Further considerations regarding PANAS: Contributions from four studies with different Argentinean samples

Abstract In this research we analyzed the psychometric properties of the Positive and Negative Affect Schedule (PANAS) in samples coming from four independent studies: university students (Study 1, n = 392; Study 2, n = 395), general adult population (Study 3, n = 316), and athletes (Study 4, n = 533). Through confirmatory analyses we evaluated the following models: (a) two-factor model (Watson, Clark, & Tellegen, 1988), (b) three-factor model by Mehrabian (1997), (c) three-factor model by Gaudreau, Sánchez and Blondin (2006) and (d) bi-factor model (Leue & Beauducel, 2011), all of them in their oblique and orthogonal variants. Several models presented an acceptable fit, but only after allowing correlated errors and excluding the items alert and excited. While orthogonal solutions of the three models showed the best fit in Studies 2, 3 and 4, oblique solutions presented the best fit in Study 1. Cronbach's alpha, composite reliability and omega indexes oscillated between .55 and .89. Considering the original theoretical model and the practical utility of the PANAS, we favor the model of two orthogonal factors, excluding the aforementioned items.

Resumen En esta investigación analizamos las propiedades psicométricas del Positive and Negative Affect Schedule (PANAS) en muestras de cuatro estudios independientes: estudiantes universitarios (Estudio 1, n = 392; Estudio 2, n = 395), población general adulta (Estudio 3, n = 316) y atletas (Estudio 4, n = 533). Mediante análisis confirmatorios evaluamos los siguientes modelos: (a) dos factores (Watson, Clark & Tellegen, 1988), (b) tres factores por Mehrabian (1997), (c) tres factores por Gaudreau, Sánchez and Blondín (2006) y (d) bi-factorial (Leue & Beauducel, 2011), todos en sus variantes oblicua y ortogonal. Varios modelos presentaron un ajuste aceptable, pero solo después de permitir correlacionar los errores y excluir los ítems alerta y excitado. Mientras que las soluciones ortogonales de los tres modelos mostraron un mejor ajuste en los Estudios 2, 3 y 4, las soluciones oblicuas presentaron un mejor ajuste en el Estudio 1. El alfa de Cronbach, la confiabilidad compuesta y el índice omega oscilaron entre .55 y .89. Considerando el modelo teórico original y la utilidad práctica del PANAS, nos inclinamos hacia el modelo de dos factores ortogonales, excluyendo los ítems antes mencionados.

El afecto y sus dimensiones: Modelos contrastados mediante análisis factorial confirmatorio de la Escala PANAS / Affection and its dimensions: Models contrasted through confirmatory factor analysis of PANAS Schedule

Dentro de las mediciones mediante autorreporte, una de las escalas más mencionadas en la literatura científica sobre las emociones es la Escala de Afecto Positivo y Negativo (PANAS). Teniendo en cuenta las investigaciones precedentes, el presente trabajo tiene un objetivo principal, el de replicar en una muestra de estudiantes universitarios argentinos los hallazgos previos encontrados en otros países acerca de la dimensión factorial del PANAS. Para esto se utilizó el Análisis Factorial Confirmatorio. Si bien en general los resultados permiten dar cuenta de las propiedades convergente, discriminante y jerárquica del Afecto propuestas por Watson (2000), no se logran corroborar del todo estos postulados. Así, se evidencia que las variables observables Alerta y Excitado presentan una ponderación cruzada, afectando el ajuste global de los modelos. Finalmente, el modelo trifactorial del Afecto es el que presenta mejores índices de ajustes, tanto para el caso oblicuo como ortogonal.

Within self-report measures, one of the most mentioned scales in the scientific literature on emotions is the Positive and Negative Affect Schedule (PANAS). Taking into account previous research, the present work has a main objective, to replicate in a sample of Argentine university students the previous findings found in other countries about the factorial dimension of PANAS. For this, it was used the Confirmatory Factor Analysis. Although in general the results showed that the hierarchical, discriminant and convergent affection properties proposed by Watson (2000) cannot be fully corroborated. Thus, it is evident that the observable variables Alert and Excited have a cross weighting, affecting the overall adjustment of the models. Finally, the trifactorial model of the Affect is the one that presents better indices of adjustments, both for the oblique and orthogonal cases

Afectividad negativa y positiva en adultos cubanos con sintomatología ansiosa, depresiva y sin trastornos / Negative and positive affect in Cuban adults with anxiety, depression and without symptoms

