Quantitative and Qualitative Studies

Advances in Developing Human
Resources
http://adh.sagepub.com/
Quantitative Research and Theory Building: Dubin’s Method
Susan A. Lynham
Advances in Developing Human Resources 2002 4: 242
DOI: 10.1177/15222302004003003
The online version of this article can be found at:
http://adh.sagepub.com/content/4/3/242
Published by:
http://www.sagepublications.com
On behalf of:
Academy of Human Resource Development
Additional services and information for Advances in Developing Human Resources can be found at:
Email Alerts: http://adh.sagepub.com/cgi/alerts
Subscriptions: http://adh.sagepub.com/subscriptions
Reprints: http://www.sagepub.com/journalsReprints.nav
Permissions: http://www.sagepub.com/journalsPermissions.nav
Citations: http://adh.sagepub.com/content/4/3/242.refs.html
>> Version of Record – Aug 1, 2002
What is This?
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
䉱 Chapter 2
Advances in Developing Human Resources
Lynham / DUBIN’S METHOD
August 2002
Quantitative Research and
Theory Building: Dubin’s Method
Susan A. Lynham
The problem and the solution. There are a number of strategies and methods that the theory builder can use to develop
applied theory. Each strategy and method is informed by
assumptions about what makes for knowledge and in turn good
theory—and each is a way of seeing and understanding the phenomenon central to the theory. Dubin, a recognized scholar in
applied theory building, advocates a theory-then-research strategy and quantitative hypothetico-deductive approach to applied
theory building. This chapter focuses on Dubin’s quantitative
method of theory building for applied disciplines, a two-part,
eight-step theory-building method as a specific variation of the
general method of theory-building research in applied disciplines presented in chapter 1.
Robert Dubin (1978), in his landmark book Theory Building, cautioned
against knowing and practicing “the form of theory-research” (p. 267) while
not knowing the substance thereof. It is the purpose of this chapter to present
the reader with both the form, that is, the shape, structure, and outward
appearance, as well as the substance, or what might be considered as the
essence, of a quantitative theory-building methodology. The form of
applied theory building is presented by way of a description of the
hypothetico-deductive method and the continual deduction-driven refrain
that is embedded in this hypothetico-deductive approach to applied theory
building. The theory-research cycle that contains Dubin’s two-part, eightstep theory-building method necessary for the development of valid, trustworthy applied theory is then presented in Figure 1. The substance of this
applied theory-building method is a detailed presentation of this two-part,
eight-step process. The chapter concludes with an illustrative summary of
the relationship between Dubin’s two-part, eight-step theory-building
method and the research questions of the study of application, together with
some implications of this explicit theory-building method for research and
practice in human resource development (HRD) and other applied
disciplines.
Advances in Developing Human Resources Vol. 4, No. 3 August 2002 242-276
Copyright 2002 Sage Publications
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
Lynham / DUBIN’S METHOD
1. Units
2. Laws of interaction
3. Boundaries
4. System States
These four
steps result in
an informed,
conceptual
framework of
the theory
Part One:
The theory
development side
of the
Theory-Research
Cycle
These four
steps result in
an empirically
verified, and
trustworthy
theory
Part Two:
The research
operation side
of the
Theory-Research
Cycle
5. Propositions
6. Empirical indicators of
key terms
7. Hypotheses
8. Testing
FIGURE 1:
Dubin’s Theory-Building Method as an Eight-Step Theory-Research Cycle
The Form of Dubin’s Quantitative
Theory-Building Method
The shape of Dubin’s (1976, 1978) theory-building research method is
based on two distinct and related conceptual and methodological components. The first component is the hypothetico-deductive approach to the
construction of knowledge. This first component in turn informs the second,
namely, a theory-to-research, or “theory-then-research” (Reynolds, 1971,
p. 144), strategy for theory development and verification.
The Hypothetico-Deductive Approach
to Knowledge Construction
The notion of the hypothetico-deductive method implies a predisposition to a
particular perspective of what makes for theory and, therefore, an appropriate
method of theory construction and verification: “A theory in science is a general
statement (or hypotheses) from which particular inferences may be deduced”
and wherein “observations can then be seen as confirming or falsifying hypotheses” (Honderich, 1995, p. 385). Kaplan (1964) described this method as “the
most widely accepted reconstruction of science” (p. 9):
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
243
244
Advances in Developing Human Resources
August 2002
According to this reconstruction, the scientist, by a combination of careful observation, shrewd
guesses, and scientific intuition arrives at a set of postulates governing the phenomena in which
he is interested; from these he deduces observable consequences; he then tests these consequences by experiment, and so confirms or disconfirms the postulates, replacing them, where
necessary, by others, and so continuing. (pp. 9-10)
The hypothetico-deductive method of knowledge construction is contained
within the empirical-analytical approach to science (Hultgren & Coomer, 1989)
and view of theory-building inquiry (see Lynham, 2002, Table 1, chap. 1). This
method therefore directs theory-building researcher-theorists in a number of
ways: that their theory-building endeavors should be motivated by a practical/
technical interest of means and ends; should assume that observational data are
the foundation of knowledge; should pursue the purpose of explanation, prediction, and control; and should result in the desired outcome of developing transferable, generalizable laws and explanations of organizational and human
behavior. This hypothetico-deductive method is therefore central to the two-part
theory-research cycle and eight-step theory-building research method advocated by Dubin (1978) to be discussed in the following sections of this chapter. It
should be noted that the discussions following are not intended to be sufficient to
convey the detailed complexity of Dubin’s theory-building research method but
rather sufficient for the reader to gain a sense of and feel for the basic form and
essence of this theory-building method. Those wishing to employ this method of
applied theory building are urged to consult Dubin’s 278-page book in this regard.
The Two-Part Theory-Research Cycle and Dubin’s
Eight-Step Applied Theory-Building Method
Dubin (1978) contributed a widely recognized research method and
accompanying language for theory building in an applied field. At a conceptual level, his method can be presented as a continuous theory-research
cycle composed of two parts, the first being the theoretical side of the cycle
and the second the research operation side of the cycle (Lynham, 2000a)
(also see Figure 1). Successful completion of the first part, or theoretical
side, of the cycle results in an informed, conceptual framework of the theory,
whereas successful completion of the second part, or research operation
side, results in an empirically verified and trustworthy theory (Lynham,
2000a). The task of the theory-researcher is to consistently and conscientiously move through each of the two parts of the cycle, by following the precise substance of Dubin’s applied theory-building method, which is made
explicit through four theory development and four research operation, or
theory verification, steps, as indicated in Figure 1 (Lynham, 2000a). Following the form and substance of this two-part theory-research cycle and eightstep applied theory-building method is considered necessary and sufficient
to ensure both rigor and relevance in the resulting theory.
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
Lynham / DUBIN’S METHOD
The first task of the theory development side of the cycle, and the first
step in the applied theory-building method, requires identification of the
units of the theory, also known as the concepts of the theory. These units represent the things or variables whose interactions constitute the subject matter, or phenomenon, that is the attention of the theory. After the units of the
theory have been identified, the next step in theory development is to specify
how the units interact and relate to one another, which is accomplished by
stipulating the laws of interaction that pertain to the units of the theory. The
laws of interaction show how changes in one or more of the units of the theory influence the remaining units. It is significant to note that describing the
units of the theory and how these units interact (that is, the laws of interaction) makes for the major contribution to knowledge that is generated by the
theory (Dubin, 1978; Lynham, 2000a; Torraco, 1994, 2000).
Next in the theory development process is the task of determining the
boundaries of the theory. This third step in the applied theory-building
method enables the researcher-theorist to determine the limited domain of
the world in which the theory is expected to hold true, or to hold up. Determining the boundaries of the theory enables the researcher-theorist to set
and clarify the aspects of the real world that the theory is attempting to
model and therefore bounds, or limits, the theoretical domain of the theory
from other aspects of the real world that the theory is not attempting to
model and explain. The fourth step in the applied theory-building method,
and the final step in the theory development side of the cycle, is the specification of the system states of the theory. System states indicate the complexity of the real world that the theory is presumed to represent and the different
conditions under which the theory operates. Theories can have numerous or
few system states, but all units of the specified system states are determinant
and measurable and are distinctive for each state of the theoretical system
(Dubin, 1978).
Attention to Steps 1 through 4 completes the theory development part of
the theory-research cycle of the theory-building method as specified by Dubin
(1978). Successful completion of this part of the applied theory-building
process constitutes the conceptual development of the theory and results in
the explicit output of a conceptual or theoretical framework of the theory. To
ensure the informed nature of this theoretical framework, the researchertheorist is compelled to rely on thorough and scholarly exploration and
study of current and related literature and research, as well as on her or his
observed experience of the phenomenon in the real world. Once the theoretical framework of the theory has been developed, the researcher-theorist is
ready to move on to the second part of the theory-research cycle, which also
makes up the research operation side of Dubin’s theory-building method.
The research operation side of the theory-research cycle of theory building in applied disciplines consists of four steps, which coincide with Steps 5
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
245
246
Advances in Developing Human Resources
August 2002
through 8 of Dubin’s theory-building method. These four steps involve
specifying the propositions of the theory, identifying the empirical indicators of the theory, constructing the hypotheses of the theory, and finally,
developing and implementing a plan of research to test and thereby confirm
the trustworthiness of the theory.
