July 28, 2021 Comment off
Middle Range Or Interdisciplinary Theory Evaluation
73
This chapter focuses on the theory (T) component of conceptual-theoretical-empirical (C-T-E) structures for research.
Chapter 6
Evaluation of Middle-Range Theories
KEYWORDS
Axiom
Deductive Reasoning
Explicit Middle-Range Theory
Hypothesis
Implicit Middle-Range Theory
Inductive Reasoning
Internal Consistency
Middle-Range Theory
Parsimony
Postulate
Premise
Reasoning
Semantic Clarity
Semantic Consistency
Significance
Social Significance
Structural Consistency
Testability
Theorem
Theoretical Significance
Recall from Chapter 2 that the T component of a C-T-E structure is the middle-range theory that was generated or tested by research. In that chapter, we defined a theory as a set of relatively concrete and specific concepts and propositions that are derived from the concepts and propositions of a conceptual model. We also pointed out that a middle-range theory guides research by providing the focus for the specific aims for the research. In Chapter 3 you began to learn where to look for information about the middle-range theory in research reports (Box 6–1) and what information you could expect to find (Box 6–2).
BOX 6-1
Evaluation of Middle-Range Theories: Where Is the Information?
Content about the middle-range theory may be found in every section of the research report.
BOX 6-2
Evaluation of Middle-Range Theories: What Is the Information?
The name of the middle-range theory
1489_Ch06_073-088.qxd 7/8/08 2:25 PM Page 73
74 Part 2 ■ Evaluation of Conceptual Models and Theories
In Chapter 4, you began to learn how to determine how good the available information about the theory is. More specifically, in Chapter 4 we presented a framework for evaluation of the different components of C-T-E structures for theory-generating research and theory- testing research.
In this chapter, you will learn more about what middle-range theories are and how to evaluate them to determine how good the information about the T component for theory-generating research and theory-testing research is. After explaining how to identify a middle-range theory, we discuss in detail the four criteria in the framework identified in Chapter 4 for evaluating the T component of C-T-E structures—significance, internal consistency, parsimony, and testability— and provide examples that should help you better understand how to apply the criteria as you read research reports. Application of the criteria will facilitate your evaluation of how good the informa- tion about the middle-range theory provided in the research report is.
HOW IS THE MIDDLE-RANGE THEORY IDENTIFIED?
We believe that generating or testing a middle-range theory is the main reason for research. Consequently, a vast number of middle-range theories exist. Sometimes, the name of the middle-range theory is stated explicitly in the research report, but sometimes the middle-range theory is not stated explicitly and is only implied.
Explicit Middle-Range Theories
A review of research guided by seven different nursing conceptual models yielded more than 50 explicitly named middle-range theories that were directly derived from the conceptual models. The theories and the conceptual models from which they were derived are listed in Table 6–1 on the CD that comes with this book.
The conceptual frames of reference for three other explicit middle-range nursing theories were extracted from publications about the theories (Fawcett, 2005b). Although none of the theories were derived from a nursing conceptual model, statements reflecting some of the nursing metaparadigm concepts—human beings, environment, health, and nursing—were identified. The theories and relevant citations are:
1. Orlando’s Theory of the Deliberative Nursing Process (Orlando, 1961; Schmieding, 2006)
2. Peplau’s Theory of Interpersonal Relations (Peplau, 1952, 1997) 3. Watson’s Theory of Human Caring (Watson, 1985, 2006)
The conceptual origins of many other explicitly named middle-range nursing theories are not yet clear; examples are listed in Table 6–2, which is included on the CD that comes with this book. Some explicitly named middle-range theories that are tested by nurse researchers come from other disciplines; examples are given in Table 6–3 on the CD. Additional information about the theories listed in Tables 6–2 and 6–3 can be found in Marriner Tomey and Alligood (2006), Peterson and Bredow (2004), Smith and Liehr (2003), and/or Ziegler (2005).
