6
Thoughts about Conceptual Models and Measurement Validity
I have long advocated the development and articulation of explicit conceptual-theoretical-empirical (CTE) structures for nursing research and nursing practice (Fawcett, 1999; Fawcett & DeSanto-Madeya, 2013). The conceptual (C) component of a CTE structure is the conceptual model that is selected to guide research or practice. The theoretical (T) component is the concepts and propositions of the middle-range theory that is generated or tested through research or applied in practice. These concepts and propositions are directly derived from or linked with the concepts and propositions of the conceptual model.
Direct derivation occurs when a new theory is proposed. For example, Tulman and Fawcett (2003) derived the concepts and propositions of their new theory of changes in health that women experience during pregnancy and the postpartum from two concepts of Roy’s adaptation model (RAM)—environmental stimuli and modes of adaptation—and the RAM proposition that environmental stimuli affect responses in the modes of adaptation.
Linkage occurs when a theory from an adjunctive discipline is logically connected to the conceptual model. For example, Villarruel, Bishop, Simpson, Jemmott, and Fawcett (2001) explained the logical connections of the concepts of the social psychological theory of planned behavior with several concepts of Neuman’s systems model—intra-, inter-, and extrapersonal stressors arising from the external and created environments, as well as the flexible and normal lines of defense--for a study of condom use. They also explained the logical congruence of the same theory with three concepts of Orem’s self-care framework—basic conditioning factors, self-care agency, and self-care—for a different study of condom use.
The empirical (E) component of CTE structures encompasses the research instruments or practice tools that are selected to measure the middle-range theory concepts. Researchers spend a great deal of time searching for instruments, including interview guides and questionnaires, that are appropriate measures of the middle-range theory concepts they plan to study. Tulman and Fawcett (2003), for example, found some existing instruments and developed new instruments to measure the concepts of their theory, as did Villarruel et al. (2001). Nursing practitioners, that is, nurses who are actively engaged in practice (Orem, 2001), spend a great deal of time searching for practice tools, including assessment formats and intervention protocols, that are appropriate measures of the middle-range theories they plan to apply when caring for people who seek nursing services. Sometimes, research instruments can be used as practice tools. For example, Fawcett and colleagues have used a combination of existing and new research instruments as practice tools in their integrated research and teaching projects focused on testing and applying theories of women’s experiences of childbirth (Aber, Weiss, & Fawcett, in press; Fawcett, Aber, & Weiss, 2003; Fawcett et al., 2011; Weiss, Fawcett, & Aber, 2009).
A major aspect of every search for research instruments and practice tools is determining measurement validity, which refers to the appropriateness of the instruments and tools, which in turn refers to what middle-range theory concept the research instrument or practice tool “really” measures (Fawcett & Garity, 2009). The question to be asked is: Does the research instrument or practice tool measure the middle-range theory concept as the nurse researcher or nursing practitioner has defined that concept? Suppose, for example, that a nurse researcher or nursing practitioner defines the concept of pain as the physiological phenomenon of disordered nerve conduction. Measurement of disordered nerve conduction requires a research instrument or practice tool that can reveal the extent of disordered nerve conduction. In contrast, suppose that a nurse researcher or nursing practitioner defines the concept of pain as the experience of an unpleasant body sensation that a person interprets as pain. Measurement of an unpleasant body sensation requires a research instrument or practice tool that can reveal the extent of the unpleasantness of the body sensation that a person interprets as pain.
A largely neglected aspect of measurement validity is determining whether the definition of the middle-range theory concept measured by the research instrument or practice tool is congruent with the focus of the conceptual model that was selected to guide a study or practice. For example, if the conceptual model does not include physiological phenomena, then a research instrument or practice tool that measures the concept of pain defined as the physiological phenomenon disordered nerve conduction is not congruent with that conceptual model. Thus, a nurse researcher or nursing practitioner who is interested in studying or assessing pain within the context of Johnson’s behavioral system model, which focuses on behaviors and does not include physiological phenomena (Fawcett & DeSanto-Madeya, 2013), would not select an instrument or tool that measures the concept of pain defined as the physiological phenomenon of disordered nerve conduction.