Specifying the propositions of the theory requires that the researchertheorist derive truth statements or logical deductions about the theory in
action. These propositions are grounded in the explanatory and predictive
power embedded in the theoretical framework of the theory constructed during the theory development part of the process. This fifth step in Dubin’s
method is of further significance in that it is often at this point in the theorybuilding process that the researcher-theorist ends the journey, for this step
marks the point at which the theory is considered developed and made ready
for further translation to and testing in practice. The taking of the rest of this
research operation side of the applied theory-building process can be the
task of the originating researcher-theorist or others who wish to further test
and confirm, or disconfirm, the theory’s explanatory and predictive capabilities and trustworthiness in practice.
Once the propositions of the theory have been specified, the next (and
sixth) step facing the researcher-theorist is to identify the empirical indicators of the theory. Identifying empirical indicators is necessary to make the
proposition statements testable and is needed for each unit in each proposition whose test is sought (Dubin, 1978).
The seventh step in quantitative theory-building process, and still part of
the research operation side of the cycle, requires that the researcher-theorist
construct hypotheses. This step involves logically substituting the empirical
indicators in the proposition statements with testable, confirmable hypotheses and enables the researcher-theorist to subject the theory to empirical
testing in the real world. The final step in Dubin’s theory-building process is
to engage in the actual testing of the theory through a thoughtfully specified
research plan of ongoing data gathering to enable adequate verification and/
or continuous refinement of the theory. Successful completion of these last
four steps completes the research operation side of the theory-research
cycle and results in an empirically verified and trustworthy theory, as well as
marks the completion of the rigorous theory-building process as outlined by
Dubin (1978).
The Substance of Dubin’s Quantitative
Theory-Building Method
As indicated in Figure 1, Dubin’s quantitative theory-building method is
composed of two parts, one theory development and one research operation,
each of which is further constituted by four specific steps. The essence of
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
Lynham / DUBIN’S METHOD
each of these eight steps that make up this particular applied theory-building
method is further illustrated by way of explicit description and a brief overview example application of Steps 1 through 6. The example comprises
extracts of a theory of responsible leadership for performance (Lynham,
2000a) (full documentation is available from the author) and illustrates the
outcomes of Steps 1 through 6 of this theory-building research method in
action.
The Four Steps of Part 1 of Dubin’s
Quantitative Theory-Building Method
As indicated earlier, and in Figure 1, the first part of the theory-research
cycle embedded in Dubin’s eight-step theory-building research method is
constituted by four theory development steps. The first step requires that the
units of the theory be identified. The second step involves establishing the
laws of interaction applicable to the units of the theory. The third step
requires that the boundaries of the theory be determined. And the fourth and
final step in theory development involves specification of the system states
and their effects on the theory. It is worth reemphasizing that the purpose of
this first, theory development part of the theory-building cycle is to develop
an informed theoretical framework, a system of knowledge that explains the
explanatory essence of the theory in action.
Part 1, Step 1: Identifying the units of the theory. The units of a theory are
sometimes described as the concepts of the theory, or the basic ideas that make
up the theory (Cohen, 1991; Dubin, 1978; Reynolds, 1971). The units represent
the things about which the researcher is trying to make sense and are informed
by literature and experience. By translating these concepts to units, the
researcher-theorist is able to identify the things or variables whose interactions
make up the subject matter of attention (Dubin, 1978), in the example provided
that of responsible leadership for performance (Lynham, 1998, 2000a). It must
be noted that units represent the properties of things rather than the things themselves (Dubin, 1978).
The units of the theory can also be described as “the things out of which
the theory is built” (Dubin, 1976, p. 26). They are the basic building blocks
from which the researcher-theorist constructs the theory, the raw conceptual
material from which the theory emerges (Cohen, 1991; Torraco, 2000).
Description of the units of the theory enables the researcher-theorist to
answer the first theory development question: What are the units of the
theory?
When identifying and selecting the units of a theory, there are five dichotomies of characteristics that must be considered by the researcher-theorist:
(a) unit versus event, (b) attribute versus variable, (c) real versus nominal,
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
247
248
Advances in Developing Human Resources
August 2002
(d) primitive versus sophisticated, and (e) collective versus member (Dubin,
1978). Further to these dichotomies, Dubin identified five types of units
among which the researcher-theorist must distinguish, namely, enumerative, associative, relational, statistical, and summative units. For further
description of these dichotomies and types of units, consult Dubin’s (1978)
text, chapter 3.
The kinds of units used in the theory are determined by the choices made
by the researcher-theorist regarding these dichotomous characteristics and
unit types. The kinds of units used in a theory are important as they can affect
the theory in a number of ways, for example, its structure, the kinds of explanations and predictions the theory can generate, and the extensiveness of the
tests that can be made of the theory (Dubin, 1978).
A further and important requirement of this step in Dubin’s method of
theory building is to consider the outcome of developing the units of the theory against a number of criteria, identified by Dubin (1978) for this first theory development step. Five criteria can be identified for this first step in the
theory development process, namely, rigor and exactness, parsimony, completeness, logical consistency, and the degree of conformity to the limitations on employment and combination of the units. A brief description of
each criterion follows.
The criterion of completeness is linked only to the use of associative units
and the resulting possible zero value of these units. By employing associative units, predictions about the theoretical system, in this example that of
responsible leadership for performance, must include states in which these
units go to zero or even become negative. This issue becomes important for
the eventual testing of the completeness of the theory, that is, of the completeness of the predictions generated by the theory.
The notion of logical consistency relates to the logic of the types of units
combined in and used to compose the theory. The use of only one type of theory unit confines the results of the theory. For example, the use of only
enumerative or only associative units confines the results “to just the first
quadrant of the Cartesian co-ordinate system” (Dubin, 1978, p. 69). On the
other hand, the combination of enumerative with associative units in a single
theory enables a better spread of data across the “four quadrants of the Cartesian co-ordinate system” (Dubin, 1978, p. 70). What units the researchertheorist decides to use in the theory therefore influences the kinds of studies
that can later be used to gather and study data on the theory and, ultimately,
be used to verify and refine the theory.
In the development of the units of a theory, the researcher-theorist is required
to uphold three limiting rules regarding the combination of types of units in a
theory. The first rule states that “a relational unit is not combined in the same theory with enumerative or associative units that are themselves properties of that
relational unit” (Dubin, 1978, p. 73). The second rule states,
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
Lynham / DUBIN’S METHOD
Where a statistical unit is employed, it is by definition a property of a collective. In the same theory do not combine such a statistical unit with any kind of unit (enumerative, associative, or relational) describing a property of members of the same collective. (pp. 74-75)
The third rule states that “summative units have utility in education and communication with those who are naïve in a field. Summative units are not employed
in scientific models” (Dubin, 1978, p. 78).
The preceding discussion concludes the description of the first step in the
theory development phase of Dubin’s (1978) theory-building method. The
outcome of this section is the identification and description of the units of
the theory, as highlighted in Figure 2 and the application example at the end
of this section, and answers the first theory development question: What are
the units of the theory? The task of the next and second step in the theory
development part of this theory-building inquiry method is to consider the
nature of the relationships between the three units of the theory.
Part 1, Step 2: Establishing the laws of interaction that govern the theory.
The second step of theory development in Dubin’s theory building research
method is the specification of the laws of interaction applicable to the units of the
theory. The laws of interaction describe the interaction among the units of the
theory and enable the researcher-theorist to answer the second theory development question: What are the laws of interaction of the theory?
The laws of interaction make explicit and specific the manner in which
the units of the theory interact with one another (Dubin, 1978; Torraco,
1994, 2000). A law of interaction is a statement by the researcher-theorist of
the relationship between units and shows how the units of the theory are
linked to each other. It is important to note that at this stage of the development of the theory, these laws of interaction relate to the relationship
between the units of the theory and not to the conceptual dimensions of each
unit of the theory. Dubin (1978) highlighted three general categories or
types of laws of interaction, namely, categoric, sequential, and determinant.
Categoric laws of interaction indicate that values of a unit of the theory
are associated with values of another unit. This type of law is common in the
social sciences and indicates “a greater-than-chance probability that the
units are related” (Dubin, 1978, p. 98). Categoric laws are symmetrical in
nature, meaning that “it does not matter whether one or the other of the units
comes first in the statement of the law” (Dubin, 1978, p. 100). Words typically employed in this kind of law of interaction are is associated with; for
example, Unit A is associated with Unit B (Dubin, 1978, p. 101).
Two categoric laws of interaction were identified for the theory of
responsible leadership for performance (Lynham, 2000a). The first pertains
to the theory as a whole and the second to the symmetrical nature of the relationship between the three units of the theory. See the corresponding discussion in the application example at the end of this section.
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
249
250
Advances in Developing Human Resources
A Framework of
“Responsibleness”
Considerations of
“Constituency”
August 2002
Effectiveness
Ethics
Endurance
Domains of
“Performance”
System Mission
Work Process/es
Social Sub-systems
Individual Performer
Inside/Outside
High/Low Authority
High/Low Impact
FIGURE 2:
Responsible Leadership for Performance: A Model of the Units of the Theory and
Supporting Conceptual Dimensions
Sequential laws are a second type of law of interaction and make use of a
time dimension to describe the relationships between two or more units. A
sequential law of interaction therefore identifies a temporal interval
between the values of two or more units and indicates that the relationship
between the units concerned is unidirectional. As a result, sequential laws
are asymmetrical, with a time lapse between the units being a characteristic
of this type of law of interaction. Words typically employed in this kind of
law of interaction are succeeded by or preceded by; for example, “specified
values of Unit A are succeeded by specified values of Unit B with a time
interval of X” (Dubin, 1978, p. 103).
