Category: Knowledge Management
The Nomological Network and the Research
Continuum
Michael J. Masterson
USAF Air War College, USA
R. Kelly Rainer, Jr.
Auburn University, USA
INTRODUCTION
Social science and management information systems (MIS)
research have been criticized for failure to integrate theory
construction and theory testing (see e.g., Subramanian &
Nilakanta, 1994). In particular, concerns with MIS as a
cohesive research discipline have long included inadequate
construct development and lack of valid, reliable measuring
instruments for those constructs (Keen, 1980). Understanding
the theoretical basis of constructs and how they are developed
and tested across the research continuum are fundamentals of
a cohesive academic discipline. To provide a common research
framework for the growth of MIS as a scientific discipline,
this chapter proposes a framework for an integrated research
continuum across the life cycle of the research process.
BACKGROUND
The growth of any scientific discipline entails the devel-
opment of a system of specialized, abstract concepts that
define the lawful relationships (or hypotheses) that repre-
sent a discipline’s theories. Scientists recognize underlying
concepts in observed phenomena and build those concepts
into the lawful relationships of a scientific theory (Blalock,
1982, Hempel, 1952).
This search for lawful relationships among natural events
is the primary function of science, and focuses on two objec-
tives: (1) describing specific events, objects, or phenomena
in the world of experience and (2) establishing theories, or
general principles, to explain or predict the specified events,
objects, or phenomena (Feigl, 1970 Hempel, 1952).
Theories of a science must support explanation and
prediction. If a scientific discipline lacks explanatory and
predictive theories, no connection can be established between
descriptions of different events, objects, or phenomena. Such
a nontheory based discipline is unable to predict or prepare
for future occurrences. The lack of explanatory principles
permits no use of theory for practical application. Practical
application requires theories or principles that explain what
particular effects occur when specific changes occur in a
Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
2827
N
given system. In addition, comparability tests of the theory
by other researchers would be impossible (Blalock, 1982).
SCIENTIFIC ThEORY AS A SPATIAL
NETWORK AND ThE RESEARCh
CONTINUUM
To develop precise theories of wide scope and high em-
pirical confirmation, scientific disciplines create and evolve
comprehensive systems describing lawful relationships of
theoretical constructs (Hempel, 1952). Some variables of
interest within these theoretical constructs cannot be directly
observed. Thus, constructs contain theorized unobservable,
latent factors measured by empirically observed indicators
(Nunnally, 1978).
A comprehensive scientific theory can be represented as
a spatial network in which hypotheses use correspondence
rules to link theoretical constructs to derived and observed
empirical concepts, which acquire meaning through op-
erational definitions. The unobservable (latent) theoretical
constructs are anchored to the empirical environment by
rules of interpretation (the correspondence rules). By virtue
of these interpretive connections, the network can function
as a scientific theory (Blalock, 1982). The ability to interpret
unobservable, underlying constructs transforms a theoretical,
spatial network into an empirically testable theory.
An important implication of this view of the structure of
theory is that the integration of theory construction and theory
testing across the research continuum is of major importance
(Feigl, 1970; Hempel, 1952; Hughes, Price, & Marrs, 1986;
Trochim, 1996). Figure 1 (below) is a graphic depiction of
this spatial framework, developed from the work of Hempel
(1952), Blalock (1982), and Trochim (1996).
Figure 1 represents a 1:1 correspondence between theo-
retical constructs and measurements, which is not always
the case. Observable measurements, which are qualitatively
different in the empirical world, can overlap or measure
the same thing if their positions in this spatial network link
them to the same theoretical construct. Also, a measurement
may serve as the interpretive link to multiple theoretical
 The Nomological Network and the Research Continuum

