Home About us Contact
|
Home About us Contact | ||
|
|
||
Regression Weights (regression + weight)
Selected AbstractsPersonality and the predisposition(s) to bipolar disorder: heuristic benefits of a two-dimensional modelBIPOLAR DISORDERS, Issue 5 2007Greg Murray Objectives:, The aim of this study was to model normal personality correlates of the predisposition(s) to bipolar disorder (BD), and in so doing explore the proposition that the tendency to bipolar depression [trait depression (T-Depression)] and the tendency to mania [trait mania (T-Mania)] can usefully be viewed as separable but correlated dimensions of BD predisposition. Methods:, A well student sample (n = 176, modal age 18,25 years, 71% female) completed the NEO Personality Inventory,Revised and the General Behavior Inventory. Results:, A good-fitting model (normed ,2 = 0.60, significance of ,2 = 0.73) was identified in which T-Depression was determined solely by neuroticism, while T-Mania was determined by extraversion and (negative) agreeableness. The pathway from T-Depression to T-Mania was also significant (standardized regression weight = 0.80), with a weaker significant reciprocal path (coefficient = 0.27). A model in which bipolar vulnerability was represented as a single dimension (T-Bipolarity) also provided a good fit to the data, but provided less heuristic power. Conclusions:, Predisposition to BD can be usefully understood in terms of two reciprocally related dimensions of vulnerability (T-Depression and T-Mania), which can be separated on the basis of their personality correlates. [source] Pitfalls of Ability ResearchINTERNATIONAL JOURNAL OF SELECTION AND ASSESSMENT, Issue 4 2001Thomas R. Carretta Ability research in psychology can be fraught with pitfalls that lead to inappropriate conclusions. We identify several issues that lead to potential misinterpretation of results and suggest corrective solutions. These issues include lack of construct validity of the measures, misinterpretation of correlations and regression weights, lack of statistical power, failure to estimate cross-validation effects, and misinterpretation of factor analytic results. [source] Structural Modeling of Car Use on the Way to the University in Different Settings: Interplay of Norms, Habits, Situational Restraints, and Perceived Behavioral Control,JOURNAL OF APPLIED SOCIAL PSYCHOLOGY, Issue 8 2009Christian A. Klöckner This manuscript presents the results of the application of an extended norm activation model to the explanation of car use on the way to the university with a sample of 430 students of 3 German universities. The proposed two-stage structural model is supported by the data. First, a norm activation process starting with awareness of consequences activates subjective and personal norms. Second, behavior is determined by car-use habits, perceived behavioral control (PBC), car access, and effort to use public transportation. The influence of personal norms on behavior is mediated by habits. Subgroup analyses of the second stage of the model show a high structural stability, but differences in the regression weights. [source] An interactive education session and follow-up support as a strategy to improve clinicians' goal-writing skills: a randomized controlled trialJOURNAL OF EVALUATION IN CLINICAL PRACTICE, Issue 1 2010Elisabeth Marsland BAppSc(OT)Hons Abstract Background, Recent research indicates that allied health clinicians have difficulty articulating client needs and priorities into specific and measurable goals. As a result, a number of strategies to facilitate improvement in allied health clinicians' goal-setting skills have been recommended in the literature. In order to assist clinicians develop the skills required to set SMART goals, it is necessary that the strategies are rigorously tested. Aim, To determine if a 50-minute education session and 3-month email and telephone support programme improves clinicians' SMART goal-writing skill and accurately predicts improved goal-writing behaviour. Methods, Concealed random allocation of participants (n = 120) into two parallel groups: (1) intervention group received education on writing goals using the SMART Goal Evaluation Method as part of a workshop on outcome measurement and received 3 months of follow-up support (n = 60); and (2) control group attended a workshop on evidence-based practice (n = 60). Results, Education and follow-up support improved clinicians' SMART goal-writing skills at both the 3- and 6-month review (Yates ,2 = 4.324, d.f. = 1, P = 0.0375). Structural equation modelling revealed education and follow-up support is an accurate predictor of SMART goal-setting behaviour change at both 3 months (standardized regression weights = 0.21; P = 0.014) and 6 months (standardized regression weights = 0.19; P = 0.02) post intervention. Changes were modest and developed over a 6-month period. Conclusion, This study provides empirical evidence that a programme of educating clinicians in a standardized method of goal setting and providing follow-up support improves allied health clinicians' SMART goal-writing skills. [source] Robust designs for misspecified exponential regression modelsAPPLIED STOCHASTIC MODELS IN BUSINESS AND INDUSTRY, Issue 2 2009Xiaojian Xu Abstract We consider the construction of designs for exponential regression. The response function is an only approximately known function of a specified exponential function. As well, we allow for variance heterogeneity. We find minimax designs and corresponding optimal regression weights in the context of the following problems: (1) for nonlinear least-squares (LS) estimation with homoscedasticity, determine a design to minimize the maximum value of the integrated mean-squared error (IMSE), with the maximum being evaluated for the possible departures from the response function; (2) for nonlinear LS estimation with heteroscedasticity, determine a design to minimize the maximum value of IMSE, with the maximum being evaluated over both types of departures; (3) for nonlinear weighted LS estimation, determine both weights and a design to minimize the maximum IMSE; and (4) choose weights and design points to minimize the maximum IMSE, subject to a side condition of unbiasedness. Solutions to (1),(4) are given in complete generality. Copyright © 2009 John Wiley & Sons, Ltd. [source] Integrating multidimensional geophysical dataARCHAEOLOGICAL PROSPECTION, Issue 1 2006Kenneth L. Kvamme Abstract Surveys that utilize multiple geophysical methods offer greater insights about the subsurface because each one generally yields different information. Common approaches to integrating or ,fusing' multidimensional geophysical data are investigated utilizing computer graphics, geographical information system (GIS), mathematical and statistical solutions. These approaches are synthesized into graphical, discrete and continuous domains. It is shown that graphical approaches allow complex visualizations of the subsurface, but only images are generated and their dimensionality tends to be low. Discrete methods incorporate any number of geophysical dimensions, allow application of powerful Boolean operations, and produce unambiguous maps of anomaly presence or absence, but many of these methods rely on arbitrary thresholds that define only robust anomalies. Continuous data integrations offer capabilities beyond other methods because robust and subtle anomalies are simultaneously expressed, new quantitative information is generated, and interpretive data are derived in the form of regression weights, factor loadings, and the like, that reveal interrelationships and underlying dimensionality. All approaches are applied to a common data set obtained at Army City, Kansas, a World War I era commercial complex that serviced troops in nearby Camp Funston (now Fort Riley). Utilizing data from six geophysical surveys (magnetic gradiometry, electrical resistivity, ground-penetrating radar, magnetic susceptibility, soil conductivity, aerial thermography), various data integrations reveal the structure of this nearly forgotten town. Copyright © 2005 John Wiley & Sons, Ltd. [source] | ||||||||||