Two sequential-type laws of interaction were identified for the theory of
responsible leadership for performance (Lynham, 2000a). These are highlighted in the corresponding discussion and Figures 3 and 4, as well as in the
application example presented at the end of this section.
At the outset of the development of the example theory, it was suggested
by the researcher-theorist that leadership could and should be thought of as a
system, that is, as a set of defined, interdependent parts that interact within
some defined boundary (Dubin, 1978). The theory of responsible leadership
for performance (Lynham, 2000a) is a systems view of leadership and presents leadership as a set of three interdependent units, interacting within the
context and boundaries of a performance system (see corresponding discussion and Figure 4, as well as application example at the end of this section).
A determinant law of interaction is one that relates determinate values of
one unit of the theory with determinate values of another unit. Determinant
laws of interaction therefore describe specific relationships between units
with determinate values, and as a result, these laws of interaction are typically used in the physical sciences where such precise relationships are
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
Lynham / DUBIN’S METHOD
A Framework of “Responsibleness”
Considerations
of “Constituency”
Domains of “Performance”
Time
FIGURE 3:
A Diagram of the First Sequential Relationship Between the Three Units of the
Theory
INPUT:
Considerations of
“Constituency”
PROCESS:
A Framework of
“Responsibleness”
OUTPUT:
Domains of
“Performance”
Time
FIGURE 4:
A Diagram of the Second Sequential Relationship Between the Three Units of the
Theory
more common than in the behavioral sciences (Dubin, 1978; Torraco, 2000).
At this time, there are no determinant laws of interaction in the example theory of responsible leadership for performance.
The two types of laws of interaction employed in the example theory,
namely, categoric and sequential laws, govern the relationships between the
three units of the theory (Dubin, 1978; Torraco, 2000). It must be noted that
laws of interaction do not necessarily indicate causality: “A statement of
interaction or relationship is not necessarily a statement of causality”
(Dubin, 1978, p. 92). Furthermore, Dubin (1978) informed us that laws of
interaction are never in themselves measured. Only the values of the units in
a relationship are measured. The scientific law is therefore tested empirically “if, and only if, values are empirically assigned to the units employed
in the law” (p. 94).
Besides being distinguishable by type, laws of interaction can also be differentiated by levels of efficiency. Dubin (1978) identified four general levels of
efficiency of a law, namely, “(1) presence-absence (lowest level of efficiency);
(2) directionality; (3) covariation; and (4) rate of change (highest level of efficiency)” (p. 109). These levels of efficiency are of a cumulative nature and tend
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
251
252
Advances in Developing Human Resources
August 2002
to correspond with the scientific sophistication of a specific scientific discipline.
According to Dubin,
As the sophistication improves in a given scientific discipline [so] the laws of interaction
employed in that discipline move to a higher level of efficiency. . . . It should be possible to measure the level of sophistication of a given discipline by the efficiency of the laws therein
employed. (p. 111)
Dubin further indicated that the lowest level and the second level of efficiency
are often employed in laws of the social and behavioral sciences and are
accepted as indicating scientific precision in these sciences.
Categoric laws of interaction are always at the lowest level of efficiency,
indicating a presence-absence relationship between the units of the theory.
Sequential laws of interaction, on the other hand, “may achieve any level of
efficiency” (Dubin, 1978, p. 111). The laws of interaction of the example
theory are at the first two (lower end) levels of efficiency of a law, namely,
presence-absence and directionality.
In addition to determining the laws of interaction of a theory, the
researcher-theorist must also consider the outcomes of this second step in
Dubin’s theory-building inquiry method against criteria of excellence for
completion of this step. The outcome of determining the laws of interaction
of a theory can be considered against a single criterion, namely, that of parsimony (Dubin, 1978). As explained in the development of the units of the theory, parsimony relates to the degree to which the theory contains a minimum
of complexity and assumptions. Parsimony, as a criterion of theory development, also refers to the complexity of the laws of interaction employed in the
theory. The higher the level of efficiency reflected in the laws of interaction,
that is, presence-absence, directionality, covariation, or rate of change, the
more sophisticated and complex the laws of interaction are considered to be.
In the example theory, briefly presented in the application example at the
end of this section, the two categoric laws are at the lowest level of efficiency
(presence-absence), whereas the two sequential laws are at the second lowest level of efficiency (directionality).
Besides the efficiency level of the laws of interaction of the theory, the
criterion of parsimony also relates to the maximum versus the minimum
number of laws required to relate the units of a theory at least once with each
other (Dubin, 1978). To this end, the key laws of interaction presented in the
example theory (see application example at end of section) are considered a
minimum to make explicit and clear the nature of the interrelationship
between the three units of the theory. By restricting the laws of interaction of
the theory to currently low levels of efficiency and minimum numbers, the
researcher-theorist attempted to ensure further parsimony of her theory of
leadership (Lynham, 2000a). This section concludes the second theory
development step in Dubin’s theory-building method. The outcome of this
step enables the researcher-theorist to answer the second theory develop-
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
Lynham / DUBIN’S METHOD
ment question: What are the laws of interaction of the theory? Once the units
and the laws of interaction of the theory have been established, the next and
third step in this applied theory–building method is to determine the boundaries of the theory.
Part 1, Step 3: Determining the boundaries of the theory. The boundaries of a
theory are established to determine and clarify the domains within which the
theory is expected to hold up and apply (Dubin, 1978). It must be remembered
that a theory is an attempt by the researcher-theorist to model some theoretical
aspect of the real world (Torraco, 1997). As such, a theory is limited to an aspect
of the real world that it is trying to model. The boundaries of a theory therefore
establish the real-world limits of the theory and in so doing distinguish the theoretical domain of the theory from those aspects of the real world not addressed or
explained by the theory.
The determination and clarification of the boundaries of a theory require
response to the third of the four theory development questions: What are the
boundaries of the theory? By addressing this theory development question,
the researcher-theorist is able to complete Step 3 of Dubin’s theory-building
inquiry method and compare the outcome to the criteria of excellence for
this third theory development step. The boundaries of a theory are important
in that they enable the researcher-theorist to make clear and explicit the limited portions of the world within which the theory is expected to hold
(Dubin, 1978). Of equal import is that the boundaries of a theory also enable
the researcher-theorist to represent the theoretical framework of the theory
as an empirical and bounded system of knowledge and explanation (Dubin,
1978).
Dubin (1978) distinguished between a closed and an open boundary (p.
126), advocating the use of an open boundary “when there is exchange over
the boundary between the domains through which the boundary extends”
(Torraco, 1994, p. 162). On the other hand, Dubin advocated a closed boundary when “exchange does not take place between the domains through which
the boundary extends” (Torraco, 1994, p. 162).
When using a theory-to-research strategy of theory building, as is inherent to the nature of Dubin’s method of theory building, the boundaries of a
theory are determined not by empirical data but rather through the use of
logic. The boundary of the theory must be chosen, through the logic of the
researcher-theorist, to indicate the “domain over which the [theory] operates as a system” (Dubin, 1978, p. 141). The boundary of the theoretical system acts to specify “the furthest extension over the empirical world that the
[theoretical] model is expected to operate” (Dubin, 1978, p. 141). The
domain of the theory therefore becomes “that portion of the empirical world
included within the boundaries” (Dubin, 1978, p. 141). Dubin (1978) indicated that “the domain of a [theory] is always bounded” and that “to determine the domain of a [theoretical] model requires the determination of its
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
253
254
Advances in Developing Human Resources
August 2002
The Broader External Environment in which the Theory Operates and Other Kinds
of Performance Systems in this Broader External Environment
THE LEGALLY DEFINED, HUMAN POPULATED
PERFORMANCE SYSTEM: The Closed Boundary of the Theory
The External Environment to the Performance System
The Social, Economic, Cultural, Technological and Political
Environment in which the Outside Constituency Resides
The Internal-External Environment Boundary of the
Performance System: The Open Boundary of the Theory
The Internal Environment to the Performance System
•
Cultural
•
Political
•
Technology
The Focused System of Responsible Leadership for Performance
Input
Process
Outputs
Considerations of
“Constituency”
A Framework
of
“Responsibleness”
Domains
of
“Performance”
© S. A. Lynham, 2000
Key:
The Closed Boundary of the theory
The Open Boundary of the theory
FIGURE 5: The Boundaries of a Theory of Responsible Leadership for Performance
Note: © S. A. Lynham, 2000.
boundaries” (p. 141). By way of example, the boundaries and corresponding
domains of the theory of responsible leadership for performance (Lynham,
2000a) are presented in the application example at the end of this section and
Figure 5.
Once the boundaries of the theory have been determined, the researchertheorist is required to compare the resulting boundaries against two steprelated criteria of excellence identified by Dubin (1978), namely, homogeneity (including the satisfaction of three specific internal and external
boundary-determining requirements) and generalization.