constructs, though the researcher would specify the opera-
tive definition in use for that interpretive relationship, for
example, measuring self-esteem requires accounting for
elements of self-image and ego (Cronbach & Meehl, 1955;
Hempel, 1952; Trochim, 1996).
In a spatial network of theory and observation, internal
principles define basic entities (concepts and constructs) of
the theory, and the hypotheses describe interrelationships of
these theoretical constructs, either within the same theory or
with other theories. (see “The CONCEPTUAL” in Figure
1). Bridge principles are the mathematical formulae and
operations, stated as the rules of interpretation, that link the
processes proposed by the theory to empirical phenomena
(measures and observations). Bridge principles enable a
theory to be used for explanations and/or prediction (see
“EMPIRICAL” and “RULES of INTERPRETATION” in
Figure 1). Without these principles, a theory has no ex-
planatory power or practical application, and no empirical
test is possible.
A fundamental requirement of science is that a statement
of fact from an appropriate theory made by one scientist
must be independently verifiable by other scientists. This
principle of scientific generalization gains wide belief and
support for any theory and is central to all scientific work
(Nunnally, 1978).

Figure 1.

SPATIAL NETWORK
OF SCIENTIFIC THEORY

The CONCEPTUAL

Theoretical Theoretical
Construct Construct

THEORY
Theoretical Theoretical
Construct Construct

MATHEMATICS*
-Formulae and Operations

Measurement Measurement

obs obs obs obs

Measurement
EMPIRICAL
Measurement

obs obs
obs obs
MEASUREMENT &
OBSERVATION

*RULES of INTERPRETATION
Bridge from Real World to Theory
Relationships from Indicators to Latent Variables

2828
The Nomological Network and the Research Continuum
Unlike the physical sciences, social and behavioral sci-
ences study living organisms to define and classify behav-
ior, and the hypotheses linking behavior to conditions that
control it. Psychological measurement, or psychometrics,
is inseparable from this process (Nunnally, 1978). Evaluat-
ing the adequacy of a theoretical spatial network requires
judging empirical testability (Popper, 1959), verifiability
(Dodd, 1968), and confirmability (Clark, 1969). Meeting
these standards requires a clear and explicit specification
of theoretical construct definitions and operationaliza-
tions across the research continuum (Hughes et al., 1986).
Statistics is the area of mathematics used in evaluations of
theoretical constructs.
Scientific behavioral studies begin with observations
or questions about some object, event, or phenomenon, in
a search for the functional relationships, the laws, and the
principles of behavior. This behavior must be described in a
manner that fully communicates all observations to others so
that they may verify the observation. (Blalock, 1982).
The Nomological Network. Cronbach and Meehl (1955)
defined a nomological network as the interlocking system of
hypotheses, principles, and laws linking the constructs that
constitute any theory. A nomological network encompasses
the theoretical constructs being measured, how the concept
is going to be measured, and the specification of the inter-
relationships between the theoretical and empirical planes
(see Figure 1). All three aspects of the spatial network are
present in a nomological network: constructs, empirical
observations/measures, and the mathematical rules of inter-
pretation that link theory to observation. Scientifically, the
purpose of the nomological network is to clarify all parts
of a theory, so a theory’s laws of explanation and prediction
can be exploited.
As a standard for tests of social and behavioral constructs,
the nomological network focuses on sampling statistics of
content (test items), not sampling statistics of people. As
empirical measures of a theory are developed, the theory
formulation that guides the development of the mathemati-
cal formulae and operations focuses on sampling content
(Blalock, 1982; Nunnally, 1978). The goal of sampling con-
tent is to generalize findings over populations of test items.
The researcher addresses what, not who, is being measured
(Cronbach, 1975). Thus, the specific focus of the nomological
network is not the subject. Rather, the nomological network
measures observable properties, the relationship of different
theoretical constructs to each other, and the relationship of
the theoretical constructs to the observable properties via
the mathematical rules of interpretation (Cronbach, 1975;
Cronbach & Meehl, 1955). Also, at least some of the laws
in the nomological network must involve observables, oth-
erwise no bridge could exist to the theoretical constructs.
(Cronbach & Meehl, 1955; Trochim, 1996).
Cronbach and Meehl (1955) defined construct validity
as the process of following all these principles of the nomo- N
logical network to ensure generalizability of scientific prin-
ciples. Construct validity specifically addresses the question,
“Does the test measure the attribute it is said to measure?”
Construct validation identifies theoretical constructs defined
by a network of relations, all of which are anchored to ob-
servables, making the constructs testable (Cronbach, 1975).
This process is a direct identification of the spatial network
(see Figure 1) and the nomological network, including the
constructs (concepts) of interest in a theory, their observ-
able manifestations, and the interrelationships among the
constructs, measures, and observations.
The social and behavioral sciences, including MIS, face
difficulties that vary considerably from those encountered
in the physical sciences. These difficulties stem from two
fundamental issues, which are:
1. the relationship between theory and research, and
2. the development of theoretically defined constructs
and measurement procedures.
When the phenomena of interest cannot be directly
observed, the development of latent constructs and instru-
ments to operationalize them provide the theoretical basis
for research in a discipline (Venkatraman & Grant, 1986).
To meet this basic criterion of science, the focus must be on
generalizability and the comparability of one’s measurements
across diverse settings (Hughes et al, 1986; Nunnally, 1978).
Blalock (1982) identifies this focus as one of conceptualiza-
tion and measurement, where
1. Conceptualization refers to the theoretical process by
which researchers move from ideas or constructs to
suggesting appropriate research operations.
2. Measurement refers to the linkage process between the
physical operation, on the one hand, and a mathemati-
cal language on the other.
Systems theory is commonly used as a generic concept
to represent the elements represented in this triple linkage
(Blalock, 1982; Trimmer, 1950). The essential point in
connecting theoretical constructs, physical measurement
operations, and mathematical formulae and operations, is
that the basic concepts or variables be specified with suf-
ficient clarity so that
1. the “systems” or unit to which they are intended to
refer is known;
2. it is clear how to distinguish them from other stimuli,
properties, or responses with which they are likely to
be confounded;
2829

