Validation procedures in radiological diagnostic models. Neural network and logistic regression

The objective of this paper is to compare the performance of twopredictive radiological models, logistic regression (LR) and neural network (NN), with five different resampling methods. One hundred and sixty-seven patients with proven calvarial lesions as the only known disease were enrolled. Clinical and CT data were used for LR and NN models. Both models were developed with cross validation, leave-one-out and three different bootstrap algorithms. The final results of each model were compared with error rate and the area under receiver operating characteristic curves (Az). The neural network obtained statistically higher Az than LR with cross validation. The remaining resampling validation methods did not reveal statistically significant differences between LR and NN rules. The neural network classifier performs better than the one based on logistic regression. This advantage is well detected by three-fold cross-validation, but remains unnoticed when leave-one-out or bootstrap algorithms are used.

A logistic regression approach to estimating customer profit loss due to lapses in insurance

This article focuses on business risk management in the insurance industry. A methodology for estimating the profit loss caused by each customer in the portfolio due to policy cancellation is proposed. Using data from a European insurance company, customer behaviour over time is analyzed in order to estimate the probability of policy cancelation and the resulting potential profit loss due to cancellation. Customers may have up to two different lines of business contracts: motor insurance and other diverse insurance (such as, home contents, life or accident insurance). Implications for understanding customer cancellation behaviour as the core of business risk management are outlined.

Adapting football to the child: an application of the logistic regression model in observational mewthodology.

Logistic regression is included into the analysis techniques which are valid for observationalmethodology. However, its presence at the heart of thismethodology, and more specifically in physical activity and sports studies, is scarce. With a view to highlighting the possibilities this technique offers within the scope of observational methodology applied to physical activity and sports, an application of the logistic regression model is presented. The model is applied in the context of an observational design which aims to determine, from the analysis of use of the playing area, which football discipline (7 a side football, 9 a side football or 11 a side football) is best adapted to the child"s possibilities. A multiple logistic regression model can provide an effective prognosis regarding the probability of a move being successful (reaching the opposing goal area) depending on the sector in which the move commenced and the football discipline which is being played.

A Hierarchical Approach for Multi-task Logistic Regression

Peer-reviewed

Determinants of science-based cooperation: evidence in a sample of small and micro firms

We study the determining factors of cience-based cooperation in the case of small and micro firms. In this research, we propose an analytical framework based on the resource-based view of the firm and we identify a set of organisational characteristics, which we classify as internal, external and structural factors. Each factor can be linked to at least one reason, from the firm¿s point of view, to cooperate with universities and public research centres. Each reason can, in turn, be used as an indicator of a firm¿s organisational needs or organisational capacities. In order to validate the theoretical model, we estimate a logistic regression that models the propensity to participate in science-based cooperation activities within a sample of 285 small and micro firms located in Barcelona. The results show the key role played by the absorptive capacity of new and small companies.

Assessing the structural change of strategic mobility: determinants under hypercompetitive environments

The primary purpose of this exploratory empirical study is to examine the structural stability of a limited number of alternative explanatory factors of strategic change. On the basis of theoretical arguments and prior empirical evidence from two traditional perspectives, we propose an original empirical framework to analyse whether these potential explanatory factors have remained stable over time in a highly turbulent environment. This original question is explored in a particular setting: the population of Spanish private banks. The firms of this industry have experienced a high level of strategic mobility as a consequence of fundamental changes undergone in their environmental conditions over the last two decades (mainly changes related to the new banking and financial regulation process). Our results consistently support that the effect of most explanatory factors of strategic mobility considered did not remain stable over the whole period of analysis. From this point of view, the study sheds new light on major debates and dilemmas in the field of strategy regarding why firms change their competitive patterns over time and, hence, to what extent the "contextdependency" of alternative views of strategic change as their relative validation can vary over time for a given population. Methodologically, this research makes two major contributions to the study of potential determinants of strategic change. First, the definition and measurement of strategic change employing a new grouping method, the Model-based Cluster Method or MCLUST. Second, in order to asses the possible effect of determinants of strategic mobility we have controlled the non-observable heterogeneity using logistic regression models for panel data.