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
Lynham / DUBIN’S METHOD
The criterion of homogeneity requires that “the units employed in the
theory and the laws by which they interact satisfy the same boundarydetermining criteria” (Dubin, 1978, p. 127). Dubin (1978) further specified
that “a theoretical model is said to be bounded when the limiting values on
the units comprising the model are known. The limiting values are always
determinate” (p. 126). In comparing the output of this third theory development step with the corresponding boundary-determining criteria, it is
important to first clarify some basic related concepts, namely, boundary
criteria, interior boundary-determining criteria, and external boundarydetermining criteria.
“The boundary-determining criteria of a [theory],” said Dubin (1978),
“apply with equal force to the units employed and the laws of interaction
among these units. The units [of a theory] must fit inside the boundaries
before the [theory] is complete” (pp. 126-127). Internal boundarydetermining criteria are those that are “derived from the characteristics of
the units and the laws employed in the [theory]” (p. 128). Dubin specified
four possible internal boundary-determining criteria, namely, truth tables of
the logician, establishing a “limit of probability on the values taken by the
units employed in a theory” (p. 130), “subsetting the property space” (p. 131),
and alignment between the relationships specified in the laws of interaction and those included in the domain of the theory. External boundarydetermining criteria, on the other hand, “are those imposed from outside
the [theory]” (p. 132).
The application of internal and external boundary-determining criteria is
related to the theory-building strategy employed by the researcher-theorist
in developing the theory. In a research-to-theory strategy of theory building,
these criteria are determined through empirical research and measurement
of the units and laws of interaction employed in the theory. In this type of
theory-building strategy, the empirical data gathered through research are
then used by the researcher-theorist to derive the boundaries of the theory
(Dubin, 1978; Reynolds, 1971). In a theory-to-research strategy of theory
building, the researcher-theorist determines the boundaries of the theory
logically (Dubin, 1978).
The criterion of generalization of a theory relates to domain size of the
theory. Thus, the bigger the domain, the more general the theory (Dubin,
1978). Reducing the number of boundary-determining criteria also serves to
enlarge the domain of the theory. Employing two boundaries in the example
theory and applying two boundary-determining criteria enabled the
researcher-theorist to limit the generalization of this theory of leadership.
As indicated in the application example at the end of this section, the theory
of responsible leadership for performance (Lynham, 2000a) is not an unlimited, unbounded theory of leadership. At the same time, the somewhat broad
domain of the theory can later be shown to have implications for understand-
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
255
256
Advances in Developing Human Resources
August 2002
ing and including knowledge from other leadership theories into that of the
theory of responsible leadership for performance.
This section concludes discussion of the essences of the third step in the
theory development part of Dubin’s theory-building method. The outcome
of this third step is the determination and clarification of the two boundaries,
one open and one closed, of the example theory, as presented in Figure 5 and
the application example at the end of this section. These boundaries make
clear and explicit the real-world domain over which the example theory is
expected to apply and hold up.
Clarification of the boundaries of the theory enables the researchertheorist to answer the third theory development question: What are the
boundaries of the theory? The next and fourth step in the first theory development part of Dubin’s applied theory-building method is that of specifying
the system states of the theory.
Part 1, Step 4: Specifying the system states of the theory. A system state is a
condition of the system being modeled in which the units of the theory interact
differently. Thus, a system state represents a condition under which the theory is
operative (Dubin, 1978; Torraco, 1994, 1999, 2000). Specifying the system
states under which the theory is expected to operate requires response to the
fourth and last theory development question: What are the system states of the
theory? Through response to this theory development question, the researchertheorist is able to complete Step 4 and the first part (the theory development side)
of Dubin’s applied theory-building method.
Dubin (1976, 1978) defined a system state as a condition of the system
being modeled in which all the units of the system take on characteristic values that have persistence through time, regardless of the length of the time
interval. All units of the system have values that are determinant, meaning
they are measurable and distinctive for that state of the system.
Dubin (1978) further identified three criteria of importance to the
researcher-theorist when identifying the system states of the theory, namely,
(a) inclusiveness, (b) persistence, and (c) distinctiveness. The criterion of
inclusiveness refers to the need for all the units of the system to be included
in the system state of the theory (Dubin, 1978; Torraco, 1994, 2000). The
criterion of persistence requires that the system state persist through a
meaningful period of time (Dubin, 1978; Torraco, 1994, 2000). And the criterion of distinctiveness requires that all units take on determinant, that is,
measurable and distinctive, values for the system state (Dubin, 1978;
Torraco, 1994, 2000).
The specification of the system states of the theory enables the
researcher-theorist to answer the fourth, theory development, question:
What are the system states of the theory? The completion of this fourth step
also enables the researcher-theorist to conclude the first theory development
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
Lynham / DUBIN’S METHOD
part of the quantitative theory-building method (Dubin, 1978). It should be
recalled that an important outcome of this first theory development part of
Dubin’s two-part, eight-step theory-building method is an informed, conceptual, or theoretical framework of the theory.
At this stage of the theory-building process, the four theory development
questions have been addressed by describing the units, specifying the laws
of interaction, determining the boundaries, and identifying the system states
of the theory. Addressing the four theory development questions thus
enabled the researcher-theorist to develop and present an informed conceptual and theoretical framework, that is, a theoretical system of knowledge,
of the example theory, as indicated in the corresponding discussion and Figure 6, as well as in the application example following.
Application example. At the outset of this section discussion, four steps were
identified as necessary for the development of the example theory of responsible
leadership for performance (Lynham, 2000a), namely, (a) a description of the
units of the theory, (b) specification of the laws of interaction of the theory, (c)
determination of the boundaries of the theory, and (d) identification of the system states of the theory. The conclusion of each of these four steps allowed the
researcher-theorist to answer the following four corresponding theory development questions and conclude the theory development side of the theory-research
cycle embedded in Dubin’s eight-step applied theory-building method: (a) What
are the units of a theory of responsible leadership for performance? (b) What are
the laws of interaction of a theory of responsible leadership for performance? (c)
What are the boundaries of a theory of responsible leadership for performance?
and (d) What are the system states of a theory of responsible leadership for
performance?
In answering the first of these four theory development questions, three
units and their conceptual dimensions were described for the theory (see
Figure 2). The first unit identified was considerations of constituency, with
three conceptual dimensions being identified as integral to this unit. These
were whether the constituency resides inside or outside the performance
system in which the leadership occurred, whether the constituency has high
or low authority over the performance system concerned, and whether the
constituency has high or low potential impact over that performance system.
The second identified unit of the theory was a framework of responsibleness, consisting of three conceptual dimensions, namely, effectiveness, ethics, and endurance (White Newman, 1993). The third unit identified in the
theory was domains of performance. The four conceptual dimensions of this
third unit of the theory included systems mission, work processes, social
subsystems, and the individual performer (Holton, 1999).
The rationale and theoretical support for each of these three units and
their conceptual dimensions were considered in detail. The three units were
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
257
258
Advances in Developing Human Resources
August 2002
External Environment to the Performance System which
Informs the Leadership System-in-Focus (RLP)
The Social, Economic, Cultural, Technological and Political Environment
in which the Outside Constituency of the Performance System Resides
The Performance System in which the Leadership System-in-Focus (RLP) Operates
(Also the Open Boundary of the Theory)
Internal Environment to the Performance System in which the
Leadership System-in-Focus (RLP) Operates
• Cultural
• Political
• Technological
The Leadership System-in-Focus: RLP
INPUT
Considerations of
“Constituency”
•
•
•
Inside/Outside
High/Low
Authority
High/Low Impact.
PROCESS
OUTPUTS
A Framework of
Domains of
“Performance”
“Responsibleness”
•
•
•
Effectiveness
Ethics
Endurance.
•
•
•
•
System Mission
Work Process/es
Social Sub-systems
Individual Performer.
© S. A. Lynham, 2000
FIGURE 6:
The Theoretical Framework of a Theory of Responsible Leadership for
Performance
Note: © S. A. Lynham, 2000.
also shown to satisfy the five criteria for identification of the units of a theory, namely, rigor and exactness, parsimony, completeness, logical consistency, and degree of conformity to the limitations on employment and combination of the units.
In answering the second theory development question, four laws of interaction were specified for the example theory of responsible leadership for performance, two of a categoric (associative) nature and two of a sequential (temporal)
nature. The first two categoric (associative) laws of interaction are as follows:
• Law 1: All three units of the theory, namely, considerations of con-
stituency, a framework of responsibleness, and domains of performance, are associated with and required for responsible leadership
for performance, and
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
Lynham / DUBIN’S METHOD
• Law 2: Each unit of the theory, namely, considerations of constitu-
ency, a framework of responsibleness, and domains of performance,
interrelates with each other unit of the theory.
These two associative laws gave rise to the first principle that informs the system states of the example theory, namely, that in the absence of any one of the
three units and interaction among the units, the theoretical system of responsible
leadership for performance is either unbalanced or destroyed. Law 3, the first
sequential (temporal) law of interaction of the theory, states that considerations
of constituency precede a framework of responsibleness and domains of performance in the theory of responsible leadership for performance (see Figure 3).
Law 4, the second sequential law of the theory, illustrated in Figure 4,
specifies that the units of the theory of responsible leadership for performance can be thought of in terms of inputs, process, and outputs, where the
unit of considerations of constituency forms the input, the unit a framework
of responsibleness forms the process, and the unit domains of performance
forms the output of the theory. These two temporal laws gave rise to the second principle that informs the system states of the theory, namely, that
responsible leadership for performance results only when the three units of
the theory are arranged in an interacting system of inputs, processes, and
outputs (see Figure 4).
