach to developing and managing IS. For ex-
ample, while one interviewee’s company tried to be on
the forefront of new and developing technologies, an-
other interviewee’s company still used a large number of
legacy systems and was cautious in its use of cutting-
edge technologies. The interviewees held roles that span
from strategic, to a mix of strategic and technical, to
more specialized technical roles.
Content analysis was done on the interviews and at-
tributes for sensing were refined as a result of content
analysis [9]. Most organizations do some sensing (some
are big on it) on a formal and informal basis. IT-level,
people-level, and organizational-level sensing was in
place. Twenty-seven attributes related to sensing were
mentioned. Most were cov ered in the or iginal list. One of
the attributes which was not specifically mentioned ori-
ginally was a priority list of things to sense. It is app arent
that not everything can be sensed and so the IS organiza-
tion needs to have a priority list of things to which it will
pay attention. The other significant attribute was having a
chief IS architect in the organization. The IS architect
can sense changes in technology relevant to the business
due to his/her inherent abilities and skills. A refined set
of attributes for sensing is presented in Table 2. New
attributes have been italicized.
8. Survey Development & Administration
Survey development and administration was done using
the well documented steps for survey development, ad-
ministration and analysis [24-27]. The development and
validation of the survey was done as per the guidelines
by Churchill [28] of generate sample items, pilo t test, and
develop final measures.
The items for the questionnaire were generated from
the list of refined attributes post-interviews. There were
23 items related to sensing (the original questionnaire
included items related to sensing, diagnosis, and selec-
tion & execution). Demographic information about the
organization and respondent was also collected. The
Table 2. Refined attributes for sensing a change in an agile
information system.
1. A priority list of things to sense.
2. Rules-based sensin g.
3. KPIs (Key Performance Indicators) for IS that need to be
monitored.
4. Chief IS architect in the organiz ation who can sense changes in
technology relevant to the bu si nes s d ue to his/her inherent
abilities.
5. IS staff skills
5.1. All staff with basic k n o w l edge of business and its
operations.
5.2. Some staff with in-depth knowledg e of business and its
operations.
5.3. Good interpersonal skills in IS staff.
5.4. Ability of the IS staff to work in cross-functional teams.
5.5. Involvement of IS staff in all bu s i n e ss-level planning.
5.6. Periodic and/or ad-hoc assessment of the skills of the IS
staff either internally or externally.
6. Direct inputs from users
6.1. Existence of a formal change management process which
logs and organizes the request for chang es.
6.2. Formal log of the helpdes k’s communication with the
users of IS.
6.3. Formal log of re quests for ad-hoc reports.
7. Input from the IS
7.1. Reports from the I S t hat provide overview of the
operations together with comparison data.
7.2. Logs of queries on the data warehouse and data marts.
7.3. Monitoring/measureme nt of performance of IT
components through performance logs.
7.4. Reports that combine data from various platform sources
and present an integrated picture of the distributed IS
environment.
8. IS interface
8.1. Open communication a nd a go od l iai son between the IS
and business functions.
8.2. Open communication a nd a go od l iai son between the IS
staff and the vendors of IT components.
8.3. Active scanning of the IS and proce ss d evel opments, and
innovations in the inform ation systems area.

PANKAJ ET AL. 155
items for demographic information used were standard
items compiled from analyses of ten surveys designed by
peer researchers. The demographic information collected
included the title and dep artment, industry, th e number of
IS personnel, annual IS budget, and annual revenue (6
items). The draft survey was pilot-tested with practitio-
ners (paper version) and students (web version) for as-
sessing the understandability of the questions, clarity of
the instructions, unambigu ity in the wo rding of th e items,
and the overall format of the questionnaire. The pilot test
used the paper questionnaire as the primary mechanism
to test for quality of content. Participants in the pilot test
of the paper survey included four practitioners and three
MIS researchers. The primary concern was the length of
the survey which was stated to be slightly long but par-
ticipants commented that the questionnaire could be
completed within a reasonable time and without exten-
sive effort. It was suggested that items be grouped to-
gether. For example, items relating to personnel may be
grouped together and items relating to the IT components
may be grouped together. The idea of contacting people
through a paper letter was supported since emails ru n the
danger of b eing dropped by the SPAM filters. Th e chan-
ges suggested by the researchers included the phrasing of
five items and the layout of the balloon graphic providing
the definition of agility on the second page.
