Resampling schemes with low resampling intensity and their applications in testing hypotheses

The paper explores statistical features of different resampling schemes under low resampling intensity. The original sample is considered in a very general framework of triangular arrays, without independence or equally distributed assumptions, although improvements under such conditions are also provided. We show that low resampling schemes have very interesting and flexible properties, providing new insights into the performance of widely used resampling methods, including subsampling, two-sample unbalanced permutation statistics or wild bootstrap. It is shown that, under regularity assumptions, resampling tests with critical values derived by the appertaining low resampling procedures are asymptotically valid and there is no loss of power compared with the power function of an ideal (but unfeasible) parametric family of tests. Moreover we show that in several contexts, including regression models, they may act as a filter for the normal part of a limit distribution, turning down the influence of outliers.     

The choice among past trends as a basis for the prediction of future trends in old-age mortality

We explored the extent to which projections of future old-age mortality trends differ when different projection bases are used. For seven European countries, four alternative sets of annual rates of mortality change were estimated with age-period log-linear regression models, and subsequently applied to age-specific all-cause mortality rates (80+) in 1999 to predict mortality levels up to 2050. On average, up to 2050, e80 is predicted to increase further by 2.33 years among men and 4.03 years among women. Choosing a historical period of 25 instead of 50 years results in higher predicted gains in e80 for men but lower gains for women. Choosing non-smoking-related mortality instead of all-cause mortality leads to higher gains for women and mixed results for men. In all alternatives there is a strong divergence of predicted mortality levels between the countries. Future projections should be preceded by a thorough study of past trends and their determinants.     

Seeking control in the midst of uncertainty: Women’s experiences of choosing mode of birth after caesarean

Problem

Clinical practice guidelines indicate that over 80% of women with a previous caesarean should be offered a planned vaginal birth after caesarean (VBAC), however only one third of eligible women choose to plan a VBAC. To support informed choices for birth after caesarean, it is necessary to understand the factors that influence women’s decision-making.

MULTIPLE IMPUTATION OF INCOMPLETE CATEGORICAL DATA USING LATENT CLASS ANALYSIS

We propose using latent class analysis as an alternative to log-linear analysis for the multiple imputation of incomplete categorical data. Similar to log-linear models, latent class models can be used to describe complex association structures between the variables used in the imputation model. However, unlike log-linear models, latent class models can be used to build large imputation models containing more than a few categorical variables. To obtain imputations reflecting uncertainty about the unknown model parameters, we use a nonparametric bootstrap procedure as an alternative to the more common full Bayesian approach. The proposed multiple imputation method, which is implemented in Latent GOLD software for latent class analysis, is illustrated with two examples. In a simulated data example, we compare the new method to well-established methods such as maximum likelihood estimation with incomplete data and multiple imputation using a saturated log-linear model. This example shows that the proposed method yields unbiased parameter estimates and standard errors. The second example concerns an application using a typical social sciences data set. It contains 79 variables that are all included in the imputation model. The proposed method is especially useful for such large data sets because standard methods for dealing with missing data in categorical variables break down when the number of variables is so large.     

Intermarriage and the risk of divorce in the Netherlands: the effects of differences in religion and in nationality, 1974-94

A textbook hypothesis about divorce is that heterogamous marriages are more likely to end in divorce than homogamous marriages. We analyse vital statistics on the population of the Netherlands, which provide a unique and powerful opportunity to test this hypothesis. All marriages formed between 1974 and 1984 (nearly 1 million marriages) are traced in the divorce records and multivariate logistic regression models are used to analyse the effects on divorce of heterogamy in religion and national origin. Our analyses confirm the hypothesis for marriages that cross the Protestant-Catholic or the Jewish-Gentile boundary. Heterogamy effects are weaker for marriages involving Protestants or unaffiliated persons. Marriages between Dutch and other nationalities have a higher risk of divorce, the more so the greater the cultural differences between the two groups. Overall, the evidence supports the view that, in the Netherlands, new group boundaries are more difficult to cross than old group boundaries.     

