Regression analysis of compositional data when both the dependent variable and independent variable are components

It is well known that regression analyses involving compositional data need special attention because the data are not of full rank. For a regression analysis where both the dependent and independent variable are components we propose a transformation of the components emphasizing their role as dependent and independent variables. A simple linear regression can be performed on the transformed components. The regression line can be depicted in a ternary diagram facilitating the interpretation of the analysis in terms of components. An exemple with time-budgets illustrates the method and the graphical features

An alternative method for dating unknown tephras based on a segmented regression model

In CoDaWork’05, we presented an application of discriminant function analysis (DFA) to 4 different compositional datasets and modelled the first canonical variable using a segmented regression model solely based on an observation about the scatter plots. In this paper, multiple linear regressions are applied to different datasets to confirm the validity of our proposed model. In addition to dating the unknown tephras by calibration as discussed previously, another method of mapping the unknown tephras into samples of the reference set or missing samples in between consecutive reference samples is proposed. The application of these methodologies is demonstrated with both simulated and real datasets. This new proposed methodology provides an alternative, more acceptable approach for geologists as their focus is on mapping the unknown tephra with relevant eruptive events rather than estimating the age of unknown tephra. Kew words: Tephrochronology; Segmented regression

Regression periods in infancy. A case study from Catalonia.

Based on Rijt-Plooij and Plooij’s (1992) research on emergence of regression periods in the first two years of life, the presence of such periods in a group of 18 babies (10 boys and 8 girls, aged between 3 weeks and 14 months) from a Catalonian population was analyzed. The measurements were a questionnaire filled in by the infants’ mothers, a semi-structured weekly tape-recorded interview, and observations in their homes. The procedure and the instruments used in the project follow those proposed by Rijt-Plooij and Plooij. Our results confirm the existence of the regression periods in the first year of children’s life. Inter-coder agreement for trained coders was 78.2% and within-coder agreement was 90.1 %. In the discussion, the possible meaning and relevance of regression periods in order to understand development from a psychobiological and social framework is commented upon

Psychosocial sources of stress and burnout in the consruction sector : a structural equation model. 'Fuentes psicosociales de estr??s y burnout en el sector de la construcci??n : un modelo de ecuaciones estructurales'

Resumen tomado de la publicaci??n. Resumen tambi??n en ingl??s

Correcting Measurement Error Bias in Interaction Models with Small Samples

Several methods have been suggested to estimate non-linear models with interaction terms in the presence of measurement error. Structural equation models eliminate measurement error bias, but require large samples. Ordinary least squares regression on summated scales, regression on factor scores and partial least squares are appropriate for small samples but do not correct measurement error bias. Two stage least squares regression does correct measurement error bias but the results strongly depend on the instrumental variable choice. This article discusses the old disattenuated regression method as an alternative for correcting measurement error in small samples. The method is extended to the case of interaction terms and is illustrated on a model that examines the interaction effect of innovation and style of use of budgets on business performance. Alternative reliability estimates that can be used to disattenuate the estimates are discussed. A comparison is made with the alternative methods. Methods that do not correct for measurement error bias perform very similarly and considerably worse than disattenuated regression

Psychological contract breach and outcomes : Combining meta-analysis and structural equation models. 'Ruptura de contrato psicol??gico : combinando meta-an??lisis y modelos de ecuaciones estructurales'

Resumen tomado de la publicaci??n

IQ heritability estimation : analyzing genetically-informative data with structural equation models. 'Estimaci??n de la heredabilidad del CI : analizando datos gen??ticamente informativos con modelos de ecuaciones estructurales'

Resumen tomado de la publicaci??n. Resumen tambi??n en ingl??s

Predicting overall service quality: a structural equation modelling approach

In this article, the results of a modified SERVQUAL questionnaire (Parasuraman et al., 1991) are reported. The modifications consisted in substituting questionnaire items particularly suited to a specific service (banking) and context (county of Girona, Spain) for the original rather general and abstract items. These modifications led to more interpretable factors which accounted for a higher percentage of item variance. The data were submitted to various structural equation models which made it possible to conclude that the questionnaire contains items with a high measurement quality with respect to five identified dimensions of service quality which differ from those specified by Parasuraman et al. And are specific to the banking service. The two dimensions relating to the behaviour of employees have the greatest predictive power on overall quality and satisfaction ratings, which enables managers to use a low-cost reduced version of the questionnaire to monitor quality on a regular basis. It was also found that satisfaction and overall quality were perfectly correlated thus showing that customers do not perceive these concepts as being distinct

Use of multilevel logistic regression to identify the causes of differential item functioning

Resumen tomado de la publicaci??n

Regression-based techniques for statistical decision making in singlecase designs

Resumen tomado de la publicaci??n

Additional regression techniques_Solutions_2009.ppt (19)

PowerPoint Slides relating to theory and use of SPSS. Used in Research Skills for Biomedical Science

Correlación SO2/FiO2 con PaO2/FiO2 en niños en ventilación mecánica a grandes alturas : estudio multicéntrico

Cambios en la PaO2 se correlacionan de manera positiva con cambios en la SO2 permitiendo determinar la severidad de la hipoxemia. La búsqueda de un predictor que de forma no invasiva detecte pacientes con mayor compromiso pulmonar ha ganando auge; estableciendo los grados de hipoxemia moderada o severa como criterios para LPA y SDRA, a partir de los valores de PaO2/FiO2 y su correlación con la SO2/FiO2. No se conocen los valores de SO2/FiO2 que a más de 2500msnm permitan identificar la severidad de la hipoxemia en pediatría. Metodología: estudio de correlación y predicción en pacientes de un mes a 18 años de edad admitidos a UCIP, con soporte ventilatorio mecánico y análisis de gases arteriales seriados en dos Hospitales de referencia. Análisis de relación lineal y determinación de la correlación SOFiO2 y PaFiO2 a partir de 430 mediciones. Resultados: el estudio mostro una media para PaO2/FiO2 de 192,12 (DS+75,62) y para SO2/FiO2 de 208,61 (DS+62,79). La correlación SO2/FiO2 y Pa/FiO2 fue positiva y moderada-alta (r= 0,702;p<0.01). A partir de la regresión lineal entre las variables se obtuvo la ecuación de determinación PaO2/FiO2 = (0.92xSO2/FIO2) - 12, con sensibilidad y especificidad de 76% para detectar hipoxemia severa (SO2/FiO2<231), y sensibilidad de 74% y especificidad de 71% para hipoxemia moderada (SO2/FiO2<340). Discusión: los hallazgos obtenidos son muy útiles desde el punto de vista clínico para detectar rápidamente pacientes con hipoxemia moderada y severa, con riesgo potencial de deterioro, cuando no se dispone de línea arterial ó gases arteriales.