The pilot test for the web questionnaire was oriented
towards visual appeal, layout and design, and not content.
These were done with the help of university students.
The first task was testing on different browsers to verify
that all the buttons and scripts worked and there were no
run time errors. The pause and resume functionality;
clarity of the images and fonts; visual appeal of colors;
amount of scrolling at standard resolution of 800 × 600;
and download times were tested. Some changes in fonts
and layout were done as a result of the tests.
Survey Administration
The survey was mailed to the potential respondents for
self-administration. A list of IS executives was purchased
from a market-research company. The purchased list
contained 5000 names and addr esses. Of these, 2718 exe-
cutives were from companies with between 100 - 1500
employees and 2282 were from companies with 1 - 100
employees. Due to budgetary constraints, it was decided
to mail the survey to 2718 executives (IS staff stren gth of
between 100 - 1500). The su rvey included a cover letter,
a paper copy of the questionnaire and a prepaid return
envelope. The only incentive for responding to the sur-
vey was to share the results of the survey. The cover let-
ter specified that the survey could also be completed
online if desired by the respondent. The reminder for the
survey was mailed approximately three weeks after the
original mailing. After mailing the questionnaires, emails
were sent to about 30 potential respondents using refer-
rals from IS executives known to the researchers. The
emails contained an executive summary of the research
concepts, a copy of the questionnaire, and a link to the
web-based questionnaire. A second mailing of the survey
was done. The survey was mailed to 2448 addresses and
a reminder postcard was mailed approximately three
weeks after the mailing.
9. Survey Analysis and Results
A total of 154 responses were received (it is estimated
that 11 responses were from referrals). Of these, 105
were paper responses and 49 were web responses. There
were a total of 2539 solicitations (accounting for incur-
rect addresses). This gives a 5.7% response rate. This
response rate was considered acceptable considering the
questionnaire had 112 items (not just pertaining to sens-
ing), budgetary constraints on the survey, and the fact
that executives in organizations receive many surveys to
complete that contain some financial incentive (gift
card).
All the data was entered into an SPSS spreadsheet for
analysis. The demograph ic data items (except for the title
and department) representing various demographic cate-
gories were assigned numerical codes. The rating for all
items for benefits and attributes were entered as marked
on the questionnaire except for “Not Applicable” which
was assigned a code of 0 (zero). To verify the correctness
of data entry from the paper survey, a random sample of
20 questionnaires was selected and checked by a fellow
researcher. No data entry errors were found. There were
few missing values in the entire data set. Missing values
were treated as pair-wise exclusive for correlational ana-
lyses like factor analysis. They were treated as list-wise
exclusive otherwise, e.g. for purposes like calculating
descriptive statistics.
9.1. Respondent Demographics
The demographics of the respondents are shown in Ta-
bles 3-6. Education, government, finance, healthcare, and
information technology companies are prominently rep-
resented in the survey. For IS personnel, 62% of the re-
spondents had fewer than 250 IS personnel. 61% of the
organizations had an annual IS budget between $1 mil-
lion and $50 million. 40% of the respondents had reve-
nues between $1 billion an d $25 billio n. In summary, the
survey had respondents from various segments of the
population and it captured opinions from a variety of
organizations in different industries both small and large
in terms of annual revenue, IS budget, and number of IS
personnel.

PANKAJ ET AL. 156
Table 3. Industry distribution of survey respondents.