In contrast, measuring the concept of pain defined as disordered nerve conduction would be of interest to a nurse who is interested in using the Roy adaptation model as a guide for studying or assessing pain as a response in the physiological mode of adaptation. However, if a nurse is interested in using the Roy adaptation model a a guide for studying or assessing pain as a self-concept mode-physical self-body sensation response, the appropriate research instrument or practice tool should measure the concept of pain defined as an unpleasant body sensation. Noteworthy is that the distinction between and categorization of responses in the physiological mode or the self-concept mode were pointed out by Roy (2009) and discussed by Aber et al. (in press).
In the real world of nursing research and nursing practice, the tendency is to find a research instrument or practice tool that is purported to measure a middle-range theory concept of interest and then determine whether the instrument or tool “fits” with the conceptual model being used to guide the research or practice. Although I fully understand and appreciate the challenges of creating new research instruments and practice tools, I propose that measurement validity is more likely to be supported if nurse researchers and nursing practitioners use a nursing conceptual model as the starting point and develop new research instruments and practice tools that are clearly congruent with the focus of the conceptual model. A recent example is the Antenatal Assessment Instrument (AAI), which is completely congruent with the Roy adaptation model. Specifically, development of the AAI items was guided by the four modes of adaptation of the Roy adaptation model (Lee, Tsang, Wong, & Lee, 2011). Another example is the Adjustment Scale, which is congruent with the Roy adaptation model concept of adaptation level (DeSanto-Madeya & Fawcett, 2009). Many other nursing conceptual model-based research instruments and practice tools are listed in the chapters of Fawcett and DeSanto-Madeya's (2013) book about Johnson's behavioral system model, King's conceptual system, Levine's conservation model, Neuman's systems model, Orem's self-care framework, Rogers' science of unitary human beings, and Roy's adaptation model.
I encourage each reader to very carefully consider measurement validity within the context of the conceptual model that guides his or her research or practice. I look forward to seeing the results of such connections in print.
References
Aber, C., Weiss, M., & Fawcett, J. (in press). Contemporary women’s adaptation to motherhood: The first three to six weeks postpartum. Nursing Science Quarterly.
DeSanto-Madeya, S., & Fawcett, J. (2009). Toward understanding and measuring adaptation level in the context of the Roy adaptation model. Nursing Science Quarterly, 22, 355-359.
Fawcett, J. (1999). The relationship of theory and research (3rd ed.). Philadelphia: F.A. Davis.
Fawcett, J. Aber, C., Haussler, S., Weiss, M., Myers, S. T., Hall, J. L., Waters, L., King, C., Tarkka, M-T., Rantanen, A., Astedt-Kurki, P., Newton, J., & Silva, V. (2011). Women’s perceptions of caesarean birth: A Roy international study. Nursing Science Quarterly, 24, 352- 362.
Fawcett, J., Aber, C., & Weiss, M. (2003). Teaching, practice, and research: An integrative approach benefiting students and faculty. Journal of Professional Nursing, 19, 17-21.
Fawcett, J., & DeSanto-Madeya, S. (2013). Contemporary nursing knowledge: Analysis and evaluation of nursing models and theories (3rd ed.). Philadelphia: F. A. Davis.
Fawcett, J., & Garity, J. (2009). Evaluating research for evidence-based nursing practice. Philadelphia: F. A. Davis.
Lee, L. Y. K., Tsang., A. Y. K., Wong, K. F., & Lee, J. K. L. (2011). Using the Roy adaptation model to develop an antenatal assessment instrument. Nursing Science Quarterly, 24, 363-369.
Orem, D. E. (2001). Nursing: Concepts of practice (6th ed.). St. Louis: Mosby.
Roy, C. (2009). The Roy adaptation model (3rd ed.). Upper Saddle River, NJ: Pearson.
Tulman, L., & Fawcett, J. (2003). Women’s health during and after pregnancy: A theory-guided study of adaptation to change. New York: Springer.
Villarruel, A. M., Bishop, T. L., Simpson, E .M., Jemmott, L. S., & Fawcett, J. (2001). Borrowed theories, shared theories, and the advancement of nursing knowledge. Nursing Science Quarterly, 14, 158-163.
Weiss, M., Fawcett, J., & Aber, C. (2009). Adaptation, postpartum concerns, and learning needs in the first two weeks after cesarean birth. Journal of Clinical Nursing, 18, 2938-2948.
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.