Health inequalities among the elderly in western Europe

Background: This paper analyses gender inequalities in health status and in social determinants of health among the elderly in Western Europe. Methods: Data came from the first wave of the “Survey of Health, Ageing and Retirement in Europe” (SHARE, 2004). For the purposes of this study a subsample of community-residing people aged 65-85 years with no paid work was selected (4218 men and 5007 women). Multiple logistic regression models separated by sex and adjusted for age and country were fitted. Results: Women were more likely to report poor health status, limitations in mobility and poor mental health. Whereas in both sexes educational attainment was associated with the three health indicators, household income was only related to poor self-rated health among women. The relationship between living arrangements and health differed by gender and was primarily associated with poor mental health. In both sexes, not living with the partner but living with other people and being the household head was related to poor mental health status (aOR=2.14; 95% CI=1.11-4.14 for men and aOR=1.75; 95% CI=1.12-2.72 for women). Additionally, women living with their partner and other(s) and those living alone were more likely to report poor mental health status (aOR=1.67; 95% CI=1.17-2.41 and aOR=1.58; 95% CI=1.26-1.97, respectively). Conclusions: Health inequalities persist among the elderly. Women have poorer health status than men and in both sexes the risk of poor health status increases among those with low educational attainment. Living arrangements are primarily associated with poor mental health status with patterns that differ by gender.

Gender inequality in health among elderly people in a combined framework of socioeconomic position, family characteristics and social support

This study analyses gender inequalities in health among elderly people in Catalonia (Spain) by adopting a conceptual framework that globally considers three dimensions of health determinants : socio-economic position, family characteristics and social support. Data came from the 2006 Catalonian Health Survey. For the purposes of this study a sub-sample of people aged 65–85 years with no paid job was selected (1,113 men and 1,484 women). The health outcomes analysed were self-perceived health status, poor mental health status and long-standing limiting illness. Multiple logistic regression models separated by sex were fitted and a hierarchical model was fitted in three steps. Health status among elderly women was poorer than among the men for the three outcomes analysed. Whereas living with disabled people was positively related to the three health outcomes and confidant social support was negatively associated with all of them in both sexes, there were gender differences in other social determinants of health. Our results emphasise the importance of using an integrated approach for the analysis of health inequalities among elderly people, simultaneously considering socio-economic position, family characteristics and social support, as well as different health indicators, in order fully to understand the social determinants of the health status of older men and women.

Self-reported social class in adolescents: validity and relationship with gradients in self-reported health

Background: Analyzing social differences in the health of adolescents is a challenge. The accuracy of adolescent's report on familial socio-economic position is unknown. The aims of the study were to examine the validity of measuring occupational social class and family level of education reported by adolescents aged 12 to 18, and the relationship between social position and self-reported health.Methods: A sample of 1453 Spanish adolescents 12 to 18 years old from urban and rural areas completed a self-administered questionnaire including the Child Health and Illness Profile-Adolescent Edition (CHIP-AE), and data on parental occupational social class (OSC) and level of education (LE). The responsible person for a sub-sample of teenagers (n = 91) were interviewed by phone. Kappa coefficients were estimated to analyze agreement between adolescents and proxy-respondents, and logistic regression models were adjusted to analyze factors associated with missing answers and disagreements. Effect size (ES) was calculated to analyze the relationship between OSC, LE and the CHIP-AE domain scores.Results: Missing answers were higher for father's (24.2%) and mother's (45.7%) occupational status than for parental education (8.4%, and 8.1% respectively), and belonging to a non-standard family was associated with more incomplete reporting of social position (OR = 4,98; 95%CI = 1,3–18,8) as was agreement between a parent and the adolescent. There were significant social class gradients, most notably for aspects of health related to resilience to threats to illness.ConclusionAdolescents can acceptably self-report on family occupation and level of education. Social class gradients are present in important aspects of health in adolescents.

"Suddenly I get into the zone": Examining discontinuities and nonlinear changes in flow experiences at work

Work-related flow is defined as a sudden and enjoyable merging of action and awareness that represents a peak experience in the daily lives of workers. Employees" perceptions of challenge and skill and their subjective experiences in terms of enjoyment, interest and absorption were measured using the experience sampling method, yielding a total of 6981 observations from a sample of 60 employees. Linear and nonlinear approaches were applied in order to model both continuous and sudden changes. According to the R2, AICc and BIC indexes, the nonlinear dynamical systems model (i.e. cusp catastrophe model) fit the data better than the linear and logistic regression models. Likewise, the cusp catastrophe model appears to be especially powerful for modelling those cases of high levels of flow. Overall, flow represents a nonequilibrium condition that combines continuous and abrupt changes across time. Research and intervention efforts concerned with this process should focus on the variable of challenge, which, according to our study, appears to play a key role in the abrupt changes observed in work-related flow.