The supporting logic to the four laws of interaction of the theory of
responsible leadership for performance was provided. The four laws of
interaction were also shown to satisfy the criterion of parsimony required in
the specification of the laws of a theory.
In response to the third theory development question, two boundaries,
one closed and one open, were determined and clarified, as indicated in Figure 5. The first boundary, a closed boundary, determined the overall domain
over which the example theory of responsible leadership for performance is
expected to hold up. This boundary was shown to have two distinctions. The
first was that of a legally defined, human-populated performance system,
and the second was the social, political, technical, economic, technological,
and cultural external environment in which the outside constituency of the
performance system resides (see the solid black border line of Figure 5).
These two boundary-determining features—of a legally defined, humanpopulated performance system and the social, political, economic, technological, and cultural external environment in which outside constituency
resides—were shown to separate the domain of the theory from other kinds
of performance systems and the broader external environment not addressed
by the theory. The second boundary, an open boundary, determined in the
development of the theory was shown to exist within the domain of all
legally defined, human-populated performance systems (shown by the dash
line in Figure 5). This second boundary allowed the system of responsible
leadership for performance to be understood as a leadership system-in-
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
259
260
Advances in Developing Human Resources
August 2002
focus and corresponded with the internal-external boundary of the performance system in which the leadership system occurred.
The clarifying logic behind the determination of these two boundaries of
the theory was presented. The two boundaries presented in the theory of
responsible leadership for performance were also shown to satisfy the two
criteria for determining the boundaries of a theory, namely, homogeneity
and generalization.
Finally, to address the fourth and last theory development question, two
system states were identified for the example theory of responsible leadership for performance, namely, that of balance and unbalance. To identify
these system states, the system being focused on in the theory was first
shown to be that of responsible leadership for performance, as presented in
Figure 6.
Some characteristic values of each of the three units of the theory in each
of the two system states of responsible leadership for performance were
identified and shown to satisfy the three criteria of inclusiveness, persistence, and distinctiveness required for the identification of the system states
of a theory. The supporting logic behind the two system states of the theory
was also presented.
This concludes the presentation and discussion of the substance of the
theory development side and the first part of the quantitative theory-building
method. Making explicit the substance of the next part of the theory-building
method requires description of the research operation side of Dubin’s
(1978) theory-building process.
The Four Steps of Part 2 of Dubin’s
Quantitative Theory-Building Method
The second part of the process requires the completion of Steps 5 to 8 of
Dubin’s two-part, eight-step theory-building inquiry method. This part of
the theory-building method therefore includes specifying the propositions
of the theory, identifying the empirical indicators for the key terms within
the theory, constructing hypotheses to enable testing of the theory, and
developing a purposeful research agenda to test the theory and/or its component parts in action.
It should be remembered that the second part of Dubin’s theory-building
method, also known as the research operation side of the two-part theorybuilding cycle, has as its main purpose the verification, or confirmation, of
the theoretical framework completed during the first theory development
part of the theory-building cycle. Through the address of the four steps central to the second part of the theory-building process, the theoretical framework that was the outcome of Part 1 of the process is operationalized by
completing Steps 5 through 7 of Dubin’s eight-step theory-building method
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
Lynham / DUBIN’S METHOD
and then tested through completion of Step 8 in Dubin’s quantitative theorybuilding process. Operationalization of the theoretical framework developed in Part 1 is achieved by first specifying propositions derived from the
framework, then identifying corresponding empirical indicators informed
by the propositions, and finally constructing hypotheses based on the propositions and informed by the empirical indicators.
Once the theoretical framework has been operationalized, the researchertheorist can begin the conduct of related research to test and confirm, or
indeed disconfirm, the theoretical framework in practice, or action. Neither
operationalization nor testing of the theory through research is a one-time
event. Rather, operationalization and testing of the theory should be understood as an ongoing, accumulative process of theory testing, confirmation,
refinement, and further development. This continual refrain between the
theory development and research operation sides of the two-part, eight-step
theory-building inquiry method is meant to ensure that the researcher-theorist
constantly adapts, tests, and readapts the theory until she or he has confidence in the trustworthiness of the theory in explaining action and outcome
in the real world and/or continues to adapt the theory to incorporate new
insights, knowledge, and experience to this predictive and explanatory end.
The sections following provide a brief description of the substance of the
research operation side, second part, involving Steps 5 through 8, of Dubin’s
two-part, eight-step theory-building method. To the extent that the example
theory of leadership was partially operationalized, an application subsection on the related theory-building steps is included in the related theorybuilding step descriptions.
Part 2, Step 5: Specification of the propositions of the theory. Propositions
are, according to Dubin (1978), “truth statements about the theory” (p. 160) and
are concerned with “the ways in which a theoretical model are put to use” (p. 159).
The propositions of a theory are considered true by virtue of being statements
that are logically derived from the theory itself (Dubin, 1978; Torraco, 2000).
Because propositions can be subjected to empirical testing, they, in turn, enable
the theory to be subjected to empirical testing (Torraco, 2000). The identification of propositions is therefore a first and necessary step in operationalizing the
theory, or in getting the theory ready to be put to the test. Propositions enable the
researcher-theorist to begin to make predictions from the theoretical framework
about the values of the units of the theoretical framework in the real world. Specification of the propositions of a theory enables the researcher-theorist to answer
the fifth question in Dubin’s theory-building method and the first research operation question in this theory-building process: What are the propositions of the
theory?
It is important to note, when considering propositions of a theoretical framework, that these truth statements “are not necessarily truth statements about
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
261
262
Advances in Developing Human Resources
August 2002
aspects of the real world that the theory represents” (Torraco, 1994, p. 166).
Indeed,
to address the problem of matching the theory with the real world the theory is intended to model,
it is necessary to convert the proposition statements, first to empirical indicators, and then into
hypotheses, and then to test the hypotheses through research. (Torraco, 1994, p. 166)
The issue of the conversion of the propositions of the theory to empirical indicators and hypotheses is dealt with in Part 2, Steps 6 and 7, of Dubin’s theorybuilding method.
Dubin (1978) suggested that the propositions of a theory are “constructed
logically and intellectually by the theorist” (p. 164). According to Dubin,
this logic and intellectual nature of propositions make proposition statements “synthetic,” meaning that “these are statements that follow as true
from the [theoretical framework] about which they are made” (p. 164). Furthermore, he indicated that “this synthetic quality of propositions makes
clear that we are not talking at this point about the empirical accuracy of the
proposition statements” (p. 164). According to Dubin (1978) and Torraco
(1994), proposition statements of a theory must be considered true if they
are logically derived from a theoretical system of specified units, laws of
interaction, boundaries, and system states. Dubin emphasized this internal
truth to proposition statements further: “The sole test of the accuracy of a
proposition is whether or not it follows logically from the [theoretical
framework] to which it applies” (p. 164).
Dubin (1978) highlighted three types of proposition statements. The first
type may be made “about the values of a single unit of the [theoretical framework], the value of that unit being revealed in relation to the value of the
other units connected to the unit in question by a law of interaction” (p. 166).
The second type of propositions “may be predictions about the continuity of
a system state that in turn involves a prediction about the conjoined values of
all units in the system” (p. 166). And the third type may involve “predictions
about the oscillation of the system from one state to another . . . and involves
predictions about the values of all units of the system as they pass over the
boundary of one system state into another” (p. 166). According to Dubin,
these three types of predictions “exhaust the logical possibilities” of the predictions and represent the “full range of truth statements” (p. 166) that can
be logically derived from the theoretical framework about the theory in
action.
“A proposition of a theoretical [framework] is, therefore, a truth statement about the [theoretical framework] in operation” (Dubin, 1978, p. 163).
Typically, proposition statements take the form of “If . . . then” (Dubin,
1978, p. 164). Cohen (1991) referred to propositions as the knowledge
claims of the theory in action. Because propositions are truth statements
about the theory in use, it can be safely assumed that propositions can be
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
Lynham / DUBIN’S METHOD
made from any theoretical framework “that has had its units, laws of interaction, boundaries and system states specified” (Torraco, 1994, p. 166). The
logic of these propositions is informed by the logic of the theoretical framework of the theory.
Further to the specification of the propositions of the theory, Dubin
(1978) indicated that in specifying propositions for a theory, there are three
criteria for consideration by the researcher-theorist, namely, consistency,
accuracy, and parsimony. The criterion of consistency in specifying the
propositions of a theory refers to the notion that the truth of the propositions
“be established by reference to only one system of logic for all the propositions set forth in the [theoretical framework]” (p. 160). To meet this criterion
in the example theory of responsible leadership for performance (Lynham,
2000a), the logic of the systems framework was used to guide the specification of the propositions. A system has specific characteristics, namely, a
name, a purpose, parts, interactions among the parts, and outcomes or outputs (Dubin, 1978; Senge, 1990; Swanson, 1994; Von Bertalanffy, 1968).
Each of the propositions of this theory is associated with one or more of
these system characteristics to ensure consistency in the propositions specified for the theory.
