Beyond qualitative and quantitative data linking: An example from a mixed method study of family recreation

Therapeutic Recreation Journal, Second Quarter 2000 by Mactavish, Jennifer B, Schleien, Stuart J

In a 1993 research note in TRJ, Henderson and Bedini discussed the challenges of linking qualitative and quantitative data within a single study. Since that time, mixed method research has appeared more frequently in the therapeutic recreation literature. Mixed method research, however, has not been fully embraced as underlying philosophical concerns persist (Datta, 1994; Samdahl, 1999). A dominant-less dominant mixed method research design and its application in a study of family recreation is presented in this research note as an approach for addressing some of these concerns. The grounding of this work in the naturalistic paradigm is discussed, as are the influences of this framework on the collection, analysis, and reporting of the data.

KEY WORDS: Research, Mixed Method Research

Introduction

Mixed method research has appeared with greater frequency in a number of social science disciplines, including therapeutic recreation (Tashakkori & Teddlie, 1998). Use of these approaches has not, however, been broadly embraced as concerns persist about the appropriateness of linking qualitative and quantitative data without attending to fundamental ontological, epistemological, and methodological issues' (Henderson & Bedini, 1993; Morse, 1991; Samdahl, 1999; Tashakkori & Teddlie). This research note addresses some of these concerns by describing a mixed method research design that was used to study recreation in families that included children with developmental disabilities (Mactavish, 1994).

Debating the Appropriateness of Mixed Method Research

For several decades, debates have raged among philosophers of science about the superiority of positivist and/or naturalistic research2 for studying society and human behaviour (Denzin & Lincoln, 1994; Guba, 1990; Reichardt & Rallis, 1994; Samdahl, 1999). One aspect of this debate has questioned the appropriateness of combining quantitative and qualitative methods within a single study (Tashakkori & Teddlie, 1998). From a traditional purist perspective, method and paradigm are inextricably linked; therefore, qualitative and quantitative approaches are viewed as distinct and incompatible (Guba; Guba & Lincoln, 1994). Capturing this perspective, Smith and Heshusius (1986) stated that "if one extends the different sets of assumptions to their logical implications, it is clear the two perspectives part company over major issues such as the conceptualization given such basic conditions as validity and reliability, the place of techniques in the inquiry process, and the interpretation of research results" (p. 9). Others, however, have articulated less rigid views arguing that qualitative and quantitative methods are not mutually exclusive research strategies and hence, the appropriateness of combining them (Bullock, 1993; Firestone, 1987; Henderson, 1991; Merriam, 1988; Patton, 1990). Concurring with this stance, Kidder and Fine (1987) stated, "There is nothing mysterious about combining quantitative and qualitative measures. This is, in fact, a form of triangulation that enhances the validity and reliability of one's study" (p. 72).

Not surprisingly, these two schools of thought have resulted in contrasting views about the weaknesses and strengths of mixed method research. From the purist perspective, the inability to reconcile inherent philosophical differences at the paradigmatic level renders mixed method research flawed and inappropriate (Guba & Lincoln, 1994). While acknowledging this concern, adherents of less rigid notions about the relationship between ontology and methodology consider this argument an abstraction that does not detract from the usefulness of mixed method designs (Tashakkori & Teddlie, 1998). Adherents of this view, often called "pragmatists," have abandoned the philosophical debate in favor of a more pragmatic approach-known as paradigm relativism. Paradigm relativism rests on the belief that methodological decisions are driven by the purpose of the research and the questions of interest, not by strict adherence to tenets of any particular worldview (House, 1994; Howe, 1988; Patton, 1990; Reichardt & Rallis, 1994). From this vantage point, the strength of mixed method research lies in its ability to facilitate the study of complex phenomena in ways that cannot be fully accommodated within a single approach (Cherryholmes, 1992; House; Howe; Reichardt & Rallis). More specifically, mixed method research has been advanced as particularly strong when the intention is to (a) seek convergent results (triangulation); (b) explore interconnected and/or distinct aspects of a phenomenon (complementarity); (c) examine similarities, contradictions, and new perspectives (initiation); (d) use methods in a ways that complement one another (e.g., interviews used to inform the development of a survey instrument); and (e) add breadth and scope to a project (Green, Caracelli, & Graham, 1989).

Mixed Method Research in Therapeutic Recreation

Based on an extensive review of recent social science research, Tashakkori and Teddlie (1998) concluded that despite the philosophical tussles, there have been dramatic increases in the use of and literature about mixed method studies and mixed model designs (e.g., Creswell, 1995; Miller & Crabtree, 1994; Morse, 1991).3 Paralleling this trend, examples of mixed method research have emerged within the therapeutic recreation literature (e.g., Bedini, Bullock, & Driscoll, 1993; Bullock, Mahon, & Welch, 1992; Mactavish, Schleien, & Taboume, 1997; Mactavish & Schleien, 1998; Malkin, Howe, & Del Ray, 1989; McAvoy, Schatz, Stutz, Schleien, & Lais, 1989). This emergence is not surprising as multiple perspectives and data collection approaches are promoted as essential to understanding the complex and diverse issues associated with therapeutic recreation practice and research (Bullock, 1993).