Prediction and Decision Making in Child Welfare

Use of computer-based statistical models to support decision making may reduce the subjectivity and uncertainty in this process of more acceptable levels. Multivariate procedures capture the collective experience of many workers by integrating information from a large number of variables and cases. Once a predictive model has been developed and validated, the profile of new cases can be compared to cases already on the data base. The probability of critical case events such as admission tocare, occurence of abuse, and placement breakdown can be identified. Such capabilities could also be incorporated into the development of computerized information systems. This would faciliate case planning and a more equitable and effective match between client needs and resource utilization.     

Supporting collegiate foster youth and alumni: A mixed-method planning approach for higher education

Despite the proliferation, of on-campus academic support programs for foster youth and alumni (e.g., those formerly in foster care), few conceptual frameworks for developing these programs exist. This study employed Group Concept Mapping (GCM) methodology to delineate a conceptual framework, from the perspective of foster youth and alumni (N = 51), for the development of on-campus support initiatives. GCM is an integrated mixed-method research method that uses non-metric multidimensional scaling and hierarchical cluster analyses, to analyze qualitative data (e.g., statement form) collected during brainstorming sessions. This research was guided by two separate queries: (1) How do foster youth/alumni conceptualize support programs to support foster youth/alumni through to degree completion; and, (2) Is there a difference in priority areas of this conceptualization between current college students, and those who have dropped out of college within the last 18 months? Analyses revealed that participants conceptualized on-campus support programs via an eight-cluster solution, including Campus Awareness, Advocacy, Data Tracking, Pre-College Supports, Fostering Family Connections, Academic Financing, Campus Life, and Peer/Mentor Supports clusters. Additionally, analyses revealed significant priority ratings between areas of this conceptualization. In terms of Importance to academic success, current students rated statements in the Academic Financing and Peer/Mentor Supports clusters significantly lower than did non-students. For Feasibility, current students viewed statements in the Data Tracking, Campus Life, and Peer/Mentor Supports clusters as significantly more feasible than non-students. Pragmatic implications derived from these findings include fostering inclusive campus environments for foster youth/alumni and advocacy undertakings by institutions, on behalf of this population.     

Mental Health Difficulties and the Support Needs of Social Work Students: Dilemmas,Tensions and Contradictions

There are a growing number of students in higher education who experience mental health difficulties and have support needs. The behaviour of small numbers of students on social work programmes indicates also a need to examine the mental health difficulties and needs of social work students. There is only limited factual information on this topic. It is rarely discussed or written about. There is a danger of discriminating against those experiencing such difficulties, especially black or female students. Yet, at the same time, undertaking a social work programme is a demanding experience, as is doing social work itself. Furthermore, on courses leading to a professional qualification, such as social work, there is an obligation to protect the rights of, and assess possible risks to users, as well as to support students. Programmes should give particular attention both to the mental health difficulties and support needs of social work students at selection, during the programme itself and upon conclusion of the programme at a wide range of levels, both within the higher education system and outside that system.     

Some asymptotic properties of point estimation with n-dimensional fuzzy data

Some asymptotic properties of point estimation with n-dimensional fuzzy data with respect to a special L ₂-metric ρ are investigated in this article. It is shown that the collection of all n-dimensional fuzzy data endowed with the ρ-metric is a complete and separable space. Some criterions for point estimation in such fuzzy environments are proposed, and the sample mean, variance and covariance with n-dimensional fuzzy data under these criteria are further studied.     

Job design and levels of physical and mental strain among Australian prison officers

Abstract

Over the past 15 years much has been made of the moderating influence of job discretion and support in occupations that are perceived to be high in demands. Such an effect seems to be most prevalent when subject populations are large and heterogeneous. One aim of this study was to examine the influence of subjective perceptions of these variables on strain reactions among a group of relatively homogeneous workers such as, in the present study, prison officers (n = 274). A second aim was to extend previous research and examine the joint influence of negative affectivity and perceptions of demands, discretion, and support on mental and physical well-being. It was hypothesized that those high in negative affect would be particularly strained by jobs perceived as being demanding. The hypothesized relationship between job demands, discretion and support were found to be weak and at times inconsistent. Social support, in particular, was found to have both positive and negative consequences for different aspects of well-being. Those high in negative affect were in general found to suffer lower mental well-being. More importantly, the interactive effect of negative affectivity and job demands was found to significantly predict the majority of the dependent variables used in this study. The valences of these relationships were inconsistent and are interpreted in light of Warr's (1987) vitamin model of stress.

There are several conclusions that can be drawn from the present study. Negative affectivity appears to have a truly interactive effect with job components, job demands in particular, to influence a range of variables relating to physical and mental health as well a job-related attitudes. The nature of these effects, however, suggests a dissociation in the ways that different independent variables influence a range of indices of well-being and affect. This result requires further investigation since there are important theoretical implications which will follow if the data are replicable.