The criterion of accuracy refers to whether the propositions follow logically from the theoretical framework to which they apply. Each of the propositions specified for the example theory is informed by and follows from the
specified system of units, laws of interaction, boundaries, and system states
that make up the theoretical framework of the theory. The logic of these
components of the theoretical framework of responsible leadership for performance is maintained in the logic of the specified propositions of the
theory.
The criterion of parsimony, when considered in relation to the specification of the propositions of a theory, refers to the use of what Dubin (1978)
called “strategic propositions” (p. 166). Dubin pointed out that “in principle, every [theoretical framework] should give rise to an infinite number of
propositions” (p. 166). However, it is not the job, in the specification of the
propositions of a theory, to identify all possible propositions for that theory.
It is, according to Dubin, more important to seek some parsimony in the
specification of propositions. This is achieved by reducing the large number
of possible propositions to a handful “that satisfies our sense of what is
important to test in determining whether or not the theory accurately models
the empirical domain it purports to represent” (p. 167). Fifty-three potential
propositions were initially specified from the theoretical framework of
responsible leadership for performance. Through application of the framework of system characteristics indicated in the above criterion of consistency, these propositions were then sorted and reduced to the 8 specified in
the application example following. It was felt by the researcher-theorist that
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
263
264
Advances in Developing Human Resources
August 2002
these 8 propositions were of a strategic (rather than a trivial) nature in terms
of future testing of the theory.
Application example. Eight propositions were specified for the example theory of responsible leadership for performance (Lynham, 2000a) and covered the
first two types of proposition statements identified by Dubin, namely, the value
of units of the theory in relation to other units and connected by the laws of interaction and predictions about the conjoined values of all the units in the theory.
The eight proposition statements logically derived from the theoretical framework of the theory of responsible leadership for performance are as follows.
1. Responsible leadership for performance is a theory of leadership as a
system-in-focus. Leadership is a purposeful, focused system not an individual or a process managed by an individual.
2. All systems have a purpose. The purpose of responsible leadership for
performance is to serve the needs and desired outcomes of the constituency of a performance system by positively affecting multiple domains
of performance in a responsible (effective, ethical, and enduring)
manner.
3. The content of responsible leadership for performance is derived from all
three units of the theory—considerations of constituency, a framework
of responsibleness, and domains of performance. If all three units are not
present and interacting, then there is not responsible leadership for
performance.
4. For leadership to be considered responsible, it must demonstrate, and be
judged to demonstrate, effectiveness, ethics, and endurance. If one of
these three attribute properties is missing from leadership, then that leadership can no longer be considered responsible.
5. The units of the theory—considerations of constituency, a framework of
responsibleness, and domains of performance—are interdependent. If
there is change in one unit, then it can be expected that there will be
changes in the other two units.
6. As responsibleness (effectiveness, ethics, and endurance) increases, performance of the whole performance system can be expected to increase.
7. Constituency is a necessary requirement for leadership. Without constituency, there is no leadership.
8. Without guiding inputs from constituency, and outputs in the form of
performance, the phenomenon of leadership collapses.
Specification of the propositions of the theory enables the researcher-theorist
to answer the research operation question of, What are the propositions of the
theory? Once the propositions of the theory have been specified, the researchertheorist is able to proceed to the second step in the operationalization of the theory, and the sixth step in Dubin’s two-part, eight-step theory-building method,
namely, to identify empirical indicators for the theory.
Part 2, Step 6: Identifying empirical indicators of the theory. In the sixth step
of Dubin’s (1978) theory-building method, the researcher-theorist is essentially
concerned with “finding the empirical indicators that produce the values on the
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
Lynham / DUBIN’S METHOD
units employed in the (theoretical) model” (p. 181). Completing this next step
enables the researcher-theorist to answer the theory-operationalizing question,
What are the empirical indicators of the theory?
An empirical indicator is, according to Dubin (1978), “an operation
employed by the researcher[-theorist] to secure measurements of values on
a unit” (p. 182). “In order that a unit be measured to indicate a value on it, the
thing itself must be apprehended by an empirical indicator” (p. 181). Identification of empirical indicators therefore involves a process of measurement
of the unit concerned. This process of measurement contains two parts;
namely, “both the operation of measuring [must be duplicable by others]
and the results or value produced [must produce equivalent values when
measured by others] by this operation” (p. 182). Empirical indicators must
therefore satisfy the two principal criteria of what Dubin called
“operationism” (p. 183) and “reliability” (p. 185).
Further to these two parts to empirical indicators, Dubin (1978) also
identified two classes of empirical indicators. The first is the class of “absolute indicators,” which refer to indicators that “have no question as to what
they measure” (p. 193). An example of an absolute empirical indicator in the
social sciences would be race, sex, or age. The second class refers to “relative indicators” (p. 195). “The primary characteristic of a relative indicator
is that it may be employed as an empirical indicator of several different theoretical units” (p. 195). An example offered by Dubin of a relative empirical
indicator is income, which can be employed as a measure of both economic
position and social class position.
The types of units used in a theory affect the empirical indicators identified. The example theory of responsible leadership for performance uses
three types of units, namely, enumerative (considerations of constituency),
associative (domains of performance), and relational (a framework of
responsibleness). It will be recalled that an enumerative unit, like that of
considerations of constituency, “is a characteristic of a thing in all its conditions” (Dubin, 1978, p. 186). “This definition suggests that any empirical
indicator used to establish the value of an enumerative unit has to generate
non-zero values for that unit in whatever conditions the unit is found”
(Dubin, 1978, p. 186). Thus, the indicator used for an enumerative unit in a
theory may never produce a zero value. In the theory of responsible leadership for performance, it is clear that without constituency, there can be no
leadership. An associative unit may be understood as “a property characteristic of a unit in only some of its conditions” (Dubin, 1978, p. 187). The difference between an enumerative unit and an associative unit is that an associative unit, like that of domains of performance, must have a real zero
value. In the theory of responsible leadership for performance, the unit of
domains of performance can show a zero value of performance output,
meaning that leadership has had no impact on performance in the perfor-
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
265
266
Advances in Developing Human Resources
August 2002
mance system concerned. Indeed, this associative unit of the theory may be
indexed by a continuum of negative, zero, and positive values.
The third type of unit employed in the example theory is that of a relational unit, which “is a property characteristic of a thing that can be determined only by the relation among the properties of the thing” (Dubin, 1978,
p. 188). The unit of responsibleness in the theory of responsible leadership
for performance is of a relational type. Responsible leadership only exists
due to the functional relationship between effectiveness, ethics, and endurance. As indicated in one of the propositions of the theory, unless leadership
demonstrates effectiveness, ethics, and endurance, that leadership cannot be
considered responsible. A zero value in any one of these three property characteristics will render the value of responsible leadership zero, or “not
responsible.”
Further to the requirements for the identification of empirical indicators
of the theory, Dubin (1978) specified two conditions or criteria for the process of measurement inherent in the identified empirical indicators of a theory. The first is that the operation of measurement is specified, and the second is that the results or values produced by this operation are identified.
Both of these criteria are partially met in the empirical indicators identified
for the example theory. However, more rigorous refinement and specification of the operations of measurement and values produced by these operations will need to be developed by the researcher-theorist. This is, in part,
the task of Step 8, the testing of the theory through research.
Further to these two criteria, it is also required that the values of the unit
indicators comply with the value requirements for the three types of units
employed in the example theory—enumerative, associative, and relational.
The requirements of empirical indicators for the three types of units of the
example theory were taken into account and adhered to in the identification
of the empirical indicators (see Table 1).
Completion of this sixth step in the theory-building method allowed the
researcher-theorist to begin to address the second research operation question of, What are the empirical indicators of the theory? It will be recalled
that the completion of the fifth step of Dubin’s (1978) theory-building
method allowed the researcher-theorist to address the first research operation question: What are the propositions of the theory? A third operationrelated question in the theory-building process is, What are the hypotheses
of the theory? Once the researcher-theorist has responded to all three of
these theory-operationalizing questions, the theory is considered ready to
take to the real world for testing, verification, or confirmation and further
refinement.
This further operationalizing of the theory must occur through the construction of hypotheses (theory-building Step 7), derived from the propositions and informed by the empirical indicators of the theory. Once the theory
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
Lynham / DUBIN’S METHOD
TABLE 1:
Example of Two Empirical Indicators for Application Example Theory
Unit
Empirical Indicator
Responsibleness
Performance
Part 1: The value of responsibleness increases as measured by
effectiveness, ethics, and endurance, where effectiveness is
measured by constituency perception of effective leadership
practices (White Newman, 1993), ethics is measured by
constituency perception of ethical leadership habits (Posner &
Kouzes, 1996; White Newman, 1993), and endurance is
measured by constituency perception of the employment of
enduring resources (Collins & Porras, 1994; White Newman,
1993).
Part 2: The value of whole-system performance increases as
measured by mission-related, work process, social subsystem,
and individual units of performance in terms of time, quality,
quantity, and alignment (Gibson, Ivancevich, & Donnelly, 1994;
Holton, 1999; Rummler & Brache, 1995; Swanson, 1996;
Swanson & Holton, 1999).
has been made ready for testing, that is operationalized, it can then be verified or confirmed through the conduct of testing research (theory-building
Step 8). Refinement of the theory is an ongoing process and occurs through
continued testing and research and the use of the outcomes of this testing
and research to both improve and further prove the theory.
