ABSTRACT
En este capítulo se presentan los modelos de ecuaciones estructurales, una técnica de análisis estadístico multivariante utilizada paracontrastar modelos que proponen relaciones causales entre las variables. Tras la definición de este tipo de modelos y la presentaciónde un ejemplo típico, se discute el concepto de causalidad, para entender su utilización en este contexto. A continuación se discute laestructura general que tiene un modelo, los tipos de variables que se pueden utilizar en ellos y su representación mediante diagramasestructurales, acompañado de la discusión de un ejemplo. Posteriormente se presentan los pasos en la elaboración de un modeloy los tipos de relaciones posibles. También se comentan brevemente el concepto de ajuste y los problemas típicos de estosmodelos. Por último se ofrecen algunos recursos adicionales (AU)
In this chapter, structural equation models (SEM) are presented. SEM is a multivariate statistical technique used to test models proposingcausal relations between their variables. After defining this type of models and presenting a typical example, the concept of causationis discussed, in order to understand its meaning in the present context. The general model structure, the types of variablesused, and how to represent them in path diagrams are discussed, accompanied with an example. Steps needed to build a model arepresented and the different types of relations are commented. Goodness of fit is also briefly commented and also typical problemsfound in these models. Some additional resources are also presented (AU)
Subject(s)
Humans , Data Interpretation, Statistical , Models, Psychological , Multivariate Analysis , CausalityABSTRACT
Los modelos jerárquicos o multinivel permiten analizar datos cuando los casos están agrupados en unidades de información más amplias y se toman medidas tanto en el nivel más bajo (los casos) como en los niveles más altos (los grupos). En este trabajo se describen los modelos multinivel comúnmente tratados en la literatura estadística y se explica cómo ajustarlos utilizando el SPSS (cualquier versión a partir de la 11) y cómo interpretar los resultados. En concreto se describen, ajustan e interpretan los siguientes modelos: (1) análisis de varianza de un factor de efectos aleatorios, (2) análisis de regresión con medias como resultados, (3) análisis de covarianza de un factor de efectos aleatorios, (4) análisis de regresión con coeficientes aleatorios y (5) análisis de regresión con medias y pendientes como resultados. Todos los modelos se describen intentando hacerlos comprensibles para los investigadores en ciencias del comportamiento y la salud
Hierarchic or multilevel models are used to analyse data when cases belong to known groups and sample units are selected both from the individual level and from the group level. In this work, the multilevel models most commonly discussed in the statistic literature are described, explaining how to fit these models using the SPSS program (any version as of the 11th) and how to interpret the outcomes of the analysis. Five particular models are described, fitted, and interpreted: (1) one-way analysis of variance with random effects, (2) regression analysis with means-as-outcomes, (3) one-way analysis of covariance with random effects, (4) regression analysis with random coefficients, and (5) regression analysis with means- and slopes-as-outcomes. All models are explained, trying to make them understandable to researchers in health and behaviour sciences
Subject(s)
Psychometrics/methods , Models, Statistical , Reproducibility of Results , Analysis of Variance , Regression AnalysisABSTRACT