3. it has been specified how concepts that refer to one
type of system are related definitionally, or by some
presumed causal law, to at least some concepts that
refer to any other systems that are also expected to
play important roles in substantive theories.
This last specification implies that it is often desirable
to use information provided at one level of analysis (e.g.,
about individual behaviors) to support, refute, or clarify
theories formulated at other, unobservable levels. Psycholo-
gists successfully integrated the methodological tools of
empirically minded investigators presented in the scaling
and measurement literature. Unfortunately, this has resulted
in an imbalance between theory development and methods
of measurement, with development of measures occupying
a much larger place in the research process than conceptual-
ization of theory (Ackoff, 1984). This creates the necessity
for explicit theoretical formulations that clearly state which
theoretical assumptions are being made in each measurement
decision across the research continuum (Blalock, 1982).
The Research Continuum. The life cycle of the research
process functions as an integrated continuum moving from
the most exploratory, descriptive kinds of observation, to
generalization, and finally to experimental applied research
focused on complete, specific answers (Ackoff, 1984; LaMott,
1997; Vandendorpe, 1997). The clear focus on the nonhi-
erarchical developmental aspect of basic research supports
a model of the research continuum in which phases of the
research life cycle are integrated, rather than performed
separately. This holistic approach lets the researcher deter-
mine research strategy, orientation, and criteria (Cronbach &
Meehl, 1955; Yin, 1994). A researcher determines the type of
research strategy by choosing one of the following familiar
series of questions to ask: “who,” “what,” “where,” “how,”
and “why” (Hedrick et al., 1993; Yin, 1994).
“What” questions are exploratory and deal with concep-
tual discussions and initial theory formulation examining
observed phenomena. “Who” and “where” questions favor
survey strategies or analysis of archival records to seek
generalizable patterns. Frameworks (i.e., a taxonomy of
classification based on standard guidelines) are developed to
explain, organize, and manage identified patterns (see e.g.,
Kochen, 1986; Merrill & Tenneyson, 1977). Frameworks in
the field of MIS, for example Nolan and Wetherbe (1980),
are typically validated by mapping MIS research publications
to the framework to determine if the framework properly
classifies the published material. This method organizes the
research and enables recognition of conceptual patterns.
These strategies support research goals that are either de-
scriptive or predictive. “How” and “why” questions are more
explanatory, and lead to the use of single or multiple case
studies and experiments as the preferred research strategy.
The experiments could be either field experiments or labo-
2830
The Nomological Network and the Research Continuum
ratory experiments, again determined by the researcher’s
orientation (Latane & Darley, 1969; Yin, 1994).
Surveys help researchers generalize their findings.
One survey development example in MIS research is the
development of the end-user computing construct (Rainer
& Harrison, 1993), which used Churchill’s (1979) survey
development process.
When researchers focus on specific “how” and “why”
criteria, field experiments are in order. Data are collected
from subjects in the environment where the construct of
interest exists. Finally, with the proper constructs of interest
(e.g., computer skills) identified, researchers set up labora-
tory experiments to obtain observable outcomes (see e.g.,
Miller, 1994).
Theory building takes place in the exploratory and gener-
alization phases of the research process, while theory testing
occurs in the generalization and experimental phases. This
blending of phases, where theory development or testing
are accomplished, demonstrates the integration of research
across the continuum, as depicted in Figure 2. The research
continuum (LaMott, 1997; Yin, 1994) supports theory build-
ing and theory testing in an interactive model.
The process of basic research as a complex, pluralistic
feedback model integrates the research life cycle across a
broad continuum. Research does not follow a rigidly struc-
tured, waterfall life cycle. Instead, a researcher can spiral
across and down the continuum on a specific research track.
In this spiral process, different strategies and phases of the
research process are determined by the orientation and criteria
set by the researcher. This process includes the peculiarity of
an observation being applied, in practice, with no theoretical
foundation (Kochen, 1986).
In 1997, R&D Magazine’s basic research survey of 4,000
researchers, 67.3% stated that the most critical element of
basic research developed an understanding of scientific prin-
ciples or phenomena. They noted that the second most critical
element was creation of a foundation for future development.
These two critical elements point out the importance and
currency of the spatial network of theory development and
the nomological network (Vandendorpe, 1997).
Different strategies can be employed across the research
continuum. The changing orientation and criteria of the
researcher determines strategy. The nature of the follow-
ing components and dimensions of the research continuum
become more specific and increase in importance in moving
from conceptual studies to laboratory experiments (Blalock,
1982; Nunnally, 1978):
1. knowledge of subjects and environment;
2. questions and objectives and data collection proce-
dures;
3. knowledge of and accuracy in operationalizing vari-
ables and relationships;
4. potential internal validity;
The Nomological Network and the Research Continuum