Part 2, Step 7: Constructing the hypotheses to test the theory. Before the theory can be tested and verified through research, hypotheses must be constructed
from the theory, a step that coincides with Step 7 of Dubin’s (1978) theory-building
method. Step 7 is required to establish the link between the theoretical framework and the researcher-theorist’s understanding of the real world (as reflected
in the theoretical framework). This linkage between the theoretical framework
and the real world is made possible by translating some of the propositions of the
theory to testable hypotheses, a process further informed by the empirical indicators identified for the units of the theory. Step 8 is then required to verify the
theory in action in the real world by putting the constructed hypotheses to the test
through research.
Step 7, construction of the hypotheses of the theory, represents the completion of the theoretical formulation of and from the theoretical framework
of the theory (Dubin, 1978). “An hypothesis,” according to Dubin (1978),
“may be defined as the predictions about the values of the units of a theory in
which empirical indicators are employed for the named units in each proposition” (p. 206). The hypotheses of a theory therefore tell us what the empirical indicators for the units—in the example theory, for constituency, respon-
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
267
268
Advances in Developing Human Resources
August 2002
sibleness, and performance—are and “mirror the propositions” (p. 206) of
the theoretical framework.
Dubin (1978) highlighted two criteria of import in constructing hypotheses of the theory. First, he stated that “every hypothesis is homologous with
the proposition for which it stands” (p. 207), indicating homology between
the hypotheses and their corresponding propositions as a necessary criterion
of excellence for this seventh step in the theory-building process. Second, he
stated that to determine hypotheses, the researcher-theorist must identify
the necessary and sufficient conditions of each unit of the theory. These necessary and sufficient conditions of each unit of the theory are obtained from
the definition, and component parts, of each unit of the theory.
In constructing the hypotheses of the theory, the researcher-theorist is
confronted with a decision of quantity; that is, How many hypotheses are
enough for testing the theory? Dubin (1978) offered some noteworthy
insights in this regard. First, he suggested that “every proposition [in the theory] has the potential of being converted to a large number of hypotheses”
(p. 208). On this point, he proceeded to say that “the general rule is that a
new hypothesis is established each time a different empirical indicator is
employed for any one of the units designated a proposition” (p. 209). He correspondingly further indicated that “this rule establishes the fact that the
number of hypotheses is rapidly expanded as skill and imagination are utilized in developing empirical indicators” (p. 209).
Second, Dubin (1978) did suggest that not all propositions of the theory need
to be converted to hypotheses:
There is no logic that insists that only those [theoretical frameworks] whose propositions are all
testable should be employed in science. Indeed this restriction, representing the extreme positions of operationalists, has proved needless. All theoretical [frameworks] need to be testable by
converting at least some of their propositions to hypotheses. It is not a requirement, however, that
all propositions of a [theoretical framework] be testable. (p. 209)
Ultimately, this question of a sufficient number of hypotheses, according to
Dubin, “poses a question of research efficiency” (p. 209).
It is therefore clear that various strategies may be employed in the formulation of hypotheses, and these strategies are usually chosen in accordance
with available research time and resources. These strategies can vary from
being extensive, where all the strategic propositions of the theory are tested;
to focusing attention on one or several strategic propositions; to being ad
hoc, where hypotheses are generalized to propositions. An important quality of these alternative strategies for hypotheses construction is that they
should not be considered mutually exclusive by the researcher-theorist.
Construction of the hypotheses of a theory clearly performs an important
theory-building and research operation function in that it allows the
researcher-theorist to answer the research operation question, What are the
hypotheses of the theory? Stated in another way, the basic question the
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
Lynham / DUBIN’S METHOD
researcher-theorist is concerned with in this seventh step in Dubin’s (1978)
theory-building process is one of, “What is the reality test we can make of
this theoretical [framework]?” (p. 206). The answer to this question is ultimately provided by the hypotheses construction strategy, or strategies,
employed by the researcher-theorist in the conduct of this step in the theorybuilding process.
Part 2, Step 8: Testing the theory through a developed plan of research. Step 8
involves the testing of the theory through empirical research. This step represents the fourth step in the second part of Dubin’s (1978) two-part, eight-step
theory-building method and allows the researcher-theorist to answer the final
research operation and theory verification question: What is the critical research
needed to verify the theory? To address this question, the researcher-theorist
must decide on his or her research stance toward the theory. This stance involves
either proving the adequacy of the theoretical framework or improving the
starting theoretical framework. The choice of research stance will inform and
determine the type and extent of research testing undertaken by the researchertheorist. For example, doing research for the purpose of improving the theoretical model involves descriptive research, whereas doing research to test the theory involves hypothesis testing (Dubin, 1978).
Outcomes, or research results, from this eighth step in Dubin’s quantitative theory-building method also inform the ongoing refinement of the theory. This overlapping link between theory verification and theory refinement is encapsulated in a research stance aimed at continuously improving
the theory.
Implications of Dubin’s Quantitative
Theory-Building Method for
Applied Theory-Building
Dubin (1978) used the concept “perpetual theory building” (p. 220) to
indicate that one is never done with theory building. This notion of ongoing
theory building, of ongoing refinement of the theory, speaks to the recursive
deductive-inductive refrain between the results of research and testing and
the use of those results to further inform and continually develop and
improve the theory. This implies an ongoing refrain between theory development, refinement, and research.
The hypothetico-deductive method of theory building, reflected in the
applied theory-building methods of Cohen (1991), Reynolds (1971), Hearn
(1958), and Dubin (1978), is based on the assumption of falsification
(Cohen, 1991; Honderich, 1995; Root, 1993), meaning that until proven
otherwise, the knowledge embedded in the theory is taken to be true. As a
result, theories of this nature are never complete and require continual discourse between the theoretical framework of the theory and the theory in
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
269
270
Advances in Developing Human Resources
August 2002
use. Purposeful testing and refining of the theory are therefore the ongoing
responsibility and challenge of theory building in an applied field.
A Summative Overview of the
Theory-Building Method
The outcomes of this chapter conclude the presentation of Dubin’s twopart, eight-step quantitative method of theory building for applied disciplines.
The essential form and essence of this theory-building method are summarized by way of an overview presented in Table 2. This table presents a summarizing overview of the relationship between the two parts, eight corresponding steps, and the essential research questions of applied theory
building encapsulated in Dubin’s quantitative theory-building method.
Closing Reflections
Clearly, the task of theory building in an applied field is a series of
repeated and ongoing conversations between research and practice,
between concept development and concept verification, through research in
the real world. As is the experience of this author, the singular completion of
this two-part theory-research cycle and eight-step theory-building method
marks the beginning and not the culmination of the lengthy business and
challenge of building theory in applied disciplines.
Dubin (1978) offered a thoughtful concluding insight to those who brave
this challenge of theory building. He cautioned that the task of building theory is both a heartening and a disheartening one. It is heartening, he suggested, in that when well done, theory building should result in a clear and
simple explanation of how a phenomenon works in the real world. On the
other hand, he explained that when well done, the outcomes of theory building should result in a clear and simple explanation of how a phenomenon
works in the real world! I remember well the experience of this paradoxical
caution when developing the theory of responsible leadership for performance. I also learned from this caution and experience that it is the task of
good applied theories to make what is often tacit or implicit knowledge
explicit. So, when you first unveil the theoretical framework of your newly
developed theory and your practitioner-colleague responds with, “Well,
what’s new about this? I knew this all along,” take that comment as the promise of one of the greatest compliments of the potential usefulness and
worthiness of your theory. Do not be disheartened by it, for it is usually a
good indicator that the explanation of your theory just helped make your
practitioner-colleague’s implicit knowledge of the phenomenon concerned
explicit for the first time, and that is an essential outcome of theory in
applied disciplines.
(text continues on p. 275)
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
TABLE 2:
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
The Relationship Between the Two Parts,Eight Steps,and Theory-Building Research Questions in Dubin’s Quantitative Theory-Building Method
A: The Two Parts
in Dubin’s TheoryBuilding Method
B: The Eight Steps
in Dubin’s TheoryBuilding Method
Part 1 Description:
The theory development
side of the theory-research
cycle
Purpose: To develop
an informed conceptual
framework of the theory
Output: An informed
theoretical or conceptual
framework of the theory
1. Identifying the units
of the theory
2. Establishing the laws
of interaction that
govern the theory
C: An Operational Description
of the Theory-Building Steps
D: Corresponding
Theory-Building
Research Questions
Addressed
• Also known as the concepts of the theory. Involves the
translation of the concepts to units, that is, to things or
variables whose interactions constitute the subject matter
of attention. A theoretical framework is used to specify the
manner in which these units interact with each other.
• Addresses Part 1, Step 1 of Dubin’s theory-building method
by developing responses to the first theory development
question, What are the units of the theory?
• Compares the outcomes of the first theory development step
to five quality criteria for Step 1: rigor and exactness, parsimony, completeness, logical consistency, and degree of conformity to the combination limitations of the units.
• Specifies the interactions among the concepts and units of the
theoretical framework.
• Addresses Part 1, Step 2 of Dubin’s theory-building method by
developing responses to the second theory development
question, What are the laws of interaction of the theory?
• Compares the outcomes of the second theory development step
to quality criterion of parsimony.
Attending to this
first part and first
four theory development steps enables the researchertheorist to answer
the theory-building
research question,
Can the theory be
developed?
