RESUMEN

Resumen La negociación de alianza es la capacidad que tienen paciente y terapeuta para resolver problemas vinculares y desacuerdos en terapia. Si bien este concepto es una noción diádica, con implicancias importantes para el tratamiento, actualmente solo existe un instrumento, desde la perspectiva del paciente, que permite su medición. El objetivo de este artículo es presentar el desarrollo y el estudio de las propiedades psicométricas preliminares de la Escala de Negociación de la Alianza, versión terapeuta (ANS-T). A estos fines, se convocó una muestra de 188 terapeutas de distintas aproximaciones teóricas que completaron un conjunto inicial de ítems tentativos del ANS-T junto con el Inventario de Alianza de Trabajo (WAI, por su sigla en inglés), otra escala para medición de la alianza. Para el análisis de datos se realizó un análisis de componentes principales, se estudiaron medidas de confiabilidad como el alfa de Cronbach y correlaciones ítem-total corregidas, así como también evidencias de validez de criterio a partir de las correlaciones entre el ANS-T y la WAI. Los resultados del análisis de componentes principales sugirieron una solución unidimensional integrada por nueve ítems. Los análisis conducidos sobre la versión final de la ANS-T mostraron una adecuada consistencia interna (a = .82) y homogeneidad de sus ítems (r ítem-total corregidas: entre .34 y .66) junto con evidencias de validez de criterio (r con WAI: .51, p < .001). Las cualidades del instrumento y sus propiedades psicométricas sugieren que la ANS-T es un recurso valioso tanto para la clínica como para la investigación.

Abstract Alliance negotiation is defined as the patient and therapist's ability to solve relational problems and disagreements (in therapeutic goals and tasks) during therapy. Although this construct is essentially a dyadic notion, with important clinical implication for treatments of different theoretical frameworks, currently there is only one instrument that can measure alliance negotiation and it is only based on the patients' perspective about therapy. The aim of this paper is to develop and examine the preliminary psychometric properties of the Alliance Negotiation Scale, therapist form. In the context of the importance attributed to analyzing relational constructs with a dyadic perspective, this instrument will not only allow to measure the perception of alliance negotiation from therapist's point of view, but also it will provide the necessary information to have a dyadic measurement of alliance negotiation by using both the original Alliance Negotiation Scale and the Alliance Negotiation Scale, therapist form in the same cases. With this purpose, a sample of 188 therapists from different theoretical frameworks completed an initial pool of twenty four preliminary items of the Alliance Negotiation Scale, therapist form. All the items of Alliance Negotiation Scale, therapist form, were developed based on the original Alliance Negotiation Scale, adapted to be completed by the therapist, and scored in a seven point Likert scale ranging from 1 (Never) to 7 (Always). The sample of therapist also completed the Working Alliance Inventory, therapist from, a widely spread standard measure of therapeutic alliance. For both instruments we asked the therapist to answer the measures thinking about a patient they have recently treated. To avoid a bias in patients' selection and ceiling effects in alliance scores, half of the sample of therapists was randomized to use as a reference a patient they considered easy, while the other half was asked to select a difficult patient. For data analyzes we first ran a principal component analysis in order to find the best factorial solution for the items and to establish the construct validity of the scale. Based on that factorial solution we analyzed the internal consistency of the final version of the Alliance Negotiation Scale, therapist form, using Cronbach's alphas coefficients and item's homogeneity using adjusted item-total correlations. Finally, to test for criterion validity we explored the associations between the Alliance Negotiation Scale, therapist form, and the Working Alliance Inventory, therapist from, using Pearson's correlations. The principal component analysis suggested a one-dimensional solution integrated by nine items, explaining a 42.16% of the total variance. All the items in this solution had factorial weights above .40. Furthermore, the analysis of reliability on this version of the scale showed an adequate internal consistency (Cronbach's a = .82). The adjusted item-total correlations presented also evidences of homogeneity (r's range: between .34 and .66), suggesting that all the items explores different aspects from a common construct, without being redundant. Finally, the correlations between the Alliance Negotiation Scale, therapist form, and the Working Alliance Inventory, therapist from, presented evidences of criterion validity (r with WAI: .51, p < .001). This result showed an association between the measures but also indicate that they do not measure the exact some construct. The specific characteristics and properties of the Working Alliance Inventory, therapist from, presented in this paper, suggested that this measure represents an important contribution both for clinical practice and research. Future studies will need to further explore and confirm the psychometric properties of the Working Alliance Inventory, therapist from, and analyze its application for dyadic measures of the therapeutic alliance negotiation in psychotherapy.