Figure 2.
N
The RESEARCH CONTINUUM

PHASES of the RESEARCH PROCESS

OUTPUT
INPUT THEORY FORMULATION /
CRITERIA EXPLORATORY DEVELOPMENT

GENERALIZATION IDENTIFY UNDERLYING

WHAT? PATTERNS / CONSTRUCTS
WHO?
WHERE? EXPERIMENTAL EXPLAIN / CAUSEEFFECT
HOW? / PREDICT
WHY? Conceptual Single Case Study Multiple Case Studies Surveys Frameworks Field Experiments Laboratory Experiments

RESEARCH STRATEGIES

IDENTIFICATION VALIDATION REPLICATION

THEORY TESTING / FORMULATION

Figure 3.

STATISTICAL METHDOLOGY and

the RESEARCH CONTINUUM

PHASES of the RESEARCH PROCESS

EXPLORATORY

GENERALIZATION

EXPERIMENTAL

Conceptual Single Case Study Multiple Case Studies Frameworks Surveys Field Experiments Laboratory Experiments

RESEARCH STRATEGIES

IDENTIFICATION VALIDATION REPLICATION

Descriptive Statistics; Inferential Statistics; Replication of Inferential

Non-Parametrics, Small Sample Reliability Determinations, Statistic Studies;
Inferential Statistics - ANOVA Exploratory Factor Analysis Confirmatory Factor Analysis