In order to examine such complex effects, in addition to the need for more longitudinal studies, it is also important that standardized measures of job components are developed so that levels of stress exposure can be equated and their impact assessed across organizational settings.

Finally, the data reported in this study suggest that some effective stress management may be undertaken by adopting selection strategies which take into consideration individual characteristics. In view of the fact that individual stress management programmes are of unproven benefit (Murphy 1986, 1988) such an approach is recommended especially for those organizations which are highly constrained in the services that must be performed.     

The role of stress and social support in Sick Building Syndrome

Sick Building Syndrome (SBS) involves seemingly healthy individuals experiencing long-lasting symptoms of physical distress in their work settings and is estimated to exist in 20 to 30% of all work settings in the USA. We examined differences in stress, social support and both physical and psychological symptoms reported by hospital personnel working in known SBS sites in Halifax, Nova Scotia (n = 297) with control employees working in relatively SBS-free settings (n = 228). We found that employees reporting that their health had been affected by the building in which they worked and those complaining of poor air quality were more likely to be found in SBS locations. In addition, those people with higher levels of organizational support and marginally higher levels of union support were also more likely to be found in SBS locations. Further analyses revealed that employees with higher role overload and greater family support, but lower levels of organizational support were more likely to report that their own health had been adversely affected by their place of work. Perceptions of poor air quality were predicted by higher levels of role conflict, role overload, and organizational stress and lower levels of organizational support. These results suggest that SBS may not be solely dependent on environmental factors.     

带有多面体扰动的半监督v-支持向量分类机

在通常建立的优化模型中,一般都假定输入的数据是精确的,而实际生活中我们得到的数据总会带有测量或统计误差,因此,本文考虑数据在多面体内扰动的半监督两类问题,以v-支持向量分类机为基础,借鉴把半监督两类分类问题转化为一个凹规划的思想,给出数据在多面体内扰动的半监督v-支持向量分类算法。该算法的参数v易于选择,而数值试验也表明该算法具有良好的稳定性和较好的分类结果。     

Estimating Social Influences from Social Networking Sites—Articulated Friendships versus Communication Interactions

Despite the ubiquity of social networking sites, the online social networking industry is in search of effective marketing strategies to better profit from their established user base. Social media marketing strategies build on the premise that the social network of online users can be predicted and social influences among online users can be estimated. However, the existence of various heterogeneous social interactions on social networking sites presents a challenge for social network prediction and social influence estimation. In this article we draw upon the literatures on self‐presentation on social networking sites and signaling in online social networking to categorize six heterogeneous online social interactions on social networking sites into two types—articulated friendships and communication interactions. This article provides empirical evidence for the differences between articulated friendships and communication interactions and the corresponding articulated and communication networks. In order to compare the impacts of the social influences based on these two networks, we utilize support vector machines to build a classifier to predict virtual community membership and we further estimate the marginal effects of these social influences using a two‐stage probit least squares method. We find significant explanatory power of social influences in predicting virtual community membership. Although the communication network is much sparser than the articulated network, social influences based on the communication network achieve similar performance as the articulated network. These findings provide important implications for social media marketing as well as the management of virtual communities.     

Interactive Multiple Objective Decision Support for Sales Force Sizing and Deployment*

An interactive decision aid is introduced for the deployment of two sales resources: salespeople and sales support staff. The aid consists of a normative sales resource allocation model with five objectives and an interactive multiple objective programming solution procedure. The specific decision problem addressed involves the assignment of salespeople and sales support people to customer accounts and the allocation of the time they spend on these accounts. The authors contribute to the existing sales resource modeling literature by dealing with the deployment of two sales resources and interactively solving this problem with respect to five short-run and long-run objectives of the firm. This approach differs from existing sales force modeling efforts in which the solution is found noninteractively by optimizing a single sales resource model with respect to a single objective, often short-run sales. An application of the decision aid to the deployment problem of an industrial sales force manager is presented. Furthermore, useful extensions of the basic sales resource allocation model are discussed.     

Artificial Intelligence and Generalized Qualitative-Response Models: An Empirical Test on Two Audit Decision-Making Domains

Machine learning methods are currently the object of considerable study by the artificial intelligence community. Research on machine learning carries implications for decision making in that it seeks computational methods that mimic input-output behaviors found in classes of decision-making examples. At the same time, research in statistics and econometrics has resulted in the development of qualitative-response models that can be applied to the same kind of problems addressed by machine-learning models—particularly those that involve a classification decision. This paper presents the theoretical structure of a generalized qualitative-response model and compares its performance to two seminal machine-learning models in two problem domains associated with audit decision making. The results suggest that the generalized qualitative-response model may be a useful alternative for certain problem domains.