Industry Frequency Percent
Retail 3 2.0%
Finance & Insurance 23 15.0%
Government 24 15.7%
Information Technology 18 11.8%
Mining & Oil 1 0.7%
Manufacturing 9 5.9%
Education 32 20.9%
Recreation & Leisure 1 0.7%
Utilities 3 2.0%
Trading (Wholesale) 4 2.6%
Media/Publishing/Broadcasting 5 3.3%
Professional Services 3 2.0%
Healthcare 20 13.1%
Other 7 4.6%
Total Valid 153 100.0%
Missing 1
Total 154
Table 4. IS personnel distribution of survey respondents.
IS Personnel Frequency Percent Cumulative Percent
1 - 100 39 25.3% 25.3%
101 - 250 56 36.4% 61.7%
251 - 500 29 18.8% 80.5%
501 - 750 12 7.8% 88.3%
751 - 1000 3 1.9% 90.3%
1001 - 1500 3 1.9% 92.2%
1501 - 2000 1 0.6% 92.9%
2001 - 5000 3 1.9% 94.8%
>5000 8 5.2% 100.0%
Total 154 100.0
Table 5. Distribution of annual is budget of survey respon-
dents.
Annual IS Budget
($millions) Frequency Percent Cumulative Percent
<0.5 11 7.2% 7.2%
0.5 - 1 10 6.5% 13.7%
1 - 50 93 60.8% 74.5%
50 -100 20 13.1% 87.6%
100 - 500 Million 15 9.8% 97.4%
500 - 1000 3 2.0% 99.3%
>1000 1 0.7% 100.0%
Total Valid 153 100.0%
Missing 1
Total 154
9.2. Respondent Bias
Since the survey was anonymous, it prevented testing for
demographical differences between respondents and non-
respondents. The respondents from the two mailings
were, however, compared for differences. The number of
responses for each medium and for each mailing is de-
tailed in Tab le 7. The number of responses in the second
mailing was about half of tho se in the first mailing.
To ensure that there were no significant unexplainable
factors operating between the first and second mailing,
cross-tabulations were computed to examine the differ-
ences on the demographic variables of industry, IS per-
sonnel, IS budget, and annual revenue. Chi-square tests
were not conducted since many of the cells had an ex-
pected count of less than 5 and combining various cate-
gories led to mitigation of the differences with respect to
the categories for the two mailings of the survey. There
was a noticeable difference in the profile of the respon-
dents between the first and second mailing. In addition,
mean ratings of benefits of an agile IS (not covered in
this manuscript) and the attributes of an agile IS were
compared using independent sample t-tests. The means
of the ratings were not found to be different between the
two mailings at a significance level of 0.01 .
Table 6. Distribution of annual revenue of survey respon-
dents.
Annual Revenue Frequency Percent Cumulative
Percent
<$1 Million 8 5.5% 5.5%
$1 Million - $50 Million 18 12.4% 17.9%
$50 Million - $100 Million9 6.2% 24.1%
$100 Million - $500 Million25 17.2% 41.4%
$500 Million - $1 Billion 21 14.5% 55.9%
$1 Billion - $25 Billion 58 40.0% 95.9%
>$25 Billion 6 4.1% 100.0%
Total Valid 145 100.0%
Missing 9
Total 154
Table 7. Paper and web responses for first and second sur-
vey mailing.
Mailing
First Second Total
Paper 67 38 105
Type of
Medium for
Survey Web 36 13 49
Total 103 51 154

PANKAJ ET AL. 157
9.3. Attribute Importance for Sensing a Change
Respondents were asked to rate the importance of 23
items for sensing in real time. The mean rating for each
of the items is significantly different (greater) from the
mid-point of the scale (rating of 4 indicating neutral) at
0.01 level of significance, on a two-tailed test. Table 8
shows the frequency, mean, and standard deviation of the
rating of the attributes for sensin g a change in real time.
The attributes relating to the skills of the IS personnel
received top ratings. Open communication between the
IS function and other business functions was rated as the
top attribute. Select IS personnel with in-depth business
knowledge was the second highest rated attribute (6.08).
