|
Article Excerpt
In this study, the authors examine the measurement invariance of the Mindful Attention Awareness Scale (MAAS) across adult attachment style. A 1-factor model and measurement invariance was supported across groups. As predicted, latent mean differences showed that securely attached individuals reported significantly more mindfulness than did insecurely attached individuals, providing construct validity evidence for the MAAS.
**********
Mindfulness refers to the process of being attentive to and aware of events and experiences occurring in the present moment (Kabat-Zinn, 1994). Cultivating mindfulness through the practice of meditation has a long history in various spiritual traditions, particularly in Buddhism (Dunn, Hartigan, & Mikulas, 1999). In those settings, mindfulness meditation is practiced because it is believed to provide insight into the nature of the mind and the roots of emotional suffering.
In recent years, mindfulness-based practices have been increasingly incorporated into clinical interventions and wellness programs that are offered in counseling, medical, and mental health settings (Baer, 2003). Frequently cited interventions incorporating mindfulness practices include Mindfulness-Based Stress Reduction (Kabat-Zinn, 1982), Mindfulness-Based Cognitive Therapy (Segal, Williams, & Teasdale, 2002), Dialectical Behavior Therapy (Linehan, 1993), relapse prevention (Marlatt & Gordon, 1985), and Acceptance and Commitment Therapy (Hayes, Strosahl, & Wilson, 1999). A growing body of empirical evidence indicates that participation in these interventions is associated with greater well-being in community, clinical, and student populations (e.g., Baer, 2003; Christopher, Christopher, Dunnagan, & Schure, 2006; Hayes, Luoma, Bond, Masuda, & Lillis, 2006; Robins & Chapman, 2004).
The growing popularity of mindfulness training in the counseling and clinical fields can be attributed to two factors. First is the realization on the part of researchers and practitioners in the areas of consciousness and health care that mindfulness is not a philosophy tied to a particular faith system, but rather a broad practice with universal applications. In particular, mindfulness increases self-awareness, which facilitates recognition of maladaptive cognitions, emotions, and behaviors. These insights tend to promote healthier ways of thinking and behaving, which results in greater well-being (Kabat-Zinn, 2003).
The second contributing factor is a growing body of research indicating that the enhancement of mindfulness is associated with various well-being outcomes, including reduced symptomatology in persons with medical disorders (Kabat-Zinn, Lipworth, & Burney, 1985; Reibel, Greeson, Brainard, & Rosenzweig, 2001); improved psychological functioning in clinical populations (Kristeller & Hallett, 1999; Teasdale et al., 2000); better immune functioning and overall well-being in both community and clinical populations (L. E. Carlson, Speca, Patel, & Goodey, 2003; Davidson et al., 2003; Williams, Kolar, Reger, & Pearson, 2001); and less stress, anxiety, and depression in student populations (Astin, 1997; Rosenzweig, Reibel, Greeson, Brainard, & Hojat, 2003; Shapiro, Schwartz, & Bonner, 1998). In addition, mindfulness has correlated positively with emotional intelligence and self-compassion and negatively with neuroticism, alexithymia, and dissociation (Baer, Smith, Hopkins, Krietemeyer, & Toney, 2006).
MEASURING MINDFULNESS
Despite the fact that mindfulness interventions are now widely available in counseling, medical, and mental health settings and that the literature strongly supports the efficacy of mindfulness training, the assessment of mindfulness has received limited attention to date. Mindfulness assessments are necessary for several important reasons. First, they are needed in the evaluation of interventions that incorporate mindfulness skills, providing a means to assess participants' acquisition of these skills. In particular, mindfulness assessments could help to identify the effective components of mindfulness-based interventions. These interventions typically include elements that are used in addition to mindfulness training, such as yoga instruction and guided relaxation, so it is important to isolate and assess the impact of each component on the acquisition of mindfulness skills. In addition, mindfulness assessments are necessary to support claims made by researchers and clinicians that greater mindfulness produces better health outcomes. For example, a mindfulness scale could be administered as a repeated-measures assessment during counseling or therapy sessions to gauge increases or fluctuations in mindfulness over time and determine if these changes are associated with improvements in various well-being domains.
In recent years, several mindfulness scales have been developed. These scales differ both in their conceptualizations of the construct of mindfulness and in their intended uses. The Freiburg Mindfulness Inventory (FMI; Walach, Buchheld, Buttenmuller, Kleinknecht, & Schmidt, 2006) is notable for being the only mindfulness scale designed to assess mindfulness specifically in a meditation context. The 30-item scale measures nonjudgmental present-moment observation and openness to negative experience. The FMI has demonstrated good internal consistency before ([alpha] = .93) and after ([alpha] = .94) completion of 3- to 14-day meditation retreats (Buchheld, Grossman, & Walach, 2001) and has discriminated between experienced and novice meditators (Walach et al., 2006). Initial exploratory factor analyses (EFAs) suggested a four-factor structure. However, because this structure was found to be unstable over time (i.e., before and after meditation retreats) and the factors were moderately correlated (rs ranging from .48 to .60), the authors recommended interpreting the scores as if they were unidimensional (Buchheld et al., 2001; Walach et al., 2006). These results suggest that further work on the scale is needed to better understand the dimensionality of these scores.