2831

5. potential external validity;
6. potential reliability;
7. possible sources of bias.
8. ability to predict;
9. ability to uncover causal models.
Also, the statistical methodologies employed by a
researcher change across the research continuum from
conceptual studies to laboratory experiments (see Figure 3).
The change is from descriptive to inferential statistics, from
initial descriptive identification of observations to indepen-
dent confirmation and replication critical to the process of
scientific generalization.
FUTURE TRENDS
Although the spatial, nomological network and the research
continuum have high face validity, more work needs to be
done in these areas. It is interesting to note that the references
here are relatively old. The fact is that researchers have not
done recent work in these areas.
One direction for future research would be to validate
the nomological network and research continuum proposed
here. Researchers could take a subject area in a discipline
(e.g., decision support systems in MIS), map its constructs
to the nomological, spatial network and map its published
articles to the research continuum.
CONCLUSION
Social and behavioral sciences face two very different
issues from those encountered in the physical sciences:
(1) the relationship between theory and research, and (2)
the development of theoretically defined constructs and
measurement procedures. By understanding the integrated
and interactive nature of the research continuum across the
life cycle of research, and the fundamental requirement for
development of theory out of that research, investigators
can advance the rigor of their chosen academic disciplines,
and build a testable body of work to support and grow their
chosen scientific discipline.
This research continuum, grounded in the triple link-
age depicted in the spatial network, is robust and flexible.
However, the challenge in not having readily observable and
measurable data outcomes across the research continuum
makes the operationalization and measurement of constructs
for a researcher difficult, challenging, and tedious (Blalock,
1982). Understanding the theoretical basis of a construct, how
it is developed and tested within a nomological network, and
the interplay across the research continuum, are fundamental
in contributing to the establishment and future development of
MIS and social sciences as cohesive academic disciplines.
2832
The Nomological Network and the Research Continuum
REFERENCES
Ackoff, R. L. (1984). Scientific method: Optimizing applied
research decisions. Malabar, FL: Robert E. Krieger Publish-
ing Company, Inc.,
Blalock, H. M. (1982). Conceptualization and measurement
in the social sciences. Beverly Hills, Ca: Sage Publica-
tions.
Churchill, G. A. Jr. (1979). A paradigm for developing bet-
ter measures of marketing constructs. Journal of Marketing
Research, 16, 64-73.
Clark, J. T. (1969). The philosophy of science and the his-
tory of science. In M. Clayett (Ed.), Critical problems in the
history of science, (pp. 103-140). Madison, WI: University
of Wisconsin Press.
Cronbach, L. J. (1975). The validation of educational mea-
sures. In D. A. Payne & R. F. McMorris, (Eds.), Educational
and psychological measurement: Contributions to theory
and practice, (pp. 75-88). General Learning Corporation,
General Learning Press.
Cronbach, L. J. & Meehl, P. E. (1955). Construct validity in
psychological tests. Psychological Bulletin, 52, 281-303.
Dodd, S. C. (1968). Systemmetrics for evaluating symbolic
systems. Systemmatics, 6, 27-49.
Feigl, H. (1970). The “orthodox” view of theories: Remarks
in defense as well as critique. In M. Radnor, & S. Winokur
(Eds.), Minnesota Studies in the Philosophy of Science, (pp.
4). Minneapolis, MN: University of Minnesota Press.
Hedrick, T., Bickman, L., & Rog, D. J. (1993). Applied
research design. Newbury Park, CA: Sage Publications.
Hempel, C. G. (1952). Fundamentals of concept formation