1489_Ch06_073-088.qxd 7/8/08 2:25 PM Page 74
Although the conceptual frame of reference for the theories listed in Table 6–3 typically is not mentioned in the published research report, one such theory—the Theory of Planned Behavior (TPB)—was linked with Neuman’s Systems Model and with Orem’s Self-Care Framework by Villarruel and her colleagues (2001). They explained that the linkage placed the TPB within a nursing context and provided direction for a program of nursing research that could progress from “an explanation of the antecedents of behavioral actions to a prediction of the effects of nursing interventions on behavioral actions that are directed toward health promotion and dis- ease prevention” (p. 160). They also explained that linkage of the TPB to a nursing conceptual model is needed if effects of interventions are to be studied, because interventions are not part of the TPB.
Implicit Middle-Range Theories
When the middle-range theory is implicit—that is, when it is not explicitly named—you may want to make up a name to increase your understanding of the theory. Finding the information
Chapter 6 ■ Evaluation of Middle-Range Theories 75
BOX 6-3
Naming an Implicit Middle-Range Theory
Example from a Theory-Generating Research Report
Study purpose
The purpose of this descriptive study was to identify patients’ perceptions of fatigue during chemotherapy for Hodgkin’s disease.
Results
Analysis of the patients’ responses to an open-ended questionnaire revealed three categories of fatigue—exhausted, sleepy, and tired.
Possible names for the middle-range theory
• Perceptions of Fatigue Theory
• Theory of Categories of Chemotherapy Fatigue
Example from a Theory-Testing Research Report
Study purpose
The purpose of this experimental study was to determine the effect of exercise on chemotherapy- related fatigue.
Hypothesis
An increase in exercise will decrease chemotherapy-related fatigue.
Possible names for the middle-range theory
• Theory of the Effects of Exercise on Fatigue
• Exercise and Fatigue Theory
1489_Ch06_073-088.qxd 7/8/08 2:25 PM Page 75
76 Part 2 ■ Evaluation of Conceptual Models and Theories
in a research report that may be used to identify a name for the theory can be challenging or even frustrating. Sometimes, the statement of the study purpose or aims can be used as the basis for the name of the theory. Or, you may have to rely on the categories or themes and their defini- tions in reports of theory-generating research and the study variables, definitions of variables, and hypotheses in reports of theory-testing research. Examples from fictitious studies are given in Box 6–3. (Recall that we discussed categories, themes, variables, definitions, and hypotheses in Chapter 2, and we identified where to look for the content of the T compo- nent in Chapters 3 and 4.)
HOW IS THE CRITERION OF SIGNIFICANCE OF A MIDDLE-RANGE THEORY APPLIED?
The criterion of significance of a middle-range theory draws attention to the importance of the theory to society and to the advancement of knowledge within a discipline. We call the importance of the theory to society its social significance, and the importance of the theory to advancement of knowledge its theoretical significance.
Application of the criterion of significance helps you determine whether enough informa- tion about social significance and theoretical significance is given in the research report. Enough information means that you can understand just how important the theory is to society and how the theory has filled a gap in or extended existing knowledge. The same amount of infor- mation about social significance and theoretical significance should be included in reports of both theory-generating research and theory-testing research.
The criterion of significance is met when you can answer yes to two questions:
• Is the middle-range theory socially significant? • Is the middle-range theory theoretically significant?
Is the Middle-Range Theory Socially Significant?
The criterion of significance requires the middle-range theory to be socially significant. That means the theory is about people experiencing a health condition that currently is regarded as having some practical importance by the general public and members of one or more dis- ciplines. The social significance of a middle-range theory is obvious when the theory focuses on a health condition, such as cancer, heart disease, or diabetes, that is experienced by a rel- atively large number of people. Social significance is also obvious when the theory focuses on a health condition that is experienced by a relatively small number of people but has a large impact on the quality of people’s lives, such as spinal cord injury or mental illness. In other words, social significance is concerned with whether the health condition experienced by people is regarded as having a considerable actual or potential impact on desired lifestyle. The social significance of a middle-range theory typically is explained in a few sentences about the incidence of a particular health condition (Cowen, 2005). An example of social significance from Newman’s (2005) study of correlates of functional status of caregivers of children in body casts is given in Box 6–4.
1489_Ch06_073-088.qxd 7/8/08 2:25 PM Page 76
Is the Middle-Range Theory Theoretically Significant?