Aunque el concepto de interacción entre factores en el análisis de varianza tiene un significado teórico inequívoco (y así está recogido de forma generalizada en la literatura estadística), en la investigación empírica se producen frecuentemente errores de interpretación que, en muchos casos, conducen a conclusiones incorrectas. En este trabajo se revisan 150 artículos: el 12,7% no presta atención a la interacción (bien porque no se analiza, bien porque se analiza pero no se interpreta); el 79,1% interpreta la interacción recurriendo al análisis de los efectos simples; sólo el 8,2% analiza e interpreta correctamente la interacción. Quizá los investigadores en Psicología tienden a analizar e interpretar la interacción entre factores de forma incorrecta porque los programas informáticos más utilizados (el SPSS a la cabeza) no permiten llevar a cabo las comparaciones necesarias para evaluar una interacción significativa en los diseños factoriales con grupos al azar. Para contribuir a erradicar este problema se muestra cómo diseñar algunas de las comparaciones lineales que permiten aislar el efecto de la interacción y se explica cómo utilizar el SPSS para llevar a cabo esas comparaciones
Although interaction in analysis of variance has an unequivocal theoretical meaning (and so it appears in the statistic literature), frequent misconceptions are found in empirical research, which, in many cases, lead to wrong conclusions. In this paper, 150 articles are reviewed: in 12.7% of them, no attention is paid to the interaction (either because it is not analysed or because it is analysed but not discussed); in 79.1%, interaction is studied through simple effects analysis; and only in 8.2% of the cases, interaction is correctly discussed. It could be that psychology researchers tend to analyse and interpret the interaction between factors incorrectly because the most widespread statistic packages (with SPSS in the lead) do not allow performing the comparisons needed to analyse a significant interaction in factorial designs with randomized groups. In order to contribute to eradicating this problem, we herein show how to design some of the linear comparisons that allow isolating the interaction effect, and we explain how to use SPSS to compute these comparisons
Subject(s)
Humans , Bias , Models, Statistical , Factor Analysis, Statistical , Linear Models , Software Validation , Numerical Analysis, Computer-Assisted/instrumentationABSTRACT
Hierarchic or multilevel models are used to analyse data when cases belong to known groups and sample units are selected both from the individual level and from the group level. In this work, the multilevel models most commonly discussed in the statistic literature are described, explaining how to fit these models using the SPSS program (any version as of the 11 th ) and how to interpret the outcomes of the analysis. Five particular models are described, fitted, and interpreted: (1) one-way analysis of variance with random effects, (2) regression analysis with means-as-outcomes, (3) one-way analysis of covariance with random effects, (4) regression analysis with random coefficients, and (5) regression analysis with means- and slopes-as-outcomes. All models are explained, trying to make them understandable to researchers in health and behaviour sciences.
Subject(s)
Data Interpretation, Statistical , Models, Psychological , Psychology/methods , Psychology/statistics & numerical data , Humans , Random AllocationABSTRACT
Although interaction in analysis of variance has an unequivocal theoretical meaning (and so it appears in the statistic literature), frequent misconceptions are found in empirical research, which, in many cases, lead to wrong conclusions. In this paper, 150 articles are reviewed: in 12.7% of them, no attention is paid to the interaction (either because it is not analysed or because it is analysed but not discussed); in 79.1%, interaction is studied through simple effects analysis; and only in 8.2% of the cases, interaction is correctly discussed. It could be that psychology researchers tend to analyse and interpret the interaction between factors incorrectly because the most widespread statistic packages (with SPSS in the lead) do not allow performing the comparisons needed to analyse a significant interaction in factorial designs with randomized groups. In order to contribute to eradicating this problem, we herein show how to design some of the linear comparisons that allow isolating the interaction effect, and we explain how to use SPSS to compute these comparisons.