The ability to work in cross-functional teams (6.01) and
good interpersonal skills (5.97) were rated as important
for sensing. Attributes with a similar theme of the pres-
ence of IS personnel in meetings (5.27) and in business
planning (5.56) as suggested by one of the interviewees,
were not rated very highly. This may be due to the pres-
ence of other mechanisms that may achieve the same
effect like departmental-level managers sense that a
change is needed and communicate it to th e IS unit. “Ex-
ternal skill assessment of the IS personnel” (4.51) was
not a priority. A possible reason for this may again be
due to the resource requirement for such a task like
budget. Overall, all attributes were considered important
with an average rating of more than 4.
Table 8. Survey ratings of the attributes for sensing a change in an agile information system.
Frequency of Rating1
Attribute for Sensing a Change in Real Time 1234 5 6 7 N/A N Mean2Std.
Dev.
Open communication between IS function and other business
functions 0 0 1 3 114591 3 154 6.47** 0.94
Select IS personnel with in-depth knowledge of
business and its operations 0 1 4 12204572 0 154 6.08** 1.12
Ability of IS personnel to work in cross-functional teams 0 1 0 8 286550 2 154 6.01** 0.91
IS change managem ent pr ocess to log and examine change requests
from users 0 1 2 12206355 0 154 6.01** 1.02
IS personnel with good inter personal skills (interface with other
parts of the organization) 0 0 1 14236550 0 153 5.97** 0.95
Open communication between the IS personnel and IS/IT ven dors 0 0 3 7 346444 2 154 5.91** 0.94
Existence of Key Performance Indicators (KPI) for IS 0 3 4 16295744 1 154 5.73** 1.18
Monitoring of various IT components through activity and
performance logs 0 0 5 17375936 0 154 5.68** 1.10
A list of IS, organizational, a nd environmental parameters to be
monitored regularly 0 1 3 16456028 1 154 5.59** 1.01
Examination of logs of helpdesk activity 0 2 6 11475829 1 154 5.57** 1.09
Existence of a “Chief IS Architect” 0 3 9 16315936 0 154 5.57** 1.23
Involvement of IS personnel in most business planning 0 6 5 21246038 0 154 5.56** 1.30
Parameter thresholds, and list of ad ditional paramet ers to be
monitored on exceptions 0 1 3 22445528 1 154 5.52** 1.05
All IS personnel with a basic knowledge of b usiness and its
operations 0 2 5 31424628 0 154 5.36** 1.16
Continuous internal sk i ll a ss es sment of IS personnel 0 3 9 1651 5123 1 154 5.35** 1.16
Presence of IS personnel in all business planning meetings 4 6 6 18395426 0 153 5.27** 1.42
Informal scanning o f the env ironment for developments in IS area
(technology, etc.) 2 2 7 25495018 0 154 5.22** 1.20
Examination of reports from transaction processing systems for
business trends 1 2 6 2747 5412 4 153 5.19** 1.11
Integrated reports synthesized from logs of various IT components 3 1 9 3047 3626 2 154 5.16** 1.31
Formal scanning of the environment for developments in IS area
(technology, etc.) 2 5 9 2353 4615 0 153 5.08** 1.26
Examination of logs of queri es to data marts and data warehouse 2 1 1023594111 7 154 5.06** 1.14
Examination of logs of requests for ad-hoc reports 0 3 161769 3214 3 154 5.01** 1.14
Continuous external skill a s s es sment of IS personnel 3 10193553239 1 153 4.51** 1.34
Notes: 1The shaded boxes show the rating with the highest frequencies. 2Items are arranged in ascending order of mean rating. **Mean is significantly greater
than 4 at 0.00 level of significance in a two-tailed test.