271
(continued)
272
TABLE 2
Continued
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
A: The Two Parts
in Dubin’s TheoryBuilding Method
B: The Eight Steps
in Dubin’s TheoryBuilding Method
3. Determining and
clarifying the
boundaries of the
theory
4. Specifying the system
states of the theory
Part 2 Description: The
research operation side of
the theory-research cycle
5. Specification of the
propositions of the
theory
C: An Operational Description
of the Theory-Building Steps
D: Corresponding
Theory-Building
Research Questions
Addressed
• Theoretical frameworks are generally limited portions of the
world; therefore, one must explore and set forth the
boundaries within which the theory is expected to hold and
apply.
• Addresses Part 1, Step 3 of Dubin’s theory-building method by
developing responses to the third theory development question,
What are the boundaries of the theory?
• Compares the outcomes of the third theory development step
to two quality criteria: homogeneity and generalization.
• Specifies the conditions under which the theory is operative and
reveals the complexity of the complex portion of the real world
that the theoretical framework is presumed to represent; each
of the units interacts differently in the system states.
• Addresses Part 1, Step 4 of Dubin’s theory-building method by
developing responses to the fourth and final theory development
question, What are the system states of the theory?
• Compares the outcomes of the fourth and final theory development step to three quality criteria: inclusiveness, persistence,
and distinctiveness.
Attending to
• Once the four basic theory development steps resulting in a
theoretical framework are addressed, the researcher-theorist
this second part
is in a position to derive conclusions that represent logical and
and four research
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
Purpose: To build an
operationalized and
empirically verifiable and
trustworthy theory
Output: an empirically verified
and trustworthy theory
•
•
6. Identifying empirical
indicators of the
theory
•
•
•
7. Constructing the
hypotheses to test
the theory
•
•
273
true deductions about the theoretical framework in operation,
operation steps
or the propositions of the theoretical framework.
enables the
Addresses the first step in Part 2 and corresponding Step 5 of
researcher-theorist
Dubin’s theory-building method by developing responses to the
to answer the
following theory operationalization question: What are the
theory building
propositions of the theory?
questions, Can
Compares the outcomes of the first theory operationalization
the theory be
step to three quality criteria: consistency, accuracy, and parsimony. operationalized?
Enables the researcher-theorist to determine whether the
and Can the theory
theoretical framework of the theory does in fact represent
be verified?
the real world by converting each term in each proposition
Also enables the
whose test is sought into an empirical indicator, which in turn
researcher-theorist
secures measures of values on the units of the theory.
to answer the
Addresses Part 2, Step 6 of Dubin’s theory-building method by
ongoing theorydeveloping responses to the second theory operationalization
building question,
question, What are the empirical indicators of the theory?
Can the theory
Compares the outcomes of the second theory operationalbe continually
ization step to two quality criteria: specification of the operation refined and
of measurement and identification of the results or values
developed?
produced.
This next and seventh theory-building step involves substituting
the empirical indicators in the proposition statement with
testable hypotheses. The research operation then involves
measuring the values on the empirical indicators of the
hypotheses to determine if the theoretically predicted values
are achieved or approximated in the research test. It is here
where issues of validity arise in the theory-research cycle.
Addresses Part 2, Step 7 in Dubin’s theory-building method
by developing responses to the third theory operationalization
(continued)
274
TABLE 2
Continued
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
A: The Two Parts
in Dubin’s TheoryBuilding Method
B: The Eight Steps
in Dubin’s TheoryBuilding Method
8. Testing the theory
through a developed
plan of research
C: An Operational Description
of the Theory-Building Steps
D: Corresponding
Theory-Building
Research Questions
Addressed
question, What are the hypotheses of the theory?
• Compares the outcomes of this third theory operationalization
Step 2, quality criteria: homology between the hypotheses and
their corresponding propositions and identification of the
necessary and sufficient conditions of each unit of the theory.
• This is the eighth and last step in theory building and involves
testing the theory in practice, that is, against the real world.
Important in this step is that the empirical indicators employed
in testing the theory have a reasonable level of reliability. The
testing of an applied theory is an ongoing process and must
result in continuous refinement of the theory. The researchertheorist is faced with two theory-testing stances, or
combinations thereof, namely, to conduct research designed
to prove the adequacy of the theoretical framework or
research to improve the starting theoretical framework.
• Addresses Step 8 of Dubin’s theory-building method by
developing responses to the following theory verification and
ongoing refinement question: What is the critical research
needed to verify the theory?
• Compares the outcomes of this final and theory verification
step to the quality criteria for the scientific method employed to
test and verify the theory.
Lynham / DUBIN’S METHOD
The task of building theory in an applied field is a demanding and difficult one; in fact, it is a grandiose and ambitious challenge (Lynham, 2000b).
Its intent is to bring trustworthy knowledge and action to practice for purposes of improved understanding, action, and outcome. As is too often the
case in applied fields, witnessed by the oversupply of popular atheoretical
literature on the bookshelves of any large bookstore, this task and responsibility should not be forgone for the sake of bringing unsubstantiated knowledge to market before its time. In an applied field, such action is irresponsible of scholars who act with the full awareness that this knowledge will be
used by practitioners to solve human and organizational problems. The pressure to bring knowledge to market should not transcend scholarship, not in
applied fields like those of HRD and management, where the actions of one
can affect so many and have such huge repercussions on human and organizational performance.
References
Cohen, B. P. (1991). Developing sociological knowledge: Theory and method (2nd ed.).
Chicago: Nelson-Hall.
Collins, J. C., & Porras, J. I. (1994). Built to last: Successful habits of visionary companies. New York: HarperBusiness.
Dubin, R. (1976). Theory building in applied areas. In M. D. Dunnette (Ed.), Handbook
of industrial and organizational psychology (pp. 17-39). Chicago: Rand McNally.
Dubin, R. (1978). Theory building (Rev. ed.). New York: Free Press.
Gibson, J. L., Ivancevich, J. M., & Donnelly, J. H., Jr. (1994). Organizations: Business,
structure, processes (8th ed.). Boston: Irwin.
Hearn, G. (1958). Theory building in social work. Toronto, Canada: University of
Toronto Press.
Holton, E. F., III. (1999). Performance domains and their boundaries. Advances in
Developing Human Resources, 1, 26-46.
Honderich, T. (Ed.). (1995). The Oxford companion to philosophy. New York: Oxford
Press.
Hultgren, F. H., & Coomer, D. L. (Eds.). (1989). Alternative modes of inquiry in home
economics research. Peoria, IL: Glencoe.
Kaplan, A. (1964). The conduct of inquiry: Methodology for behavioral science. San
Francisco: Chandler.
Lynham, S. A. (1998). The development and evaluation of a model of responsible leadership for performance: Beginning the journey. Human Resource Development International, 1(2), 207-220.
Lynham, S. A. (2000a). The development of a theory of responsible leadership for performance (Tech. Rep.). St. Paul: University of Minnesota, Human Resource Development Research Center.
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013
275
276
Advances in Developing Human Resources
August 2002
Lynham, S. A. (2000b). Theory building in the human resource development profession.
Human Resource Development Quarterly, 11(2), 159-178.
Lynham, S. A. (2002). The general method of theory-building research in applied disciplines. Advances in Developing Human Resources, 4(3), 221-241.
Posner, B., & Kouzes, J. (1996). Ten lessons for leaders and leadership developers. Journal of Leadership Studies, 3(3), 3-10.
Reynolds, P. D. (1971). A primer in theory construction. New York: Macmillan.
Root, M. (1993). Philosophy of social science. Cambridge, MA: Blackwell.
Rummler, G. A., & Brache, A. P. (1995). Improving performance: How to manage the
white space on the organization chart (2nd ed.). San Francisco: Jossey-Bass.
Senge, P. M. (1990). The fifth discipline: The art and practice of the learning organization. New York: Currency Doubleday.
Swanson, R. A. (1994). Analysis for improving performance: Tools for diagnosing organizations and documenting workplace expertise. San Francisco: Berrett-Koehler.
Swanson, R. A. (1996). What I learned this year. St. Paul: University of Minnesota,
Human Resource Development Research Center.
Swanson, R. A., & Holton, E. F., III. (1999). Results: How to assess performance, learning, and perceptions in organizations. San Francisco: Berrett-Koehler.
Torraco, R. J. (1994). The development and validation of a theory of work analysis. St.
Paul: University of Minnesota, Human Resource Development Research Center.
Torraco, R. J. (1997). Theory-building research methods. In R. A. Swanson & E. F.
Holton III (Eds.), Human resource development handbook: Linking research and
practice (pp. 114-137). San Francisco: Berrett-Koehler.
Torraco, R. J. (Ed.). (1999). Performance improvement: Theory and practice. Advances
in Developing Human Resources, 1.
Torraco, R. J. (2000). A theory of knowledge management. Advances in Developing
Human Resources, 5, 38-62.
Von Bertalanffy, L. (1968). General systems theory: Foundations, development, application. New York: George Braziller.
White Newman, J. B. (1993). The three E’s of leadership: A model and metaphor for
effective, ethical and enduring leadership. Unpublished manuscript, College of St.
Catherine, St. Paul, MN.
Downloaded from adh.sagepub.com at UNIV NORTH TEXAS LIBRARY on June 28, 2013