Subject(s)
Analysis of Variance , Data Interpretation, Statistical , Models, Psychological , Psychology/methods , Psychology/statistics & numerical data , HumansABSTRACT
El constructo sistema de memoria transactiva (SMT) es una de las aproximaciones teóricas más extendidas para comprender el papel de los mecanismos cognitivos a nivel grupal en la efectividad de los equipos de trabajo. Este estudio examinó las relaciones entre SMT, rendimiento, satisfacción y comunicación funcional en equipos de toma de decisiones. 40 equipos de cuatro personas realizaron una tarea que requería asesorar en el papel de equipos expertos a una compañía ficticia sobre cómo gestionar un problema con los empleados. Los resultados mostraron un efecto de mediación de los SMT en las relaciones entre comunicación funcional, rendimiento percibido y satisfacción. En contra de lo esperado, los SMT no se relacionaban con el rendimiento evaluado por observadores externos. Estos hallazgos sugieren la importancia de la comunicación grupal en la formación de los SMT, así como las implicaciones de éstos para la efectividad de los equipos de toma de decisiones
The transactive memory system (TMS) construct is one of the most extended theoretical approaches to understand the role of cognitive processes at group level in the effectiveness of work teams. This study examined the relationships among TMS, performance, satisfaction and functional communication in decision-making teams. 40 four-person teams were asked to suppose they were expert consulting teams advising a ficticious organization on how to solve a problem with their employees. Results showed a mediating effect of TMS on the relationships between functional communication, perceived team performance and satisfaction. Contrary to our expectations, TMSs were not related to performance as evaluated by external raters. These findings suggest the importance of team communication in the formation of TMSs, as well as their implications for the effectiveness of decision-making teams
Subject(s)
Humans , Group Processes , Memory , Decision Support Techniques , Employee Performance Appraisal , Achievement , Cognition , Mental Processes , Job Satisfaction , Interpersonal RelationsABSTRACT
The transactive memory system (TMS) construct is one of the most extended theoretical approaches to understand the role of cognitive processes at group level in the effectiveness of work teams. This study examined the relationships among TMS, performance, satisfaction and functional communication in decision-making teams. 40 four-person teams were asked to suppose they were expert consulting teams advising a fictitious organization on how to solve a problem with their employees. Results showed a mediating effect of TMS on the relationships between functional communication, perceived team performance and satisfaction. Contrary to our expectations, TMSs were not related to performance as evaluated by external raters. These findings suggest the importance of team communication in the formation of TMSs, as well as their implications for the effectiveness of decision-making teams.
Subject(s)
Decision Making , Efficiency, Organizational , Group Processes , Interpersonal Relations , Adult , Communication , Cross-Sectional Studies , Decision Making, Organizational , Female , Humans , Male , SpecializationABSTRACT
As a result of the pressure to succeed in global markets and the accelerating pace of change in the business environment, organizational flexibility has become an unavoidable issue. We carried out a descriptive study covering 660 firms based in Spain with the objective of finding out how they deal with and use the different facets of organizational flexibility, in particular qualitative (functional) and quantitative (temporal) flexibility. Our results provide favourable evidence for the possibility of combining temporal and functional flexibility. We discuss these results with reference to the «core-periphery» model, which interrelates both kinds of flexibility. Based on these results, we also make three specific propositions concerning the manner in which both kinds of flexibility may be made compatible, and impacts on temporary workers and the development of the labour market in the next years
Las presiones por mantener ventajas competitivas en mercados globales y las consecuencias del acelerado ritmo de cambio que caracteriza el entorno de nuestras organizaciones han convertido la flexibilidad organizacional en un aspecto de máxima relevancia. Con el objetivo de aclarar cómo las empresas españolas están combinando la flexibilidad cualitativa (funcional) con la flexibilidad cuantitativa (temporal) llevamos a cabo un estudio descriptivo con 660 empresas que operan en España. Los resultados aportan evidencia favorable respecto a la posibilidad de combinar diferentes facetas de la flexibilidad (temporal y funcional). Discutiremos los resultados obtenidos con relación al modelo «centro-periferia» que interrelaciona ambos tipos de flexibilidad. Con base en los resultados realizamos también tres proposiciones específicas relativas a la forma en que ambos tipos de flexibilidad puede compatibilizarse, sobre el impacto que tendrán sobre los trabajadores temporales y sobre el desarrollo del mercado de trabajo en años venideros
