Beyond criticism, hostility, and emotional overinvolvement: Development and psychometric validation of the Comprehensive Expressed Emotion Scale (CEES) in schizophrenia

Expressed Emotion (EE) is a well-established predictor of schizophrenia relapse, but concerns remain regarding the content validity of its traditional components—criticism, hostility, and emotional overinvolvement (EOI). Recent research suggests a covert EE dimension (disassociation, apathy) unmeasured by current scales. This study developed and validated a self-report EE scale, the Comprehensive Expressed Emotion Scale (CEES), to reconceptualize EE. 319 adults with schizophrenia in Mainland China completed CEES, relevant Family Environment Scale—Chinese Version (FES-CV) subscales, and SF-12. Ten patient–family dyads also underwent the Camberwell Family Interview (CFI). Exploratory and confirmatory factor analyses determined factor structure; reliability and validity were evaluated using established indices and correlations. EFA supported a three-factor model—overt (criticism, hostility), covert (disassociation, apathy), and EOI—explaining 64 % of variance. CFA confirmed strong loadings for overt and covert items (0.70–0.91 and 0.72–0.90; all p < 0.001), but weaker, mostly non-significant loadings for EOI items. The three-factor model showed good fit (χ 2 (402) = 657.17, CFI = 0.971, TLI = 0.968, RMSEA = 0.061, SRMR = 0.079). The total scale and overt and covert subscales showed high internal consistency and strong construct and concurrent validity; EOI subscale reliability was moderate. CEES operationalizes covert EE in Chinese families, broadening EE assessment to include implicit conflict. It demonstrates strong psychometric properties and potential for targeted interventions, though EOI subscale reliability and generalizability remain limitations. • A novel three-factor structure of Expressed Emotion (EE) was identified in Chinese families of individuals with schizophrenia, consisting of: (1) an overt factor (criticism and hostility), (2) a covert factor (disassociation and apathy), and (3) emotional over-involvement (EOI) as an independent construct. • Criticism and hostility, traditionally viewed as separate EE components, were found to load onto a single overt dimension, highlighting their continuum in overt abrasive behavior and challenging the established tripartite EE model. • The study is the first to quantitatively operationalize covert abrasive behaviors in EE, demonstrating that emotional disengagement—previously described only qualitatively—can be reliably measured in Chinese schizophrenia-affected families. • The Comprehensive Expressed Emotion Scale (CEES) demonstrated excellent reliability (total scale α = 0.95, overt α = 0.94, covert α = 0.94) and strong construct, convergent, and concurrent validity; the EOI subscale showed moderate reliability (α = 0.71). • These findings expand the conceptualization of EE by incorporating both overt and covert dimensions, offering a more nuanced and empirically validated assessment of family emotional climate in schizophrenia research.