PANKAJ ET AL. 158
9.4. Exploratory Factor Analysis for
Identification of Dimensions Underlying the
Sensing Attributes
An exploratory factor analysis (EFA) was conducted to
identify possible dimensions underlying the benefits of
an agile IS and the attributes for sensing. Though broad
categories were defined when arriving at attributes and
so it would be possible to do a confirmatory factor analy-
sis, it was felt that an EFA is appropriate given the ex-
ploratory nature of the study. For sensing, the number of
attributes was 23 for a sample of 154. This satisfied the
rule of thumb of 5:1for the sample size. To test the ap-
propriateness of the correlation matrix for factor an alysis
the Kaiser-Meyer-Olkin’s Measure of Sampling Ade-
quacy (KMO MSA) was examined. The KMO MSA for
sensing was 0.816 or at a meritorious level. An EFA was
conducted using principal axis factoring since the object-
ive was to identify underlying dimensions based on the
common variance shared by the attributes. Varimax rota-
tion was used to arrive at a simple structure for the fac-
tors. Other rotations oblimin, quatrimax, and promax
were also tested. Varimax rotation provided the simplest
and most interpretable factor structure. The number of
factors to be extracted was based on a combination of the
scree test, eigen-value criterion, and simple structure.
Simplicity of structure was given preference over parsi-
mony. Initial factor extraction was based on the criteria
of the eigen-value greater than 1. Then, the scree plot
was examined for a visible elbow such that all factors
with an eigen-value of greater than 1 were included in
the solution. The factor analysis was run again with the
number of factors indicated by the elbow. Both solutions
were examined and the factor structure that was simple
and interpretable was chosen. The cut-off for factor load-
ing was taken as 0.4. Guidelines suggest [29] loadings of
greater than ±0.3 to meet the minimal level while load-
ings of ±0.40 are considered more important.
For the EFA of the attributes for sensing, nine factors
were extracted. Using the criterion of eigen-values of
greater than one, six factors were extracted. The scree
plot had a distinct elbow at factor 9 with a possible elbow
at factor 7. Examining the factor structure with 6, 7, and
9 factors, the solution with nine factors was retained. The
nine factors together explain 74% of the variance. Table
9 provides details of the factors and the items that load
on the factors along with their loadings. The factors have
been provided with some indicative names.
The first factor was labeled as “Institutional Sensing”.
The items under this factor related to organizational level
sensing done both objectively and subjectively. The sec-
ond factor was labeled as “Trend Sensing” as the items
that loaded on this factor were related to looking for
trends that can be inferred by examining various logs
and/or reports. The item “examination of helpdesk activ-
ity” cross-loaded onto this factor. This may be explained
by the fact that helpdesk activity logs may also point to
trends. The third factor was labeled as “External Sens-
ing” and included formal and informal environmental
scanning. The fourth was labeled as “IS Presence” since
it included items relating to sensing internal organiza-
tional stimuli through attendance at meetings and par-
ticipation in business planning. The fifth factor was la-
beled as “Interface” since it included items that related to
the interface of IS personnel with the rest of the organi-
zation. IS personnel may sense various stimuli through
working with personnel from other functions. The sixth
factor was labeled as “Communication” and the items
that loaded onto this factor related to signals or stimuli
commu nicat ed from the organization and external partie s.
The seventh factor was labeled as “Assessment” since
the items that loaded on this factor exclusively related to
sensing change in the IS personnel. The eighth factor was
labeled as “Logs” since the items that loaded on this fac-
tor related to recording signals in various logs. The ninth
factor was labeled as “Business Skills”. Apart from the
item “select IS personnel with in-depth business knowl-
edge”, the item “examination of reports from the Trans-
action Processing Systems (TPS) for business trends”
cross-loaded on this factor. The latter item relates to ex-
amining TPS reports for business trends which would
need significant business knowledge on the part of IS
personnel.
The item “all IS personnel with basic knowledge of
business and its operations” did not load on any factor.
