Test of Emotion Comprehension: Exploring the underlying structure through Confirmatory Factor Analysis and Similarity Structure Analysis

Some decades of research on emotional development have underlined the contribution of several domains to emotional understanding in childhood. Based on this research, Pons and colleagues (Pons & Harris, 2002; Pons, Harris & Rosnay, 2004) have proposed the Test of Emotion Comprehension (TEC) which assesses nine domains of emotional understanding, namely the recognition of emotions, based on facial expressions; the comprehension of external emotional causes; impact of desire on emotions; emotions based on beliefs; memory influence on emotions; possibility of emotional regulation; possibility of hiding an emotional state; having mixed emotions; contribution of morality to emotional experiences. This instrument was administered individually to 182 Portuguese children aged between 8 and 11 years, of 3rd and 4th grades, in public schools. Additionally, we used the Socially in Action-Peers (SAp) (Rocha, Candeias & Lopes da Silva, 2012) to assess TEC’s criterion-related validity. Mean differences results in TEC by gender and by socio-economic status (SES) were analyzed. The results of the TEC’s psychometric analysis were performed in terms of items’ sensitivity and reliability (stability, test-retest). Finally, in order to explore the theoretical structure underlying TEC a Confirmatory Factor Analysis and a Similarity Structure Analysis were computed. Implications of these findings for emotional understanding assessment and intervention in childhood are discussed.

Medidas individualizadas de mudança: análise das propriedades psicométricas do PSYCHLOPS

Neste estudo apresentam-se dados referentes à fiabilidade, validade e sensibilidade para detetar mudança de uma medida individualizada de mudança (MIM), o PSYCHLOPS. Os dados foram recolhidos em quatro países, e dividem-se por seis amostras, totalizando 1053 participantes. Pretende-se aprofundar o conhecimento acerca das propriedades psicométricas e da sua adequabilidade, em MIM no geral, e no PSYCHLOPS, em particular. Após avaliar as propriedades psicométricas separadamente por amostra, procedeu-se à sua integração mediante métodos de meta-análise. O valor global de alfa de .82 indica uma boa consistência interna, mas a sua interpretação é dificultada pela não equivalência de itens. A correlação entre o PSYCHLOPS e o CORE-OM é de .63, e a correlação entre os scores de mudança das duas medidas é .67. Os dados suportam as hipóteses referentes à validade, mas suscitam dúvidas quanto à estrutura do PSYCHLOPS, e quanto à adequabilidade da avaliação psicométrica tradicional em MIM; Abstract: Various evidence regarding reliability, validity, and responsiveness of PSYCHLOPS, a patient-generated outcome measure (PGOM), are presented. Data comes from three countries, in a total of seven data sets comprising 1053 participants. The purpose of this study is to deepen the understanding of psychometric properties in PGOMs, and in PSYCHLOPS in particular. Following psychometric analysis by sample, meta-analysis procedures were used to calculate overall values. Overall mean internal consistency (alpha) was .82. Overall mean correlation between PSYCHLOPS and CORE-OM is r = .63, and overall mean correlation between change scores is r = .67. Overall mean internal consistency is good, but item non-equivalency makes interpretation difficult. Data support hypotheses regarding correlations, but cast doubt over PSYCHLOPS’ internal structure, and over suitability of classical psychometric evaluation in PGOM.

Quinn’s Leadership Roles: A Confirmatory Factor Analysis Study in Portuguese Health Services

Purpose: To evaluate psychometric properties of Quinn’s leadership questionnaire (CFV questionnaire; 1988) to the Portuguese health services. Design: Cross-sectional study, using the Quinn’s leadership questionnaire, administered to registered nurses and physicians in Portuguese health care services (N = 687). Method: Self-administered survey applied to two samples. In the first (of convenience; N = 249 Portuguese health professionals) were performed exploratory factor and reliability analysis to the CFV questionnaire. In the second sample (stratified; N = 50 surgical units of 33 Portuguese hospitals) was performed confirmatory factor analysis using LISREL 8.80. Findings: The first sample supported an eight-factor solution accounting for 65.46% of the variance, in an interpretable factorial structure (loadings> .50), with Cronbach’s α upper than .79. This factorial structure, replicated with the second sample, showed reasonable fit for each of the 8 leadership roles, quadrants, and global model. The models evidenced, generally, nomological validity, with scores between good and acceptable (.235 < x2/df < 2.055 e .00 < RMSEA < .077). Conclusions: Quinn’s leadership questionnaire presented good reliability and validity for the eight leadership roles, showing to be suitable for use in hospital health care context. Key-Words: Leadership; Quinn’s CVF questionnaire; health services; Quinn’s competing values.