Abstract
Across cognitive behavioral treatments for stress management, individuals' ability to effectively learn intervention components is necessary for improved outcomes. The Management of Current Stress (MOCS)-Part A, formerly known as the Measure of Current Status-Part A, captures the uptake of such "active ingredients" or perceived stress management abilities. The MOCS-A is widely used, yet its psychometric properties are not well-established. Little is known about its generalizability across populations. In the current study, we sought to test the MOCS-A reliability among cancer survivors, examine measurement invariance across sex and language (English/Spanish), and verify the measurement structure. We aggregated participants from six randomized controlled trials of stress management interventions, which ran from 1999 to 2024 and enrolled males with prostate cancer (n=649) and females with breast cancer (n=517). Five trials administered the English MOCS-A, and one administered the measure in Spanish. The MOCS-A consists of a total score and four subscales: ability to relax, awareness of bodily tension, assertiveness, and coping confidence. Depending on model assumptions, we calculated omega or alpha reliability estimates. Confirmatory factor analysis tested measurement invariance through incremental constraints added across subgroups. We compared unidimensional, four-factor, second-order, and bifactor models for best fit. The total measure (ω=.93, α=.91) and four subscales (ω=.75-.91; α=.75-.90) demonstrated acceptable reliability. Strong scalar invariance existed between males and females and Spanish and English versions. The bifactor model best fit the measure structure. The MOCS-A is internally consistent, statistically independent, appropriately scored in their current form, and may be implemented across sexes and languages.