One possible reason for this may be that many other
items (presence at meetings, cross-functional team, etc.)
may subsume this item and its variance may be explained
through other items. Two items cross-loaded and the
cross-loading could be justified to the extent that the
items appeared to be relevant to both the factors on
which the items loaded. These items could be split into
items that are specific and meaningful with respect to
each of the factors. For instance, “examination of help-
desk log” may be rephrased to “examination of helpdesk
logs to identify tren ds” in th e context of the “Trend Sens-
ing Factor”. These issues may be addressed in future
studies in this area.
10. Summary and Comments
The survey data support the importance of attributes per-
taining to sensing in an agile information system. These
can be used by practitioners to improve the ability of
their IS to sense. The EFA identifies dimensions that
make logical sense at this early stage of exploration. The
major finding from the study is that people matter and
that IS agility is people-driven. The attributes related to
personnel in the areas of sensing were amongst the top
ated. In the comments that accompanied the survey r

PANKAJ ET AL. 159
Table 9. Factors for sensing a change in an agile information system.
Factors
Attribute for Sensing a Change
in Real Time 1
(Institutional
Sensing)
2
(Trend
Sensing)
3
(External
Sensing)
4
(IS
Presence)
5
(Interface) 6 (Communication) 7 (Assessment) 8 (Logs)
9
(Business
Skills)
Parameter thresholds, and list of
additional parameters to b e
monitored on exceptions 0.844
A list of IS, organizational, a nd
environmental parameters to be
monitored regularly 0.741
Existence of a “Chief IS Architect” 0.450
Existence of Key Performance
Indicators (KPI) for IS 0.428
Integrated reports synthesized from
logs of various IT com ponents 0.402
Examination of logs of queri es to
data marts and data warehouse 0.809
Examination of logs of requests for
ad-hoc reports 0.779
Examination of reports from
transaction processing syst ems for
business trends 0.578 0.409
Informal scanning of the
environment for developments
in IS area (technology, etc.)
0.817
Formal scanning of the environment
for developments in IS area
(technology, etc.)
0.601
Involvemen t of IS personnel in most
business planning
0.853
Presence of IS personnel in all
business planning meetings
0.636
Ability of IS personnel to work in
cross-functional teams
0.762
IS personnel with good
interpersonal skills (interface with
other parts of the organization)
0.690
All IS personnel with a basi c
knowledge of business and its
operations
Open communication between IS
function and other business
functions 0.913
Open communication between the
IS personnel and IS/IT vendor s 0.533
Continuous external skill
assessment of IS personnel 0.712
Continuous internal skill
assessment of IS personnel 0.682
Monitoring of various IT
components through activity and
performance logs
0.580
Examination of logs of helpdesk
activity 0.462 0.551
IS change management process to
log and examine change requests
from users
0.520
Select IS personnel with in-depth
knowledge of business and its
operations
0.464

PANKAJ ET AL. 160
respondents stressed the importance of people. The stress
on personnel implies that organizations need to focus on
issues like training and retention of good IS personnel.
A valuable area where the research has implications is
in the area of IS personnel. IS personnel were stated to be
the key to achieving IS agility. There is a trend in the
industry to outsource and to reduce personnel-related
costs. This may move a significant amount of knowledge
pertaining to IS to external parties. This knowledge is
seldom shared between partners. Hence the outsourcing
organizations run the risk of losing this knowledge and
thereby their ability to ch ange IS on th eir own accord and
in time. So organizations need to put in some thought on
the retention and building of a team of quality of IS per-
sonnel and to restrict outsourcing to just non-critical ar-
eas if they have resources to follow such a strategy. In
case outsourcing has to be used, the organization should
take steps to establish an internal core of skilled IS per-
sonnel.
10.1. Limitations
There are some limitations in this study. This research
represents the state at a particular point in time. In the IS
area where technologies, practices, and concepts are con-
tinuously evolving, the research may need to be continu-
ally refreshed to be more meaningful with the times.