The criterion of significance also requires the middle-range theory to be theoretically signif- icant. That means the theory offers new insights into the experiences of people who have a certain health condition. The theoretical significance of a nursing theory typically is explained in a concise summary of “what is known, what is not known, and how the results from [the research] advance . . . knowledge” (Cowen, 2005, p. 298). In other words, the information given in the research report about theoretical significance should tell you that the research focuses on the next meaningful step in the development of a theory about people with a certain health condition. Sometimes, a researcher will write that the research was conducted because nothing was known about the research topic. Such a statement does not meet the criterion of significance because it is possible that the topic is trivial and, there- fore, the research is trivial. An example of an explanation of theoretical significance from Newman’s study of correlates of functional status of caregivers of children in body casts is given in Box 6–5.
Chapter 6 ■ Evaluation of Middle-Range Theories 77
BOX 6-4
Example of Statement of Social Significance
The number of children who are placed in body casts each year is unknown. Observations in orthopedic clinics, however, indicate that a relatively small number of children are so treated. Mothers, fathers, and others who care for children in body casts face challenges that disrupt their usual pattern of daily living (Newman, 2005, p. 416).
(In this example, although a large number of children do not have a health condition requir- ing a body cast, their caregivers face considerable challenges.)
BOX 6-5
Example of Statement of Theoretical Significance
Developmental needs of the child, specific care requirements related to the body cast, and changes in parental functional status, health, psychological feelings, and family needs comprise typical challenges that must be faced by caregivers (Newman, 1997b; Newman & Fawcett, 1995). Previous studies of functional status during normal life transitions and serious illness have revealed that alterations in performance of usual role activities are influenced by demographic, health, psychological, and family variables (Tulman & Fawcett, 1996, 2003). This pilot study extended the investigation of correlates of functional status by examining the relation of personal health and self-esteem to functional status of caregivers of children in body casts [from] birth up to 3 years of age and [from] 3 to 12 years of age. The pilot study also provided data to determine the feasibility of a large-scale study. The long-term goal of the research is to assist caregivers to attain optimal functional status while caring for children in body casts (Newman, 2005, p. 416).
(In this example, the first two sentences tell you what is already known and include citations to previous research. The remaining three sentences tell you how the study extends knowledge, why it was conducted, and the long-term goal of the research.)
1489_Ch06_073-088.qxd 7/8/08 2:25 PM Page 77
78 Part 2 ■ Evaluation of Conceptual Models and Theories
BOX 6-6
Example of Semantic Clarity of a Middle-Range Theory Concept
• Concept: Self-esteem
• Constitutive definition: Self-esteem “is defined as the caregiver’s feelings of personal worth and value” (Newman, 2005, p. 417).
• Operational definition: Self-esteem was measured by Rosenberg’s Self-Esteem Scale (Newman, 2005).
HOW IS THE CRITERION OF INTERNAL CONSISTENCY OF A MIDDLE-RANGE THEORY APPLIED?
Internal consistency draws attention to the comprehensibility of the middle-range theory. Application of the criterion of internal consistency helps determine whether enough information about the theory concepts and propositions is given in the research report. Enough informa- tion means that you can identify each concept and how the concepts are described and linked. The same amount of information about internal consistency should be included in reports of both theory-generating research and theory-testing research.
The criterion of internal consistency is met when you can answer yes to three questions:
• Is each concept of the middle-range theory explicitly identified and clearly defined? • Are the same term and same definition used consistently for each concept? • Are the propositions of the middle-range theory reasonable?
Is Each Middle-Range Theory Concept Explicitly Identified and Clearly Defined?
The criterion of internal consistency requires every concept of the theory to be explicitly iden- tified and clearly defined. This requirement, which is called semantic clarity (Chinn & Kramer, 2004; Fawcett, 1999), is met when each concept can be identified and both theoret- ical and operational definitions for each concept are included in the research report. An exam- ple from Newman’s (2005) study of correlates of functional status of caregivers of children in body casts is given in Box 6–6. (Recall from Chapter 2 that a constitutive definition provides meaning for a concept, and an operational definition indicates how the concept was measured.)
