ABSTRACT
Resumen El modelo dimensional alternativo para los trastornos de personalidad incluye 25 facetas (rasgos patológicos) organizadas en cinco dominios de orden superior (Desapego, Afectividad Negativa, Psicoticismo, Antagonismo y Desinhibición). Para evaluar este modelo, se desarrolló el Personality Inventory for DSM-5 (PID-5), que posee dos versiones: una extensa (220 ítems) que evalúa dominios y facetas, y una breve (25 ítems) que evalúa solo los dominios. En un trabajo anterior, se brindó evidencia favorable para una versión breve (31 ítems) adaptada para ser utilizada en población argentina. En el presente trabajo se estudian las propiedades psicométricas de una versión reducida y modificada del PID-5 que permite evaluar ambos componentes por medio de una cantidad de ítems (108). La validez convergente se evaluó a través de la relación con una medida de rasgos de personalidad normal del Modelo de los Cinco Grandes Factores. Se trabajó con una muestra de tipo no probabilística de n = 525 sujetos de población general, que respondieron la versión adaptada del PID-5 y el Listado de Adjetivos para Evaluar la Personalidad. Los resultados brindaron evidencia de validez y confiabilidad para el instrumento. El Análisis Factorial Exploratorio y Confirmatorio sugirió un buen ajuste de la estructura pentafactorial. La consistencia interna resultó adecuada y los ítems presentaron buenos índices de discriminación. Se observaron diferencias de género y edad, y correlaciones con los factores correspondientes de los cinco grandes. Esta versión puede ser utilizada para evaluar el modelo, con fines tanto clínicos como de investigación, y con ventajas respecto al tiempo de administración respecto a la versión extensa original.
Abstract The official classification of personality disorders in the latest edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) remains categorical. However, a dimensional alternative for personality disorders is presented as an emerging model. The model is organized in five higher order domains (Negative Affectivity, Detachment, Antagonism, Disinhibition and Psychoticism), with relationships with the Big Five Model of Personality, strongly established within the Personality Psychology. The proposal also includes 25 facets or second-order traits, included within the main domains. Domains and facets represent psychopathological traits with clinical relevance. To assess this model, the Personality Inventory for DSM-5 (PID-5) was developed. PID-5 has two forms: extensive (220 items) that assesses domains and facets, and brief (25 items) that assesses only the domains. In a previous study, evidence for a short version (31 items) adapted to the Argentine population was provided, that overcomes some of the limitations of the original one. In this work, the psychometric properties of a reduced and modified version of the PID-5 are studied, which allows evaluating five domains and 25 facets, through a reduced number of items (108). We worked with a non-probabilistic sample of n = 525 subjects from the general population, who answered the adapted version of the PID-5 and the Adjectives Checklist to Assess the Big Five Personality Factors (AEP), a Big Five Model measure. The following data analyses were performed: (1) Exploratory and Confirmatory Factor Analysis to evaluate the internal structure of PID-5; (2) reliability analysis to assess the internal consistency of the PID-5 scales; (3) item analysis to assess discriminating power; (4) multivariate analysis of variance (MANOVA) to examine significant differences due to gender and age; and (5) bivariate correlation analysis to analyze PID-5 convergent validity. The results provided evidence of validity and reliability. Exploratory and Confirmatory Factor Analysis suggested a five-factor structure. The facets presented factor loadings in the domain theoretically expected, with some exceptions: Suspiciousness (loaded in Psychoticism), Hostility (loaded in Disinhibition), Depressivity (loaded in Detachment) and Insensitivity (loaded in Detachment). CFA also suggested a good model fit (CFI = .98; RMSEA = .04; SRMR = 0.083). Psychoticism, Detachment, and Disinhibition facets had their higher factor loadings in the expected domain. Negative affectivity showed higher correlations with the rest of the scales. Internal consistency was satisfactory, especially at the domain level, and the items had good discrimination indices. Correlations with the corresponding of the Big Five factors were observed, similar to previous studies. The five PID-5 domains were also found positively correlated. Additionally, gender and age differences were found. In line with previous literature, results suggest that some facets scales are "pure" markers of these domains (e. g., Psychoticism and Antagonism facets), whereas others (e. g., Negative Affectivity facets such as Depressiveness, Suspicion, Hostility), are located "in between" domains since they share features of more than one domain. Psychoticism facets presented higher loadings in their domains and lower in the rest. This is not surprising; although most of psychopathology cannot be understood as categories, schizophyte and Schizotypal Personality Disorder are exceptions, and Psychoticism would be the representation of these categories in the APA model. Findings also provide evidence of convergent validity for the instrument, as well as theorical evidence regarding the relationship between normal and pathological personality traits. This version can be used to evaluate the model, both in research and clinical practice. It has advantages over the original longer version, in terms of administration time and participants' fatigue, while maintaining its psychometric properties. The results are also expected to contribute to the recent literature on the dimensional approach to personality psychopathology. However, complementary studies, particularly with a clinical population, are needed.