Items may need to be further refined to avoid cross-load-
ings and non-loading items in EFA. The response rate for
the survey was 5.7% and though this response rate was
considered acceptable for the purpose of this survey there
is still room to improve the response rate. This could be
done, for example, through collaboration with some spe-
cial interest groups and research groups with an interest
in the area of IS agility.
10.2. Future Research
The attributes for sensing in an agile IS identified in this
study are based on a continuous and rigorous monitoring
of IS industry developments. The shifting ideas and
thoughts have been conso lidated into the attribute list for
sensing in an agile IS in this study. This list could serve
as a starting point for further research.
The instrument and the items may be examined and
revised based on the data collection and analyses in this
research, and the study may then be replicated. The rep-
licated results may be used to come up with a scale for
measuring sensing abilities of an agile IS.
REFERENCES
[1] R. Mason-Jones and D. R. Towill, “Total Cycle Time
Compression and the Agile Supply Chain,” International
Journ al of Production Economics, Vol. 62, No. 1-2, 1999,
pp. 61-73. doi:10.1016/S0925-5273(98)00221-7
[2] M. Christopher, “The Agile Supply Chain: Competing in
Volatile Markets,” Industrial Marketing Management,
Vol. 29, No. 1, 2000, pp. 37-44.
doi:10.1016/S0019-8501(99)00110-8
[3] R. I. V. Hoek, “The Thesis of Leagility Revisited,” In-
ternational Journal of Agile Manufacturing Management
Systems, Vol. 2, No. 3, 2000, pp. 196-201.
[4] N. Bajgoric, “Web-Based Information Access for Agile
Management,” International Journal of Agile Manage-
ment Systems, Vol. 2, No. 2, 2000, pp. 121-129.
doi:10.1108/14654650010337131
[5] H. Sharifi and Z. Zhang, “A Methodology for Achieving
Agility in Manufacturing Organizations: An Introduc-
tion,” International Journal of Production Economics,
Vol. 62, No. 1, 1999, pp. 7-22.
doi:10.1016/S0925-5273(98)00217-5
[6] Y. Y. Yusuf, M. Sarahadi and A. Gunasekaran, “Agile
Manufacturing: The Drivers, Concepts, and Attributes”,
International Journal of Production Economics, Vol. 62,
No. 1, 1999, pp. 33-43.
doi:10.1016/S0925-5273(98)00219-9
[7] J. Bal, R. Wilding and J. Gundry, “Virtual Teaming in the
Agile Supply Chain,” The International Journal of Logis-
tics Management, Vol. 10, No. 2, 1999, pp. 71-82.
doi:10.1108/09574099910806003
[8] A. C. Boynton, “Achieving Dynamic Stability through In-
formation Technology,” California Management Review,
Vol. 35, No. 2, 1993, pp. 58-77. doi:10.2307/41166722
[9] Pankaj, “An Analysis and Exploration of Information
Systems Agility,” Ph.D. The sis, Southern Illinois Univer-
sity, Carbondale, 2005, 343 pp.
[10] M. H. Pankaj and A. Ramaprasad, “Revisiting Agility to
Conceptualize Information Systems Agility,” In: M. Ly-
tras and P. O. D. Pablos, Eds., Emerging Topics and Te-
chnologies in Information Systems, IGI-Global, Hershey,
2009, pp. 19-54.
[11] “Comp.realtime: Frequently Asked Questions (FAQs)
(Version 3.5),” 2002.
http://www.faqs.org/faqs/realtime-computing/faq/
[12] J. Edwards, T. Milea, S. Mcleod and I. Coults, “Agile
System Design and Build.” IEEE Colloquium on Manag-
ing Requirements Change: A Business Process Reengi-
neering Perspect i ve, London, 11 June 1998, pp. 1-5.
[13] T. A. Byrd and D.E. Turner, “Measuring the Flexibility of
Information Technology Infrastructure,” Journal of Man-
agement Information Systems, Vol. 17, No. 1, 2000. pp.