Semantic clarity requires that even concepts that are generally understood in everyday language must be clearly defined when used in theories. As Chinn and Kramer (2004) pointed out,
Words like stress and coping have general common language meanings, and they also have specific theoretic meanings. . . . If words with multiple meanings are used in the- ory and not defined, a person’s everyday meaning of the term, rather than what is meant in the theory, often is assumed; therefore, clarity is lost. (p. 110)
1489_Ch06_073-088.qxd 7/8/08 2:25 PM Page 78
Are the Same Term and Same Definition Used Consistently for Each Middle-Range Theory Concept?
Semantic clarity is enhanced when the same term and same constitutive definition are used for each concept throughout the research report. The requirement for use of the same term and same constitutive definition is called semantic consistency (Chinn & Kramer, 2004; Fawcett, 1999). Although requiring use of the same term for the same concept may seem obvious, some- times a researcher uses different labels for the same concept. For example, a researcher may reduce clarity by referring to both self-esteem and self-confidence in the same research report, although the theory focuses only on self-esteem. Chinn and Kramer (2004) explained,
Normally, varying words to represent similar meanings is a writing skill that can be used to avoid overuse of a single term. But, in theory, if several similar concepts are used interchangeably when one would suffice, . . . the clarity of the [concept] is reduced rather than improved. (p. 110)
A researcher also may reduce clarity by using different constitutive definitions for the same concept. For example, if self-esteem is defined as “feelings of personal worth and value,” that concept should not also be defined as “feelings of self-confidence” in the same research report. Different definitions of the same concept that are explicit are, as Chinn and Kramer (2004) noted, “fairly easy to uncover” (p. 111). In contrast, when a different definition is not explicit but only implied, the inconsistency may be more difficult to identify. Suppose, for example, that a researcher explicitly defined self-esteem as “feelings of personal worth and value” and then wrote about caregivers’ feeling self-confident when bathing a child in a body cast. It would be difficult to know whether the researcher was referring to the caregivers’ self- esteem or another concept when discussing feelings of self-confidence.
Sometimes a researcher may use more than one operational definition for the same concept. If all of the operational definitions identify instruments that measure the same constitutive def- inition of the concept, the requirement of semantic consistency is met. For example, using the constitutive definition given in Box 6–6, a researcher might operationally define self-esteem as measured by both Rosenberg’s Self-Esteem Scale and a Personal Worth and Value Questionnaire that asks caregivers to rate their feelings of personal worth and value on a scale of 1 to 10, with 1 equivalent to feelings of very low personal worth and value and 10 equiva- lent to feelings of very high personal worth and value.
However, if the instruments identified in the operational definitions measure different constitutive definitions of the concept, the requirement of semantic consistency is not met. For example, again using the constitutive definition of self-esteem given in Box 6–6, a researcher might operationally define self-esteem as measured by the Personal Worth and Value Questionnaire, as well as a Self-Confidence Inventory, which measures self-esteem con- stitutively defined as “feelings of self-confidence.”
Are the Middle-Range Theory Propositions Reasonable?
The criterion of internal consistency also requires the propositions of the theory to be reasonable. This requirement is called structural consistency (Chinn & Kramer, 2004;
Chapter 6 ■ Evaluation of Middle-Range Theories 79
1489_Ch06_073-088.qxd 7/8/08 2:25 PM Page 79
Fawcett, 1999). Propositions are reasonable when they follow the rules of inductive or deduc- tive reasoning. Reasoning is defined as “the processing and organizing of ideas in order to reach conclusions” (Burns & Grove, 2007, p. 16).
Inductive Reasoning
Inductive reasoning encompasses a set of particular observations and a general conclu- sion. This type of reasoning is “a process of starting with details of experience and moving to a general picture. Inductive reasoning involves the observation of a particular set of instances that belong to and can be identified as part of a larger set” (Liehr & Smith, 2006, p. 114). Inductive reasoning is most often found in reports of theory-generating research. Observations typically are quotations from study participants or are made by the researcher; the conclusion usually is referred to as a category or theme. The general form of inductive reasoning and an example from a fictitious study are given in Box 6–7.