El Modelo Bifactorial de Afectividad de Watson y Tellengen se ha postulado para la diferenciación de la ansiedad y la depresión en Europa y Norteamérica, pero no se han reportado estudios en población cubana de edad media. El objetivo de esta investigación fue investigar la Afectividad Positiva (AP) y la Afectividad Negativa (AN) en personas sin psicopatología, con sintomatología predominantemente ansiosa, depresiva o mixta. La muestra fue de 219 personas; 19,2% con sintomatología ansiosa, 17,8% depresiva, 23,7% mixta y el resto sin síntomas. Se utilizó el Inventario de Afecto Positivo y Negativo (PANAS) para estudiar la afectividad y la Escala de Ansiedad y Depresión de Goldberg (EADG) para diferenciar los subgrupos diagnósticos. Se utilizó el ANOVA de un factor con el Método de la Diferencia Significativa Honesta de Tukey para establecer las diferencias. Se encontró que la AP era menor en los depresivos tal y como predice el Modelo, pero no se cumplió la predicción de que la AP diferenciaría a los ansiosos de las personas sin síntomas, y se encontró que en el grupo con sintomatología mixta había mayor AN y menor AP que en los otros subgrupos patológicos, lo cual alerta sobre la necesidad de evaluar los modelos desarrollados en otras culturas.

Two-Factor Model of Emotion of Watson and Tellengen has been investigated for the differentiation of anxiety and depression in Europe and North America, but no studies have been reported on middle-aged Cuban population. The objective of this research was to investigate the negative (NA) and positive affectivity (PA) in individuals without psychopathology, predominantly anxious, depressive or mixed symptoms. The sample was 219 subjects; 19.2% with symptoms of anxiety, 17.8% has depression, 23.7% mixed anxiety and depression symptoms, and the rest without symptoms. Positive and negative affect schedule (PANAS) was used to study emotion; and Anxiety and Depression Scale Goldberg (GADS) to differentiate diagnostic subgroups. The one-way ANOVA was used with the method of Tukey Honest Significant Difference. We found that AP was lower in depressive as predicted by the model, but the prediction that the AP would differentiate anxious people without symptoms was fulfilled, which highlights the need to evaluate the models developed in other cultures.

Funcionamento Diferencial de Itens (DIF) e experiência de afeto: questões de gênero / Differential Item Functioning and affect: gender issues

O presente estudo teve por objetivo contemplar o critério de justiça dos itens da Escala de Afeto Positivo e Afeto Negativo (PANAS), considerando contribuições das últimas décadas sobre a evidência de diferenças de gênero na manifestação dos traços em questão. Com esse objetivo, a PANAS foi aplicada em 354 participantes, de ambos os sexos. Os dados foram avaliados sob o paradigma da TRI, segundo o modelo para itens politômicos sugerido por Andrich (1988), e o critério de justiça de acordo com o Funcionamento Diferencial de Itens, conforme os parâmetros sugeridos por Linacre (2011), comparando amostras com sujeitos do sexo feminino e masculino. As evidências do presente estudo apontam para itens que favorecem determinadas características dos grupos em questão.

This article is an investigation into the justice criterion for the items of the Positive Affect and Negative Affect Schedule (PANAS) considering contributions from the last decades on gender differences as to emotion and affect experience. For this purpose, the PANAS was administered to 354 participants of both genders. Data was evaluated under the paradigm of IRT, according to the model of polytomous responses suggested by Andrich (1988), and the criterion of justice in accordance with the Differential Item Functioning on the parameters suggested by Linacre (2011) comparing male and female samples. Evidence indicates differential item functioning for specific items, favoring groups according to their characteristics.

Confirmatory factor analysis and temporal invariance of the Positive and Negative Affect Schedule (PANAS) / Análise fatorial confirmatória e invariância temporal da Positive and Negative Affect Schedule (PANAS)

This paper analyzes the structure and the temporal invariance of the Portuguese version of the Positive and Negative Affect Schedule. Previous studies are not consensual whether PANAS measures two or three affect factors and whether such factors are independent or correlated. In order to fill in this gap, we compared the original PANAS, as an independent bi-dimensional structure, with several other alternative structures. Two hundred forty five university students and professional trainees answered the questionnaire in two distinct moments with a two month interval. The model of PANAS with a structure of two independent factors, Positive Affect (PA) and Negative Affect (NA), as proposed by the authors of the scale, was tested. However, the best model consisted of two independent factors, PA and NA, with the cross-loading of the item "excited" between PA and NA, and specified error correlations between the same categories of emotions. Another gap in the literature is the temporal invariance analysis of the PANAS. This paper assesses the temporal invariance of the scale, using the structural equation modeling analysis. Although it was used in its state form version, the PANAS scale showed temporal stability in a two month interval...