The Kentucky Inventory of Mindfulness Skills (KIMS; Baer, Smith, & Allen, 2004) represents a conceptualization of mindfulness as a set of skills that can be taught and practiced (Dimidjian & Linehan, 2003; Segal et al., 2002). The 39-item, four-factor scale assesses four facets of mindfulness: Observing, Describing, Acting With Awareness, and Accepting Without Judgment. The KIMS has demonstrated adequate psychometric properties, including a four-factor structure, internal consistency reliability (ranging from .76 to .91 for the four subscales), and test-retest reliability (ranging from .65 to .86 for the four subscales; Baer et al., 2004).
The Toronto Mindfulness Scale (TMS; Lau et al., 2006) differs from the KIMS and the FMI in that it assesses mindfulness as a state-like quality, whereas the other two scales measure trait mindfulness. The TMS is a 13-item, two-factor scale assessing the capacity to invoke a mindful state. Items for the TMS reflect Bishop et al.'s (2004) two-component definition of mindfulness: (a) the intentional self-regulation of attention to facilitate greater awareness and (b) a quality of attention characterized by curiosity, acceptance, and openness to experience. There has been very limited study of this measure, but preliminary evidence suggests
that the TMS has adequate internal consistency reliability (Cronbach's [alpha] = .95) and criterion and incremental validity, and there is support for the two-factor structure (Lau et al., 2006).
The mindfulness scale most frequently cited in the literature is the Mindful Attention Awareness Scale (MAAS; Brown & Ryan, 2003). The MAAS measures "the presence or absence of attention to and awareness of what is occurring in the present" (Brown & Ryan, 2003, p. 824); this quality is considered to form the foundation of mindfulness (Brown & Ryan, 2003). As such, the MAAS differs from the FMI and the KIMS in that it is designed to assess mindful states regardless of meditation or mindfulness skills training.
The MAAS developers relied on several sources for generating items for the scale: their personal experience and knowledge of mindfulness, existing scales designed to measure states of consciousness, and publications in the areas of mindfulness and attention. Items were written to reflect both general and specific experiences of mindfulness. Specifically, scale items were designed to capture individual variations in attention and awareness across various domains, including interpersonal relationships and cognitive, affective, and physical states. The initial pool of 184 items was reduced to 15 by using feedback from expert raters, pilot studies with college students, and EFA results (Brown & Ryan, 2003).
Next, the scale developers conducted a series of factor analytic, reliability, and validation studies to provide an initial examination of the properties of the 15-item scale (Brown & Ryan, 2003). The MAAS was shown to have adequate psychometric properties in both student and adult community samples. Specifically, confirmatory factor analyses (CFAs) supported the one-factor model initially revealed in an EFA. The scale also demonstrated adequate internal consistency reliability (as ranging from .82 to .87 across samples) and temporal stability over a 4-week period (interclass correlation coefficient = .81, p < .001). Moderate positive correlations between the MAAS and theoretically related constructs (i.e., emotional intelligence, openness to experience, engagement, and novelty seeking) provided convergent validity evidence for the scale scores. In addition, the lack of correlation between the MAAS and theoretically unrelated constructs (i.e., private self-consciousness and reflection) provided evidence of discriminant validity.
Because the MAAS had been studied only in nonclinical populations, L. E. Carlson and Brown (2005) were interested in examining the measurement invariance of this scale across patient populations with and without cancer. Establishing measurement invariance is a source of construct validity evidence because it addresses the equivalence of measurement across different populations. Researchers often make the assumption of measurement invariance instead of formally testing it (Vandenberg & Lance, 2000). If measurement invariance is not established, differences between groups cannot be unambiguously interpreted (M. S. Thompson & Green, 2006). More specifically, a series of invariance tests (i.e., configural, metric, and scalar invariance) are conducted to allow researchers to investigate whether observed differences in means between groups are due to true differences in the underlying latent construct or are the result of differential item functioning (Little, 1997; Steenkamp & Baumgartner, 1998; M. S. Thompson & Green, 2006; Vandenberg & Lance, 2000; Widaman & Reise, 1997). In the L. E. Carlson and Brown study, the MAAS displayed configural and metric invariance, although no mention was made of scalar invariance. The presence of configural invariance indicated that the...
|