Flaws in Inductive Reasoning
Flaws in inductive reasoning occur when a relevant observation is excluded (Kerlinger & Lee, 2000). For example, suppose that a researcher observed many white swans and con- cluded that all swans are white. The flaw would be discovered when another observation revealed a black swan. Or, suppose that a nurse observed that several people with a medical diagnosis of depression cried a lot and concluded that all people who cry are depressed. The flaw would be discovered when another observation revealed that people who were happy also cried. Consequently, when you evaluate the structural consis- tency of a middle-range theory in a theory-generating research report, consider whether the report includes a sufficient number and variety of observations to support each conclusion.
80 Part 2 ■ Evaluation of Conceptual Models and Theories
BOX 6-7
Inductive Reasoning
General form: Proceeds from the particular to the general Observation: A is an instance of x. Observation: B is an instance of x. Observation: C is an instance of x. Conclusion: A, B, and C make up x.
Example Observation: Doing household chores is a usual activity that is performed less frequently when a person is ill. Observation: Visiting friends is a usual activity that is performed less frequently when a person is ill. Observation: Exercising is a usual activity that is performed less frequently when a person is ill. Conclusion: All usual activities are performed less frequently when a person is ill.
1489_Ch06_073-088.qxd 7/8/08 2:25 PM Page 80
Chapter 6 ■ Evaluation of Middle-Range Theories 81
BOX 6-8
Deductive Reasoning
General form: Proceeds from the general to the particular Premise: If x is related to y, and Premise: if y is related to z, Hypothesis: then x is related to z.
Example Premise: If personal health status is related to self-esteem, and Premise: if self-esteem is related to functional status, Hypothesis: then personal health status is related to functional status.
Example constructed from Newman (2005).
Deductive Reasoning
Deductive reasoning encompasses a set of general propositions and a particular conclusion. This type of reasoning is “a process of starting with the general picture . . . and moving to a specific direction” (Liehr & Smith, 2006, p. 114). The general propositions of deductive rea- soning typically are referred to as premises, axioms, or postulates; the particular conclusion is called a theorem or hypothesis. Premises, axioms, and postulates typically are drawn from literature reviews of previous research and are regarded as empirically adequate statements that do not have to be empirically tested again. A theorem or hypothesis, in contrast, must be test- ed by research. Deductive reasoning is most often found in reports of theory-testing research. The general form of deductive reasoning and an example constructed from Newman’s (2005) study of correlates of functional status of caregivers of children in body casts are given in Box 6–8.
Flaws in Deductive Reasoning
Flaws in deductive reasoning occur when there is an error in a general proposition. Suppose, for example, that a researcher started with the premise that personal health status was related to functional status without providing any supporting research findings, added a premise that functional status was related to self-esteem, and then hypothesized that personal health status was related to self-esteem. The deduction in this example is flawed because the initial premise (personal health status is related to functional status) cannot be regarded as empirically ade- quate prior to testing the statement by conducting research. Although sets of deductive rea- soning statements such as those seen in Box 6–8 are not usually found in research reports, the researcher should provide sufficient support for each hypothesis by citing relevant previous research as part of a critical review of the theoretical and empirical literature. Consequently, when you evaluate the structural consistency of a middle-range theory in a theory-testing research report, consider whether the report includes sufficient information to support any premises and each hypothesis.
1489_Ch06_073-088.qxd 7/8/08 2:25 PM Page 81
HOW IS THE CRITERION OF PARSIMONY OF THE MIDDLE-RANGE THEORY APPLIED?
Parsimony draws attention to the number of concepts and propositions that make up a middle-range theory. Application of the criterion of parsimony helps you determine whether the middle-range theory is stated as concisely as possible. The same standard of simplicity should be used to evaluate theories that were generated or tested.
The criterion of parsimony is met when you can answer yes to one question:
• Is the middle-range theory stated concisely?
Is the Middle-Range Theory Stated Concisely?
Parsimony requires that a middle-range theory be made up of as few concepts and propositions as necessary to clearly convey the meaning of the theory. Glanz (2002) referred to parsimony as “selective inclusion” of concepts (p. 546). Walker and Avant (2005) explained, “A parsimonious theory is one that is elegant in its simplicity even though it may be broad in its content” (p. 171).