ABSTRACT
Actualmente uno de los debates centrales en el campo de la suicidología refiere a si los comportamientos suicidas y las autolesiones no suicidas (NSSI) representan clusters diferentes o un continuum en el espectro autodestructivo. Si se las considerara entidades nosológicas separadas debiera ser factible identificar características diferenciales entre ellas de manera que, por ejemplo, fuera posible predecir en qué casos de NSSI sería más probable que un adolescente cometiera un intento suicida. Este estudio se propuso establecer la capacidad predictiva de habilidades emocionales e interpersonales para ambos tipos de comportamientos. Para ello se administraron el Inventario de Orientaciones Suicidas ISO-30, la Escala Rasgo de Metaconocimientos sobre Estados Emocionales TMMS-24 y las escalas Recursos Sociales de las Escalas de Recursos Psicológicos a 510 adolescentes escolarizados de la ciudad de Mar del Plata, Argentina. Se seleccionaron 81 casos con autolesiones no suicidas (Grupo NSSI) y 61 con ideaciones suicidas (Grupo IS). Los resultados muestran que los modelos de regresión obtenidos para cada grupo son diferentes. El ajuste general del modelo del Grupo IS resultó adecuado (χ² = 10.54; p = .22), los coeficientes de determinación revelan una capacidad explicativa que da cuenta de entre un 25 y 46% de la varianza total y la variable de mayor peso en el modelo fue la Incapacidad para Pedir Ayuda. En cambio el modelo del grupo NSSI mostró un pobre ajuste y solamente quedaron incluidas las variables emocionales. Se discuten los resultados según consecuencias aplicadas y futuras líneas de investigación.
Currently, one of the central debates in the field of suicidology refers to whether suicidal behavior and non suicidal self-injury (NSSI) represent different clusters or a continuum in the self-harm spectrum. The first approach, which runs in the same line than DSM-5, defines the NSSI as self-inflected injuries on any surface the own body with the goal of releasing negative thought or feelings, as a reaction to an interpersonal conflict, aiming at the achievement of a positive affective state, clearly understanding that in any case these injuries could cause death. The second approach states that the non-suicidal qualification represents a restriction and is potentially dangerous, given the close link between NSSI and suicidal behavior. If they were separate nosological entities it should be possible to identify differential features in each one. Literature focused on the study of pathological variables, suggest that it seems to be a difference of degree rather than qualitative since adolescents showing suicidal behaviors report more severe psychopathological symptoms, more depression, more hopelessness and loneliness, many family conflicts, and finding less reasons to stay alive compared to those who only exhibit NSSI. These data, somehow, offer support to both approaches. Following this debate, this study aimed at establishing the predictive value of emotional and interpersonal skills as inde-pendent variables for both types of behaviors. The selection of these variables was based on previous research findings, which provide robust evidence regarding the huge difficulties showed by adolescents with suicidal behaviors and NSSI in emotional expression and regulation. As for of interpersonal variables, diverse studies offer consistent results about the low perception of social support, the presence of intense feeling of loneliness and alienation, as well as the communication difficulties manifested by adolescents with suicidal behaviors. Based on this background, the hypothesis states that such difficulties with emotional skills will be useful predictors for both groups, only showing differences of degree. This way, adolescents with NSSI will show fewer difficulties in this aspect than the suicidal ideation group; mean while difficulties in interpersonal skills will arise as predictors only for the suicidal ideation group. The Inventory for Suicidal Orientations, the Trait Meta-Mood Scale and the Social Resources Scales from the Psychological Resources Scales were used to examine to 510 adolescent students from the city of Mar del Plata, Argentina. From the total sample, 142 participants were selected and separated into two groups. Suicidal Ideation Group -IS- (82% females, M = 16.16; SD = 1.11) accomplished at least one of the following criteria for high suicidal risk with active suicidal ideation: a) total score of ISO-30 ≥ 45; b) score ≥ 2 in the item #30 of the ISO-30 scale, "If my problems get worst, I will kill myself"; c) score≥ 1 in item 30 and total score ≥ 37 (P 84) in ISO-30, or d) explicit information on ideation and/or suicidal attempt in responses to personal data survey. The other group -NSSI- was composed of 81 cases (71.6% females, M = 16.35; SD =1.13) according to the following criteria: (a) having reported self-injury thoughts and/or committing self-injuries; (b) not achieving any of the 4 criteria established for the suicidal ideation group. Results identified different regression models for each group. The overall fit showed for IS-Group model was adequate (χ²= 10.54; p =.22). As for coefficients of determination reported, 25 to 46% of the total variance was explained, and the inability to ask for help played a major role within predictive variables. Instead, the model for the NSSI Group achieved a poor fit, including only emotional variables, as hypothesized. Results and further research lines are discussed.