O artigo analisa a estrutura e a invariância temporal da versão portuguesa da Positive and Negative Affect Schedule. A investigação científica anterior não é consensual acerca da estrutura da PANAS, se é constituída por dois ou três fatores e se esses fatores são independentes ou correlacionados. De forma a colmatar esta lacuna, procedemos à comparação da estrutura da PANAS como tendo duas dimensões independentes com várias outras estruturas alternativas. Duzentos e quarenta e cinco estudantes universitários e de cursos de formação profissional responderam ao questionário em dois momentos, com dois meses de intervalo. Foi testado o modelo da PANAS que representa uma estrutura de dois fatores independentes, Afeto Positivo (AP) e Afeto Negativo (AN), tal como definida pelos autores da escala. Contudo, o melhor modelo foi o que representa uma estrutura de dois fatores independentes, com a ponderação dupla do item "excitado" entre o AP e o AN e especificando correlações entre os erros dos itens que pertencem às mesmas categorias de emoções. Outra lacuna na investigação é a análise da invariância temporal da PANAS. Este artigo analisa a invariância temporal da escala, utilizando a análise de modelos de equações estruturais. Embora a escala tenha sido aplicada na sua versão afeto estado, a PANAS revelou invariância temporal num intervalo de dois meses...

Adaptación de la Escala de Afectividad Positiva y Negativa (PANAS) a la población de adultos de la ciudad de Córdoba: análisis psicométricos preliminares / Adaptation of Positive and Negative Affectivity Scale (PANAS) to adults in Cordoba city: Preliminary psychometric analysis

El afecto o emoción como tema de investigación científica ha despertado un gran interés en los últimos años, sin embargo, en nuestro medio no se cuenta con instrumentos psicométricos que permitan una medición válida y confiable. El presente trabajo tuvo por objetivo realizar una adaptación psicométrica de la Escala de Afectividad Positiva y Negativa ([PANAS], Watson, Clark & Tellegen, 1988) a la población de adultos de la ciudad de Córdoba, Argentina. Los estudios de estructura interna sugieren la existencia de dos dimensiones subyacentes homogéneas (α= 0.73; α= 0.82) que explican un 39 % de la variabilidad de la prueba. Asimismo, los estudios de correlación ítem-total, de discriminación de ítem y de contrastación de grupos resultaron semejantes a los reportados por trabajos antecedentes. En conjunto, los resultados obtenidos son promisorios y avalan el uso del PANAS en el contexto cordobés.

The affect or emotion as a topic of scientific investigation has been of great interest in recent years, however in our environment there are no psychometric instruments which allow a valid and reliable measurement. The object of this work was to create a psychometric adaptation of the Positive and Negative Affectivity Scale ([PANAS], Watson, Clark & Tellegen, 1988) for the adult population of the city of Cordoba, Argentina. The internal structure studies suggest the existence of two subjacent homogenous dimensions (α= 0.73; α= 0.82) which explain 39% of the variability of the test. Studies of the correlation item total, of item discrimination and contrasted groups resulted similar to those reported in previous works. Altogether, the results obtained are promising and support the use of PANAS in Cordoba.

Multidimensional Comparison of Personality Characteristics in Genetic High Risk for Schizophrenia, First Episode Psychosis and Healthy Controls / 대한정신분열병학회지

OBJECTIVES: Previous studies have suggested that some personality characteristics are associated with the onset, prognosis and social function in schizophrenia patients. However few is known about the personality and affective characteristic of genetic high risk group (GHR) for schizophrenia. This study aimed to investigate the personality and the affective characteristic of GHR group for schizophrenia. METHODS: Participants were 54 healthy controls (HC), 26 subjects with GHR for schizophrenia and 28 subjects with first episode psychosis (FEP). We performed three self-report questionnaires; NEO-Personality Inventory-Revised, State and Anger Expression Inventory and Positive Affect and Negative Affect Schedule. RESULTS: The GHR group showed higher score in agreeableness item than the HC (p=0.028). In extraversion item, the FEP group showed significantly lower scores than the HC (p=0.001). The GHR group showed lower scores in neuroticism item compared with FEP group in trend level. The FEP group showed higher trait-anger, lower positive affect and higher negative affect compared with the others. CONCLUSION: The GHR group seem to share certain vulnerable personality and affective characteristics for schizophrenia with the FEP group. On the other hands, the GHR group appeared to be more agreeable than the other groups, which might act as the compensation for other impaired functions.