The criterion of parsimony should not be confused with oversimplification of the content needed to convey the meaning of the theory. A theory should not be stated so simply that its meaning is lost. “Parsimony that does not capture the essential features of the [theory] is false economy” (Fawcett, 1999, p. 93). In other words, “A parsimonious theory explains a complex [thing] simply and briefly without sacrificing the theory’s content, structure, or completeness” (Walker & Avant, 2005, p. 172).
A challenge in theory-generating research is to include all relevant data that were collected in one or just a few meaningful categories, rather than a large number of categories, subcategories, and sub-subcategories. For example, a researcher who regards household chores, visiting friends, and exercising as usual activities will present a much more parsimonious theory than a researcher who regards each of those activities as a separate category.
A challenge in theory-testing research is to determine whether the middle-range theory becomes more parsimonious as the result of testing. For example, Tulman and Fawcett (2003) found that several concepts and propositions of their Theory of Adaptation During Childbearing were not supported by their research. They concluded, “The collective quantitative results of our study revealed a somewhat more parsimonious version of the theory” (p. 151). Sometimes, a research report will include diagrams depicting the connections between the middle-range theory concepts before and after testing. Such diagrams can be helpful visual aids to evaluation of parsimony. Figure 6–1 depicts an example from a correlational study of the relations between type of cesarean birth and perception of the birth experience, perception of the birth experience and responses to cesarean birth, and type of childbirth and responses to cesarean birth (Fawcett et al., 2005).
As can be seen in the diagram, the middle-range theory before testing includes links between type of cesarean birth (unplanned and planned) and perception of the birth experi- ence, perception of the birth experience and responses to cesarean birth, and type of cesarean birth and responses to cesarean birth (see Figure 6–1 part A). After testing, the theory includes
82 Part 2 ■ Evaluation of Conceptual Models and Theories
1489_Ch06_073-088.qxd 7/8/08 2:25 PM Page 82
Chapter 6 ■ Evaluation of Middle-Range Theories 83
Perception of the Birth Experience
Responses to Cesarean Birth
Type of Cesarean Birth
A. The Theory Before Testing
Perception of the Birth Experience
Responses to Cesarean Birth
Type of Cesarean Birth
B. A More Parsimonious Theory After Testing
Figure 6-1. Diagrams of middle-range theory propositions before and after testing.
links only between perception of the birth experience and responses to cesarean birth, and type of cesarean birth and responses to cesarean birth (see Figure 6–1 part B). After testing, the the- ory is more parsimonious because no support was found for a link between type of cesarean birth and perception of the birth experience.
HOW IS THE CRITERION OF TESTABILITY OF THE MIDDLE-RANGE THEORY APPLIED?
Testability draws attention to whether the middle-range theory can be empirically tested. Application of the criterion of testability helps you determine whether enough information about the measurement of theory concepts is given in the research report. Enough informa- tion means that you can identify how each concept was operationally defined and how any associations between concepts were determined. The same amount of information about testa- bility should be included in reports of both theory-generating research and theory-testing research.
The criterion of testability is met when you can answer yes to two questions:
• Was each concept measured? • Were all assertions tested through some data analysis technique?
Was Each Concept Measured?
The criterion of testability requires each middle-range theory concept to be empirically observ- able—that is, measurable. The operational definition of the concept identifies the way in which it was measured. A diagram of the C-T-E structure for the research will help you to answer this question. If the research report does not include a C-T-E structure diagram, you can try to
1489_Ch06_073-088.qxd 7/8/08 2:25 PM Page 83
84 Part 2 ■ Evaluation of Conceptual Models and Theories
BOX 6-9
Applying the Criterion of Testability for Theory- Generating Research
Conceptual Model
Roy’s Adaptation Model
Conceptual Model Concept
Role function mode
Proposition Linking the Conceptual Model Concept to the Empirical Indicator
Development of the Usual Activities Interview Schedule was guided by the role function mode of adaptation.
Operational Definition
Content analysis of data from the Usual Activities Interview Schedule revealed one category, which was labeled “usual activities of ill people.”
