Josie Geller1, Megumi M Iyar2, Allison C Kelly3, Suja Srikameswaran4. 1. St. Paul's Hospital Eating Disorders Program, Vancouver, Canada; University of British Columbia, Department of Psychiatry, Vancouver, Canada. Electronic address: jgeller@providencehealth.bc.ca. 2. St. Paul's Hospital Eating Disorders Program, Vancouver, Canada; University of British Columbia, Department of Psychology, Kelowna, Canada. 3. University of Waterloo, Department eof Psychology, Ontario, Canada. 4. St. Paul's Hospital Eating Disorders Program, Vancouver, Canada; University of British Columbia, Department of Psychiatry, Vancouver, Canada.
Abstract
Individuals with eating disorders (EDs) endorse fears of being self-compassionate, as assessed by the Fears of Compassion for Self scale (FCSelf; Gilbert, McEwan, Matos, & Rivis, 2011), and higher fears predict poorer treatment outcomes. Previous research using the FCSelf demonstrated a single factor solution in a non-clinical sample. OBJECTIVES: This research sought to explore the factor structure of the FCSelf in a clinical population and to examine correlations with associated constructs. The factor structure obtained was compared with its structure in a student sample. METHODS: Individuals with EDs (n = 251) completed the FCSelf, and measures of readiness to change, self-compassion, quality of life, as well as psychiatric and ED symptom severity. A female student sample (n = 314) completed the FCSelf. RESULTS: In the clinical sample a two-factor solution exhibited best fit. Factor 1, labeled "Meeting Standards," contained items expressing concern about showing flaws and losing achievements and relationships. Factor 2, labeled "Emotional Vulnerability," contained items focusing on difficult emotions such as grief, unworthiness, and vulnerability. The two subscales exhibited acceptable internal consistencies and were significantly correlated with clinical outcome variables. In the student sample, the two factor model was superior to a single factor model but given that the latent correlation was so high (r = 0.89) a single summary score is recommended. CONCLUSION: Future research is needed to determine the value of addressing barriers related to meeting standards versus emotional vulnerability in helping individuals with EDs develop self-compassion and recover from their illness.
Individuals with eating disorders (EDs) endorse fears of being self-compassionate, as assessed by the Fears of Compassion for Self scale (FCSelf; Gilbert, McEwan, Matos, & Rivis, 2011), and higher fears predict poorer treatment outcomes. Previous research using the FCSelf demonstrated a single factor solution in a non-clinical sample. OBJECTIVES: This research sought to explore the factor structure of the FCSelf in a clinical population and to examine correlations with associated constructs. The factor structure obtained was compared with its structure in a student sample. METHODS: Individuals with EDs (n = 251) completed the FCSelf, and measures of readiness to change, self-compassion, quality of life, as well as psychiatric and ED symptom severity. A female student sample (n = 314) completed the FCSelf. RESULTS: In the clinical sample a two-factor solution exhibited best fit. Factor 1, labeled "Meeting Standards," contained items expressing concern about showing flaws and losing achievements and relationships. Factor 2, labeled "Emotional Vulnerability," contained items focusing on difficult emotions such as grief, unworthiness, and vulnerability. The two subscales exhibited acceptable internal consistencies and were significantly correlated with clinical outcome variables. In the student sample, the two factor model was superior to a single factor model but given that the latent correlation was so high (r = 0.89) a single summary score is recommended. CONCLUSION: Future research is needed to determine the value of addressing barriers related to meeting standards versus emotional vulnerability in helping individuals with EDs develop self-compassion and recover from their illness.