Subject(s)
Humans , Competitive Behavior , Organization and Administration , Efficiency, Organizational , Organizational Innovation , 16360 , 16054/psychologyABSTRACT
As a response to certain chronic weaknesses in organizational science, we propose a general theoretical alternative that will permit us to redefine our conception of organizations in terms of meaning systems. We discuss the theoretical underpinnings and the basic definitions involved in this approach, identifying the components of organizations and the relationships between them from a representational point of view. Finally, we also consider some implications of this approach for a series of relevant topics in the study of organizations (AU)
Como respuesta a determinadas deficiencias crónicas en la ciencia organizacional, proponemos una alternativa teórica que nos ofrezca una oportunidad para redefinir nuestra concepción de las organizaciones en términos de sistemas de significados. En el presente artículo estableceremos los antecedentes y las definiciones implicadas en esta aproximación e identificaremos desde un punto de vista representacional los componentes de dicho sistema y sus interrelaciones. Finalmente, consideraremos las implicaciones de esta aproximación para diferentes tópicos relevantes en el estudio de las organizaciones (AU)
Subject(s)
Humans , Psychology/methods , Psychology/organization & administration , Models, Organizational , Semantics , Organizational Innovation , Organizational Culture , Psychology, Social/organization & administration , Organizations/organization & administration , Cognition/physiologyABSTRACT
El propósito de este estudio es valorar el efecto de la duración del desempleo sobre la salud mental. Los resultados confirman un comportamiento diferente de las variables resultado de la Salud Mental a lo largo del tiempo. La salud percibida tiende a empeorar progresivamente a medida que aumenta el tiempo del desempleo. El deterioro en la Salud Mental aumenta entre los 7 y los 12 meses y mejora levemente en períodos más largos de desempleo. La respuesta afectiva no se modifica significativamente. El consumo de alcohol y cigarrillos no sufre un incremento significativo. La calidad de vida empeora significativamente después de los seis meses de desempleo y se estabiliza posteriormente. La utilización de servicios de salud aumenta solamente en visitas a profesionales no médicas, no en cuanto a visitas médicas (AU)
The aim of this study is to assess the effect of unemployment duration on Mental Health. Results confirm a differential behaviour of the Mental Health output variables over time. Perceived Health tends to progressively get worse as time of unemployment increases. Mental impairment is increased at 7 to 12 months and improves slightly at longer unemployment periods. Affective response is not modified significantly. Alcohol and cigarette consumption do not suffer a significant increase. Quality of life gets significantly worse after the sixth month of unemployment and levels afterwards. The usage of Health Services is increased only for non medical professional visits, but not for medical visits (AU)
Subject(s)
Adolescent , Adult , Female , Male , Middle Aged , Humans , Unemployment/psychology , Mental Health/statistics & numerical data , Quality of Life/psychology , Tobacco Use Disorder/epidemiology , Alcohol Drinking/epidemiology , Health Status , AffectABSTRACT
This article summarizes the main characteristics of Response Surface Methodology (RSM) and their connections with linear regression analysis. It also presents an example of the application of RSM to the study of person-organization (P-O) fit with real data from 591 subjects. First and second-order polynomial equations were estimated to describe the relationship between a measure of intent to leave and two person and organization commensurate measures. Results show the advantages of using RSM to avoid the traditional methodological problems in the study of P-O fit and to provide an appropriate interpretation of the three-dimensional relationship between O, P and a work outcome (AU)
Aplicación de la metodología de superficies de respuesta en el estudio del ajuste persona - organización. Este artículo resume las principales características de la Metodología de Superficies de Respuesta (MSR) y sus conexiones con el análisis de regresión lineal. Asimismo, presenta un ejemplo de la aplicación de la MSR en el estudio del ajuste persona - organización (P-O) con datos reales de una muestra de 591 sujetos. Se estimaron las ecuaciones polinomiales de primer y segundo orden para describir la relación entre una medida de intención de abandono y dos medidas conmensurables de la persona y la organización. Los resultados muestran las ventajas del uso de la MSR para evitar los tradicionales problemas metodológicos en el estudio del ajuste P-O y proporcionar una interpretación adecuada de la relación tridimensional entre O. P y un resultado de trabajo (AU)