Middle-Range Theory Concept
Usual activities of ill people
Descriptive Research Design 30 People Who Had Self-Reported Illness
Usual Activities Interview Schedule Content Analysis
Usual Activities of Ill People
Roy’s Adaptation Model
Role Function Mode
Conceptual Model
Middle-Range Theory
Empirical Research Methods
C-T-E structure for a theory-generating study.
draw one from the written information included in the report. The diagram will enable you to determine whether each concept is connected to an instrument or experimental conditions.
The example in Box 6–9 contains information from a fictitious theory-generating research report. The written information and the C-T-E diagram indicate that the criterion of testability was met. Suppose, however, that another category was mentioned in the report, such as special activities of ill people, and that no information about how the data used to generate the special activities category was given. In that instance, the diagram would not be complete and the cri- terion of testability would not have been met.
1489_Ch06_073-088.qxd 7/8/08 2:25 PM Page 84
Chapter 6 ■ Evaluation of Middle-Range Theories 85
BOX 6-10
Applying the Criterion of Testability for Theory- Testing Research
Conceptual Model
Roy’s Adaptation Model
Conceptual Model Concepts
• Physiological mode
• Self-concept mode
• Role function mode
Propositions Linking the Conceptual Model and Middle-Range Theory Concepts
• The physiological mode was represented by personal health.
• The self-concept mode was represented by self-esteem.
• The role function mode was represented by functional status.
Middle-Range Theory Concepts
• Personal health
• Self-esteem
• Functional status
Operational Definitions
Personal health was measured by the Personal Health Questionnaire (PHQ). Self-esteem was measured by Rosenberg’s Self-Esteem Scale (RSES). Functional status was measured by the Inventory of Functional Status–Caregiver of a Child in a Body Cast (IFSCCBC).
PHQ
Physiological Mode
Personal Health
Self-Concept Mode
Self-Esteem
RSES
Role Function Mode
Functional Status
IFSCCBC
Roy’s Adaptation Model
Correlational Research Design Correlational Statistics 30 Parents of Children in Body Casts
Conceptual Model
Middle-Range Theory
Empirical Research Methods
C-T-E structure for a theory-testing study.
Example constructed from Newman (2005).
1489_Ch06_073-088.qxd 7/8/08 2:25 PM Page 85
The example in Box 6–10 contains information found in Newman’s (2005) theory-testing research report. Although Newman did not include a C-T-E structure diagram, it was easily constructed from the written information in the conceptual framework and instruments subsections of the report. The written information and diagram reveal that the criterion of testability was met. Suppose, however, that Newman had not included an operational defini- tion for one of the concepts. In that instance, the diagram would not be complete and the cri- terion of testability would not have been met.
Were All Assertions Tested Through Some Data Analysis Technique?
The criterion of testability requires each assertion made by the middle-range theory propo- sitions to be measurable through some data analysis technique. Although most theory- generating research focuses on the description of a health-related experience in the form of one or a few concepts that are not connected to one another, some theory-generating research reports include propositions that state an association between two concepts. Suppose, for example, that a researcher generated a theory of usual activities of ill people from data col- lected from a group of chronically ill people and a group of acutely ill people. Suppose also that the researcher looked at the list of usual activities for each group, concluded that acute- ly ill people performed different usual activities than chronically ill people, and included a proposition stating that there is an association between the type of illness and the type of usual activities performed. In this example, a proposition stating an association between two concepts—type of illness and usual activities—was generated simply through visual inspec- tion of the data.
Theory-testing research, in contrast, frequently involves use of statistical procedures to sys- tematically test associations between two or more concepts. In theory-testing research, propo- sitions stating associations between concepts, especially when the names of the instruments used to measure the concepts (i.e., the empirical indicators) are substituted for the names of the concepts, are referred to as hypotheses. Each hypothesized association between concepts is tested using a statistical procedure to determine if there is an association between scores from the instruments used to measure the concepts.
The example in Box 6–11 gives the information you should look for in the research report to determine whether the proposition was testable.