Finite mixture analysis of beauty-contest data using generalised beta distributions

This paper introduces a mixture model based on the beta distribution, without preestablishedmeans and variances, to analyze a large set of Beauty-Contest data obtainedfrom diverse groups of experiments (Bosch-Domenech et al. 2002). This model gives a bettert of the experimental data, and more precision to the hypothesis that a large proportionof individuals follow a common pattern of reasoning, described as iterated best reply (degenerate),than mixture models based on the normal distribution. The analysis shows thatthe means of the distributions across the groups of experiments are pretty stable, while theproportions of choices at dierent levels of reasoning vary across groups.

Social factors related to the clinical severity of influenza cases in Spain during the A(H1N1)2009 virus pandemic

Background During the 2009 influenza pandemic, a change in the type of patients most often affected by influenza was observed. The objective of this study was to assess the role of individual and social determinants in hospitalizations due to influenza A (H1N1) 2009 infection. Methods We studied hospitalized patients (cases) and outpatients (controls) with confirmed influenza A (H1N1) 2009 infection. A standardized questionnaire was used to collect data. Variables that might be related to the hospitalization of influenza cases were compared by estimation of the odds ratio (OR) and 95% confidence intervals (CI) and the variables entered into binomial logistic regression models. Results Hospitalization due to pandemic A (H1N1) 2009 influenza virus infections was associated with non-Caucasian ethnicity (OR: 2.18, 95% CI 1.17 − 4.08), overcrowding (OR: 2.84, 95% CI 1.20 − 6.72), comorbidity and the lack of previous preventive information (OR: 2.69, 95% CI: 1.50 − 4.83). Secondary or higher education was associated with a lower risk of hospitalization (OR 0.56, 95% CI: 0.36 − 0.87) Conclusions In addition to individual factors such as comorbidity, other factors such as educational level, ethnicity or overcrowding were associated with hospitalization due to A (H1N1) 2009 influenza virus infections.

La decisión de endeudamiento y los factores de mayor relevancia. Una aplicación empírica en la empresa industrial española

El presente trabajo aborda el estudio de los factores determinantes del endeudamiento empresarial para contrastar empíricamente la hipótesis del Pecking Order. El endeudamiento empresarial se mide junto a su madurez y para los diferentes tamaños empresariales dada la importancia de diferenciar sus posibles efectos contrapuestos o compensados. Los modelos utilizados para el contraste de hipótesis se han estimado con una muestra de 1.320 empresas manufactureras españolas proporcionada por la Encuesta sobre Estrategias Empresariales (ESEE), para el período 1993-2001. El análisis empírico aplica un modelo multivariante de regresión logística que permite concluir que la teoría del Pecking Order es la de mejor cumplimiento, además de constatarse que las empresas de menor tamaño tienen mayores dificultades de acceso a la financiación con deuda a largo plazo.

Stratification of the severity of critically ill patients with classification trees

Background: Development of three classification trees (CT) based on the CART (Classification and Regression Trees), CHAID (Chi-Square Automatic Interaction Detection) and C4.5 methodologies for the calculation of probability of hospital mortality; the comparison of the results with the APACHE II, SAPS II and MPM II-24 scores, and with a model based on multiple logistic regression (LR). Methods: Retrospective study of 2864 patients. Random partition (70:30) into a Development Set (DS) n = 1808 and Validation Set (VS) n = 808. Their properties of discrimination are compared with the ROC curve (AUC CI 95%), Percent of correct classification (PCC CI 95%); and the calibration with the Calibration Curve and the Standardized Mortality Ratio (SMR CI 95%). Results: CTs are produced with a different selection of variables and decision rules: CART (5 variables and 8 decision rules), CHAID (7 variables and 15 rules) and C4.5 (6 variables and 10 rules). The common variables were: inotropic therapy, Glasgow, age, (A-a)O2 gradient and antecedent of chronic illness. In VS: all the models achieved acceptable discrimination with AUC above 0.7. CT: CART (0.75(0.71-0.81)), CHAID (0.76(0.72-0.79)) and C4.5 (0.76(0.73-0.80)). PCC: CART (72(69- 75)), CHAID (72(69-75)) and C4.5 (76(73-79)). Calibration (SMR) better in the CT: CART (1.04(0.95-1.31)), CHAID (1.06(0.97-1.15) and C4.5 (1.08(0.98-1.16)). Conclusion: With different methodologies of CTs, trees are generated with different selection of variables and decision rules. The CTs are easy to interpret, and they stratify the risk of hospital mortality. The CTs should be taken into account for the classification of the prognosis of critically ill patients.