(PS7-C66) Altered Factor Structures of Contextual Behavioral Measures of Self in a Diverse Sample

The concept of the self is fundamental within acceptance and commitment therapy (ACT), comprised of three aspects of the self: self as content, self as process, and self as context (SAC). Self as content consists of self-discriptions which are socially reinforced and can vary by culture and self as process evaluates the content in the moment and asserts that one’s experience is constantly changing. By contrast, SAC is a perspective from which one can notice the self-descriptions of self as content, utilizing the perspective of changing situations and environments of self as process, and recognize the self-descriptions are not permanent attributes. However, these components of the self, their relationships to each other, to the ACT model more generally, and across cultures specifically, would benefit from further investigation. Therefore, this study aimed to examine the underlying factor structure of measures of SAC utilizing an intentionally diverse sample. We investigated the factor structure of four self-report functional contextual measures of the self: the self as context scale (SACS; Zettle et al., 2018), the self experience questionnaire (SEQ; Yu et al., 2016), the Reno inventory of self-perspective (RISP; Jeffcoat, 2015), and the context-sensitive self (CSSELF; Kashima et al., 2004). A total of 1083 participants from a mixed community-university sample (52.7% identifying as men; 42.9% identifying as women; 66.1% identifying as a race or ethnicity other than only White; Mage=27.71 [SD=9.47]) completed an online survey. Exploratory and confirmatory factor analyses were conducted in R using the Lavaan package. The two-factor model of the SAC (transcending and centering) was replicated with adequate fit in the  EFA (χ2 = 318.38, p< .001; RMSR = .08) and CFA (χ2(34) = 154.99, p< .001; CFI =.94, TLI =.9, RMSEA =.08, SRMR =.05). The two-factor model of the SEQ (distinction and centering) demonstrated a single factor structure in the EFA (χ2 = 834.35, p< .001; RMSR = .09), and was not replicated with adequate fit in the CFA (χ2(89)=854.92, p< 0.001; CFI = .84, TLI = .83, RMSEA = .12, SRMR = .07) . The three-factor model of the RISP (transcended, centered, entangled) demonstrated a two-factor structure in the EFA (χ2 (53) = 260.7, p< .001) with two items eliminated for high uniqueness, and was not replicated with adequate fit in the CFA (χ2 (41) = 333.16, p< .001; CFI = .89, TLI = .85, RMSEA = .11, SRMR = .09). The two-factor model of the CSSELF (consistent and true) was partially replicated in the EFA (χ2 (8) = 114.08, p< .001) with one item eliminated for high uniqueness, and was not replicated with adequate fit in the CFA (χ2 (13) = 219.13, p< .001; CFI =  .87, TLI = .79, RMSEA = 0.17, SRMR = 0.09). Item analysis of the SEQ, RISP, and CSSELF will be presented in addition to additional measurement invariance analyses. Sample differences between the present study and the original measurement development studies will be described. Future directions and limitations will be addressed.