167-208.
[14] P. Ein-Dor and E. Segev, “Organizational Context and the
Success of Management Information Systems,” Man-
agement Science, Vol. 24, No. 10, 1978, pp. 1064-1078.
doi:10.1287/mnsc.24.10.1064
[15] D. M. S. Lee, E. Trauth and D. Farwell, “Critical Skills
and Knowledge Requirements of IS Professionals: A
Joint Academic/Industry Investigation,” MIS Quarterly,
Vol. 19, No. 3, 1995, pp. 313-340. doi:10.2307/249598
[16] K. M. Eisenhardt, “Building Theories from Case Study

PANKAJ ET AL. 161
Research,” Ac ade my of M anagement Review, Vol. 14, N o.
4, 1989, pp. 532-550.
[17] J. B. Quinn, “Managing Strategic Change,” Sloan Man-
agement Review, Vol. 21, No. 4, 1980, pp. 3-20.
[18] S. H. Haeckel and R. L. Nolan, “Managing by Wire: Us-
ing I/T to Transform a Business from Make-and-Sell to
Sense-and-Respond,” In: J. N. Luftman, Ed., Competing
in the Information Age: Strategic Alignment in Practice,
Oxford University Press Inc., London, 1996, p. 220.
[19] R. S. Billings, T. W. Milburn and M. L. Schaalman, “A
Model for Crisis Perception: A Theatrical and Empirical
Analysis,” Administrative Science Quarterly, Vol. 25, No.
2, 1980, pp. 300-316. doi:10.2307/2392456
[20] S. Kiesler and L. Sproull, “Managerial Response to
Changing Environments: Perspectives on Problem Sens-
ing from Social Cognition,” Administrative Science Quar-
terly, Vol. 27, No. 2, 1982, pp. 548-570.
doi:10.2307/2392530
[21] S. P. Weil, M. Subramani and M. Broadbent, “Building
IT Infrastructure for Strategic Agility,” MIT Sloan Man-
agement Review, Vol. 44, No. 1, 2002, pp. 57-65.
[22] B. R. Allen and A. C. Boynton, “Information Architec-
ture: In Search of Efficient Flexibility,” MIS Quarterly,
Vol. 15, No. 4, 1991, pp. 435-445. doi:10.2307/249447
[23] T. V. Bonoma, “Case Research in Marketing: Opportuni-
ties, Problems, and a Process,” Journal of Marketing Re-
search, Vol. 22, No. 2, 1985, pp. 199-208.
doi:10.2307/3151365
[24] J. E. Bailey and S. W. Pearson, “Development of a Tool
for Measuring and Analyzing Computer User Satisfac-
tion,” Management Science, Vol. 29, No. 5, 1983, pp.
530-545. doi:10.1287/mnsc.29.5.530
[25] D. L. Goodhue, “Supporting Users of Corporate Data,”
Ph.D. Thesis, Massachusetts Institute of Technology,
Boston, 1988, p. 300.
[26] B. Ives, M. H. Olson and J. J. Baroudi, “The Measure-
ment of User Information Satisfaction,” Communications
of the ACM, Vol. 26, No. 10, 1983, pp. 785-793.
doi:10.1145/358413.358430
[27] J. A. Ricketts and A. M. Jenkins, “The Development of
an MIS Satisfaction Questionnaire: An Instrument for
Evaluating User Satisfaction with Turnkey Decision Sup-
port Systems,” Discussion Paper No. 2961985, Indiana
University School of Business, Bloomington.
[28] G. A. Churchill, “A Paradigm for Developing Better
Measures of Marketing Constructs,” Journal of Market-
ing Research, Vol. 16, No. 1, 1979, pp. 64-73.
doi:10.2307/3150876
[29] J. H. Hair Jr., R. E. Anderson, R. L. Tatham and W. C.
Black, “Multivariate Data Analysis,” Prentice Hall, Upper
Saddle River, 1995.