Validation of the Faith Development Scale Using Longitudinal and Cross-Sectional Designs

Article excerpt

Two studies were conducted to further evaluate the validity of the Faith Development Scale (FDS). Study 1 used a longitudinal design and evaluated the change in scores on the scale in response to normal maturational and faith-based experiences over 4 years. Study 2 examined the differences in level of faith development between entering freshmen and graduating senior college students. Both studies found differences consistent with the hypothesis that the FDS is sensitive to: (a) between groups expected to differ in faith development and (b) to changes in faith development within groups over time. Despite limitations within both studies, it was concluded that the results offer further support for the validity of the FDS. Implications of these results are discussed.

Theologian James Fowler's (1981) theory of faith development focuses on how one construes "ultimate reality", rather than the specific content of one's beliefs. Religious growth is seen as a process of evolution through six stages of reasoning about ultimate reality. According to his theory, faith development is characterized by increasing complexity, differentiation, autonomy, humility, and activism in one's faith. His theory is congruent with other influential perspectives on faith development, such as those of Allport (1950) and Batson Batson, Schoenrade, & Ventis, 1993) with their emphasis on open-mindedness coupled with complexity and differentiation of belief as the foundations of a mature faith.

Typically, faith development is assessed through a lengthy interview (Fowler, 1981). While this procedure may describe someone in faith stage terms, it has not yet been shown to possess acceptable psychometric properties (Batson Schoenrade, & Ventis, 1993; Wulff, 1997) as well as being time intensive. Thus, the development of a brief, objective measure of faith styles or faith types based on Fowler's theory was warranted. Further, many social scientists may be interested in an alternative to the two frequently used - but controversial - measures of religious maturity: the Allport and Ross Intrinsic Religiousness (IR) scale (Allport &Ross, 1967), and Batson's Quest scale (Batson, et al., 1993). Kirkpatrick and Hood (1990) have criticized the IR scale for its poor psychometric properties, while others have maintained that the Quest scale does not measure religion-as-quest, but existential anxiety and personal conflict (Donahue, 1985)

In response to this need, Leak, Loucks, and Bowlin (1999) developed and provided initial validation for an 8-item, forced-choice measure of global faith development or faith style: the Faith Development Scale (FDS). One response option reflects Stages 4 or 5 faith development and the other is keyed for Stages 2 or 3 faith development. As a brief summary of their research, one study reported on the construction of the scale with an emphasis on establishing its content validity as well as presenting initial convergent validity evidence. They reported a second study that explored the validity of the new scale through its associations with theoretically important religious and personality characteristics. Other studies used a variety of methodological approaches to establishing construct validity, specifically peer-ratings and "known-group" differences. The results were generally supportive of the validity of the new scale. For example, the scale was related in theoretically expected ways to measures of religious and personality openness as well as to peer ratings of faith development, while remaining uncontaminated by socially desirable response set tendencies.1

RATIONALE FOR THE PRESENT RESEARCH

While the validity of the FDS appears promising, there has been one neglected aspect of its validation strategy to date: there is no direct evidence that the scale can detect changes in faith development over time. It was the purpose of present research to assess such changes, both directly and indirectly, through the use of longitudinal and cross-sectional research designs, respectively. …