Bartlett's Test Applied to Variance Component Models

Exact methods for testing equality between variance components obtained from several cases of the same type of balanced orthogonal design are discussed. In particular, methods for successively testing equality of a number of components using Bartlett's tests are outlined for univariate and multivariate responses. Two clinical trial examples of repeated‐measures data are presented.     

Understanding the use of rural space: the need for multi-methods

Although the late 1990s saw increasing use of qualitative data in rural studies and a turn towards issues such as identities and the construction of rurality, many rural researchers still rely on a range of different methods and use both qualitative and quantitative data. However, the challenge of combining quantitative and qualitative data and using different methods is a theme not often dealt with in rural studies, at least not explicitly. This paper (re-)turns the attention to implications of using various methods and combining different types of data for studying a subject matter called ‘the use of rural space’. It concerns both physical land use and the practice and values of individual actors influencing the land use. We emphasise interplay between methodology and philosophy throughout the research process and argue for using multi-methods without compromising the integrity of the different methods. The methodological approach is a combined study of practice and values of individual actors. Two examples—one concerning Senegalese pastoralists’ livelihoods and their use of mobility and one concerning landowners’ location of field afforestation in Denmark—illustrate how the approach facilitates quite different studies of both practice and values and how quantitative and qualitative data can be combined in a non-eclectic way.     

Age-specific survivorship analysis ofHeliothis spp. Populations on cotton

Summary Population dynamics ofHeliothis virescens (F.) andHeliothis zea (Boddie) (Lepidoptera: Noctuidae) eggs and larvae were studied for two years in a small plot of cotton,Gossypium hirsutum (L.). Due to morphological and ecological similarities, the pooledHeliothis population was considered for most of the analyses. Two generations ofHeliothis eggs and larvae were completed during each year. Stage recruitment was estimated for the eggs and larval instars 2–6, and recruitment variances were estimated by a Monte Carlo method. A modified form of the Weibull distribution was developed and used as a model to characterize survivorship curves for each of the fourHeliothis generations. A Type I survivorship curve (mortality rate increasing with age) was inferred for both Generation 1 (early season) data sets, whereas a Type II survivorship curve (mortality rate constant and thus independent of age) was inferred for both Generation 2 (late season) data sets. The shapes of the survivorship curves for the individualH. virescens andH. zea populations were inferred to be the same as those for the pooled populations. Analysis of the contributions of various factors toHeliothis stage-specific mortality indicated that natural enemies (predators and parasites) and the availability of food for larvae were responsible for between-generation differences in survivorship patterns.     

PREDICTING MORTALITY AMONG INDEPENDENTLY LIVING RURAL ELDERLY:

The problem of predicting mortality among rural elderly (65 years and older) living independently is examined using 20-year panel data derived from a random multistage cluster sample. Fifteen independent variables, including social networks, age, sex, and health status, were hypothesized on theoretical and empirical grounds to predict mortality. Face-to-face structured interviews were conducted with the same respondents in 1966, 1974, and 1986/87. Logistic regression establishes that a model comprising age, sex, participation in formal organizations, relative and children association, and general health status is a powerful predictor of mortality. The authors conclude that the more heterogeneous formal, or secondary, social networks which may enhance self esteem are most functional for the elderly.     

Demographic Characteristics as Predictors of Quality of Life in a Population of Psychiatric Outpatients

Studies examining relationships between demographic variables in a general population of psychiatric outpatients and quality of life (QOL), in which QOL was assessed according to current recommendations, have not been performed yet. The aim of this study was to examine one particular aspect of this relationship: the question to what extent QOL scores can be predicted by demographic variables. In a sample of adult Dutch psychiatric outpatients (n=495), demographics were recorded and the participants completed a questionnaire for measuring QOL (WHOQOL-100). The relationships of the demographic variables with the WHOQOL-100 domains Social Relationships and Environment, were stronger than those with the domains Physical Health and Psychological Health. The latter had only significant relationships with educational level and sick leave, which explain little of the variance of the concerning QOL domain. In general, the demographic characteristics used, explained only a relatively small part of the variance in QOL scores. An exception was sick leave, which, in participants with a job, explained an extensive part (27.4%) of the variance of scores on the domain Physical Health.     

Implications of survey design for generalized regression estimation of linear functions

The paper presents a general randomization theory approach to point and interval estimation of Q linear functions T_q = Σ^N₁c_kqY_k(q = 1,…,Q), where Y₁,…,Y_N are values of a variable of interest Y in a finite population. Such linear functions include population and domain means and totals, population regression coefficients, etc. We assume that some auxiliary information can be exploited. This suggests the generalized regression technique based on the fit of a linear model, whereby is created approximately design unbiased estimators T?_q. The paper focuses on estimation of the variance-covariance matrix of the T?_q for single stage and two stage designs. Two techniques based on Taylor expansions are compared. Results of Monte-Carlo experiments (not reported here) show that the coverage properties are good of normal-theory confidence intervals flowing from one or the other variance estimate.     

The Role of Joint Attention in Later Development Among Preterm Children: Linkages Between Early and Middle Childhood

Joint attention is a hallmark of human cognition. It refers to the capacity to coordinate attention to objects and events with attention to other people. Infants display considerable individual differences in this capacity. This longitudinal study of 13‐month‐old preterms was conducted to examine the hypothesis that two different types of joint‐attention skills assessed in an infant–tester paradigm are related to verbal and nonverbal IQ measures through middle childhood. Data are reported separately for the children's tendency to initiate such skills and to respond to an experimenter's offers to share in such behaviours. The results provide support for the hypothesis that the initiation of joint attention makes a unique contribution to later nonverbal IQ apart from variance associated with biomedical risk status and infant development. The results of this study have implications for the conceptualisation of joint‐attention skills, as well as for understanding the relation between joint attention and later cognition.     

HIV-infected African families living in Stockholm/Sweden: their social network, level of disclosure and knowledge about HIV

Most HIV-infected parents in Stockholm are of African origin. Many of them have several children, but their social network is often limited. This article looks at the networks of these families in relation to different aspects of HIV in order to improve the support strategies. The theoretical basis of the study rests on network and ecological systems theory. The social networks of the families in the study were found to be generally of similar size in Sweden and in the home country. However, of 47 HIV-infected parents only 21 (45%) had disclosed their HIV status to friends and relatives in Sweden. The lack of a social network was especially prominent among single women. Contact with counsellors at the HIV clinic and the social welfare office increased the probability of disclosure about HIV infection. The lack of knowledge about HIV transmission and about whom to inform about the infection stresses the importance of improved counselling. The support strategies aimed at broadening the patient's network and breaking their isolation need to be continued and strengthened.     

The Role of Forgiveness in Divorce Adjustment and the Impact of Affect

The role of forgiveness and affect was investigated in relation to divorce adjustment. Ninety-one participants separated from their ex-spouse less than 30 months completed a background questionnaire, the Enright Forgiveness Inventory (EFI), the Self Forgiveness Scale (SFS), the Center for Epidemiological Studies–Depression Inventory (CES–D), the Spielberger State–Trait Anger Inventory (STAXI–2), and the Fisher Divorce Adjustment Scale (FDAS). There were significant relationships between depression and the capacity to forgive and between being in an angry feeling state and harboring a lack of forgiveness toward the ex-spouse. The relationship that emerged as the most powerful was that between trait anger and lack of self-forgiveness. Forgiveness intervention that occurs early in the divorce adjustment process is implicated, particularly in relation to the self.