Log-Multiplicative Association Models As Latent Variable Models For Nominal and/or Ordinal Data

Associations between multiple discrete measures are often due to collapsing over other variables. When the variables collapsed over are unobserved and continuous, log-multiplicative association models, including log-linear models with linear-by-linear interactions for ordinal categorical data and extensions of Goodman's (1979, 1985) RC(M) association model for multiple nominal and/or ordinal categorical variables, can be used to study the relationship between the observed discrete variables and the unobserved continuous ones, and to study the unobserved variables. The derivation and use of log-multiplicative association models as latent variable models for discrete variables are presented in this paper. The models are based on graphical models for discrete and continuous variables where the variables follow a conditional Gaussian distribution. The models have many desirable properties, including having schematic or graphical representations of the system of observed and unobserved variables, the log-multiplicative models can be read from the graphs, and estimates of the means, variances, and covariances of the latent variables given values on the observed variables are a function of the log-multiplicative model parameters. To illustrate some of the advantageous aspects of these models, two examples are presented. In one example, responses to items from the General Social Survey (Davis and Smith 1996) are modeled, and in the other example, panel data from two groups (Coleman 1964) are analyzed.     

The role of vertical conflict in the relationship between leader self-enhancement and leader performance

Although studies have shown that inflated self-perceptions of transformational leadership behavior negatively affect leader performance, insight into the underlying processes explaining this relationship is lacking. The current study addresses this gap by identifying vertical conflict between leaders and subordinates as one such underlying process. Using a sample of 52 leaders supervising 259 subordinates, it was found that leader self-enhancement of transformational leadership behavior is significantly and positively related to vertical conflict about both task and relationship issues. Results further indicate that vertical task conflict rather than vertical relationship conflict mediates the negative relationship between leader self-enhancement and leader performance.     

Who Enjoys the Fruits of Growth? Impact of Governments and Markets on Living Standards in Germany,The Netherlands and the U.S.A., 1987–1996

The 1980s and 1990s have been decades of quitegood economic growth in North America and muchof Western Europe. But how have the fruits ofgrowth been shared? This paper reviews changingincome distributions in the U.S., Germany and theNetherlands. These three countries may be takenas exemplars and leading economic performers in``the three worlds of welfare capitalism''(Esping-Andersen, 1990). The U.S. is a liberalwelfare-capitalist state, Germany a corporatiststate, and the Netherlands (less clearly) asocial democratic welfare-capitalist state. Thepaper focuses particularly on income changes inthe bottom, middle and top quintiles and takesa ten year period into account.Previous analyses have shown that labor andmarket income dispersion are increasing, withincreased returns to human capital. Thepotential impact of government through thetax-transfer system has been largely ignored.All three governments redistribute income fromthe rich to the poor. However, the paper showsthat only the Dutch government hasredistributed sufficiently to ensure that thebottom quintile has gained along with others.In Germany and the U.S. the poorest quintile wasconsiderably worse off in absolute terms at theend of the decade.than the beginning. TheGerman government somewhat counteracted thetrend towards greater income dispersion byredistributing to the poorest quintile, so theloss of market income was partly compensated. In the U.S. the impact of government on thepoorest quintile stayed about the same, so thisgroup ended up with about the same decrease indisposable income as market income.The U.S., Germany and the Netherlands are theonly three countries for which ten or moreconsecutive years of panel data are available.The data come from the PSID-GSOEP EquivalentFile 1980-97 and from a comparable fileconstructed from the Dutch SEP data.

     

Validation of prognostic indices using the frailty model

A major issue when proposing a new prognostic index is its generalisibility to daily clinical practice. Validation is therefore required. Most validation techniques assess whether “on average” the results obtained by the prognostic index in classifying patients in a new sample of patients are similar to the results obtained in the construction set. We introduce a new important aspect of the generalisibility of a prognostic index: the heterogeneity of the prognostic index risk group hazard ratios over different centers. If substantial variability between centers exists, the prognostic index may have no discriminatory capability in some of the centers. To model such heterogeneity, we use a frailty model including a random center effect and a random prognostic index by center interaction. Statistical inference is based on a Bayesian approach using a Laplacian approximation for the marginal posterior distribution of the variances of the random effects. We investigate different ways to summarize the information available from this marginal posterior distribution. Our approach is applied to a real bladder cancer database for which we demonstrate how to investigate and interpret heterogeneity in prognostic index effect over centers.