in empirical science. Foundations of the Unity of Science,
vol. II, #7. Chicago: University of Chicago Press.
Hughes, M. A., Price, R. L., & Marrs, D. W. (1986). Linking
theory construction and theory testing: Models with multiple
indicators of latent variables. Academy of Management
Review, 11, 128-144.
Keen, P. G. W. (1980). MIS research: Reference disciplines
and a cumulative tradition. In E. R. McLean (Ed.), Proceed-
ings of the First International Conference on Information
Systems (pp. 9-18).
Kochen, M. (1986). Are MIS frameworks premature? Journal
of Management Information Systems, 2, 92-100.
LaMott, E. E. (1997). The research continuum. Retrieved
from http://www.csp.edu/scp/people/lamott/sac581/Re-
search1/index htm
The Nomological Network and the Research Continuum
Latane, B., & Darley, J. M. (1969). Bystander apathy. Ameri-
can Behavioral Scientist, 57, 244-268.
Merrill, M. D., & Tennyson, R. D. (1977). Teaching con-
cepts: An instructional design guide. Englewood Cliffs, NJ:
Educational Technology Publications.
Miller, M. D. (1994). The extended technology acceptance
model: Theory and empirical test. Unpublished Dissertation
Auburn, Al: Auburn University.
Nolan, R. L., & Wetherbe, J. C. (1980). Toward a comprehen-
sive framework for MIS research. MIS Quarterly, 4, 1-19.
Nunnally, J. C. (1978). Psychometric theory, [2nd ed.]. New
York: McGraw-Hill Book Company.
Popper, K. R. (1959). The logic of scientific discovery. New
York: Harper & Row.
Rainer, R. K., Jr., & Harrison, A. W. (1993). Toward devel-
opment of the end user computing construct in a university
setting. Decision Sciences, 24, 1187-1202.
Studt, T. (Ed.). (1997). Basic research white paper. R&D
Magazine.Highlands Ranch, Co: Reed Elsevier, Inc.
Subramanian, A., & Nilakanta, S. (1994). Research measure-
ment: A blueprint for theory-building in MIS. Information
& Management, 26, 13-20.
Trimmer, J. D. (1950). Response of physical systems. New
York: Wiley.
Trochim, W. M. K. (1996). The multitrait-multimethod
matrix.Retrieved from http://trochim.human.cornell.edu/
kb/mtmmmat htm
Vandendorpe, L. (1997). Basic research white paper. R&D
Magazine. Highlands Ranch, Co: Reed Elsevier.
Venkatraman, N., & Grant, J. H. (1986). Construct measure-
ment in organizational strategy research: A critique and
proposal. Academy of Management Review, 11, 71-87.
Yin, R. K. (1994). Case study research: Design and methods,
[2nd ed.]. Thousand Oaks, CA: Sage Publications.
KEY TERMS
N
Basic Research: Research performed without thought
of practical ends, producing general knowledge and an
understanding of nature and its laws.
Bridge Principles: Mathematical formulae and op-
erations, stated as the rules of interpretation, that link the
processes proposed by the theory to empirical phenomena
(measures and observations).
Conceptualization: Refers to the theoretical process by
which researchers move from ideas or constructs to suggest-
ing appropriate research operations.
Construct Validity: Identifies theoretical constructs
defined by a network of relations, all of which are anchored
to observables, making the constructs testable.
Internal Principles: Define basic entities (concepts
and constructs) of the theory and the lawful relationships,
hypothesized as part of theory, describing interrelationships
of these theoretical constructs, either within the same theory
or with other theories.
Measurement: Linkage process between the physical
operation, on the one hand, and a mathematical language
on the other.
Measurement by Proclamation: Theoretical claims of
lawful relationships with no link to empirical tests (Blalock,
1982).
Nomological Network: Interlocking system of hy-
potheses, principles, and laws linking the constructs that
constitute any theory.
Theories: General principles to which individual events,
objects, or phenomena conform, and by which the occurrence
of these events, objects, or phenomena is systematically
anticipated.
2833