Hypothesis Testing
Theory-testing research involves tests of hypotheses. Sometimes, the hypothesis is explicit, and sometimes it is implicit. Explicit hypotheses are, of course, easy to identify because they are labeled as such. For example, a researcher may state that the purpose of the study was to test a particular hypothesis, or a few hypotheses will be listed in the research report. You can iden- tify any implicit hypotheses by systematically examining the research findings and listing all the statistical procedures mentioned in the report. For example, examination of Newman’s
86 Part 2 ■ Evaluation of Conceptual Models and Theories
1489_Ch06_073-088.qxd 7/8/08 2:25 PM Page 86
(2005) research report revealed that she used a correlation coefficient to test the implicit hypothesis of a relation between scores on the Personal Health Questionnaire (PHQ) and scores on the Inventory of Functional Status–Caregiver of a Child in a Body Cast (IFSCCBC).
Hypotheses should be falsifiable (Popper, 1965; Schumacher & Gortner, 1992). That means that the way in which the hypothesis is stated should allow the researcher to conclude that the hypothesis was rejected if the data do not support the assertion made in the hypoth- esis. For example, suppose that a researcher hypothesized that all mothers and fathers have high, medium, or low scores on the IFSCCBC and high, medium, or low scores on the PHQ. The hypothesis cannot be falsified because it does not eliminate any logically or practically possible results. In contrast, the hypothesis that all mothers and fathers have medium scores on the PHQ and low scores on the IFSCCBC can be falsified because it asserts that the moth- ers and fathers will not have high or low scores on the PHQ and will not have high or medium scores on the IFSCCBC.
In addition, it is not correct to conclude that a hypothesis was partially supported. For example, suppose that a researcher hypothesized that both mothers’ and fathers’ scores on the PHQ were related to their scores on the IFSCCBC and that the results indicated that the hypothesis was supported only by the data from the mothers. It would not be correct to con- clude that the hypothesis was partially supported because the mothers’ data supported the hypothesis. Rather, the correct conclusion is that the hypothesis is rejected.
Chapter 6 ■ Evaluation of Middle-Range Theories 87
BOX 6-11
Example of Testability of a Proposition Stating an Association Between Two Concepts
Middle-Range Theory Concepts
• Personal health
• Functional status
Proposition
There is a relation between personal health and functional status.
Operational Definitions
• Personal health was measured by the Personal Health Questionnaire (PHQ).
• Functional status was measured by the Inventory of Functional Status–Caregiver of a Child in a Body Cast (IFSCCBC).
Hypothesis
There is a relation between scores on the PHQ and scores on the IFSCCBC.
Statistical Procedure
A Pearson coefficient of correlation was used to determine the correlation between scores from the PHQ and the IFSCCBC.
Example constructed from Newman (2005).
1489_Ch06_073-088.qxd 7/8/08 2:25 PM Page 87
Conclusion
In this chapter, you continued to learn about how to determine how good the information about a middle-range theory given in a research report is. Specifically, you learned how to evaluate the T component of C-T-E structures using the criteria of significance, internal consistency, parsi- mony, and testability. The questions to ask and answer as you evaluate the middle-range theory are listed in Box 6–12. Application of these four criteria should help you to better understand the link between the T and E components of C-T-E structures. The learning activities for this chapter will help you increase your understanding of the four criteria and their application to the contents of research reports.
References
Full citations for all references cited in this chapter are provided in the Reference section at the end of the book.
Learning Activities
Activities to supplement what you have learned in this chapter, along with practice examina- tion questions, are provided on the CD that comes with this book.
88 Part 2 ■ Evaluation of Conceptual Models and Theories
BOX 6-12
Evaluation of Middle-Range Theories: How Good Is the Information?
Significance
• Is the middle-range theory socially significant?
• Is the middle-range theory theoretically significant?
Internal Consistency
• Is each concept of the middle-range theory explicitly identified and clearly defined?
• Are the same term and same definition used consistently for each concept?
• Are the propositions of the middle-range theory reasonable?
Parsimony
• Is the middle-range theory stated concisely?
Testability
• Was each concept measured?
• Were all assertions tested through some data analysis technique?
1489_Ch06_073-088.qxd 7/8/08 2:25 PM Page 88
