50716 research outputs found
Sort by
A shape-enhanced outlier-detection framework for regression tasks
No full textOutliers significantly undermine the robustness of regression analysis. However, classical robust regression methods often fail to effectively handle diverse data scenarios. To address this limitation, we propose a novel outlier-detection framework that fundamentally differs from conventional approaches by leveraging data morphological features and residual structural characteristics to ensure robust performance across various outlier types, sample sizes, dimensions, and contamination rates. This framework operates through three sequential stages: partition, partial purification, and full recovery. The partition stage employs a recursive clustering-based approach to segment data into spatially adjacent and linearly structured components, facilitating outlier isolation. The partial purification stage then extracts a reliable subset of clean samples through a strategy based on least median of squares (LMS), providing an initial robust estimate. Finally, the full recovery stage identifies clustered and sparse outliers by examining the shape and magnitude of robust residuals, enabling comprehensive clean data reconstruction. Extensive evaluations on synthetic and real-world datasets confirm the superior performance of the method across challenging conditions, as validated through statistical analysis, sensitivity analysis, and scalability analysis.</p
A Decolonising Critical Discourse Analysis Framework for Positive Behaviour Support Plans
No full textPositive Behaviour Support Plans (Plans) are documents written by practitioners about and for Service Users who are exhibiting “challenging behaviours”, which may risk harm to the self or others. These Service Users are subject to restrictive practices, including physical or environmental restraints, seclusion, and psychotropic medication. This article presents a Decolonising Critical Discourse Analysis (DCDA) Framework for the analysis of Plans and other disability-related texts, iteratively developed through analysis of 16 Plans, conversations with disability stakeholders, reflexive team discussions, and a review of the literature. The aims of the research discussed in this article are to document and analyse discourses evident in the sample Plans and to inform critical and socially just Plan authorship practices. The purpose of this article is to report on the Framework development and offer an analysis of early findings. Plans are viewed as texts that create and maintain a complex interplay between macrolevel forces (discourse and policy) and microlevel practices of plan writing. The DCDA Framework emphasises decolonising language, making Whiteness visible, enacting disability justice and neurodivergent-affirming practice, and upholding a body politic analysis. This methodological Framework is applicable to a range of texts and contexts to explore how language can be utilised to construct Service User identities. IMPLICATIONS A Decolonising Critical Discourse Analysis (DCDA) Framework and method can be applied effectively to analyse Positive Behaviour Support Plans (Plans). Recognition is needed that risk–centric language pathologises neurodivergence and frames Service Users as requiring close regulation. Psychotropic medication is a frequently utilised restrictive practice. Cultural identity is not well–considered in this sample of Plans. Social workers have a responsibility to use their advocacy skills to critically engage with the power of language and to disrupt dominant narratives about disabled Service Users.</p
Evaluation of a residential group therapy program for Canadian first responders
No full textObjectives: To assess the effects of a four-day, group-based residential treatment program focused on enhancing psychological health and social functioning of firefighter and police first responders in British Columbia, Canada. Methods: Using a repeated-measures design, participants completed seven validated self-report questionnaires at baseline (pre-test T0), two weeks after the session (post-test T1), and six months follow-up (T2). A multilevel approach to the analysis of repeated measures examined the effects of the program on several indicators. These included symptoms of major depressive disorder, symptoms of generalized anxiety, symptoms of post-traumatic stress disorder, social role functioning, social support (giving and receiving), quality of life, and health related impairments. Results: A total of 106 police and 114 firefighters undertook the program. The baseline measures were completed by 207 (94 %) participants and 175 (80 %) completed at least one questionnaire at the longest follow-up. All outcomes measured improved from baseline to two-weeks post intervention (T0 to T1, p < 0.001), and sustained reductions at six months follow-up (T0 to T2, p < 0.001), except for giving social support. The highest standardized effect size (Cohen's d) observed at six months (T2) was for symptoms of major depressive disorder (d = −0.90), followed by symptoms of generalized anxiety disorder (d = −0.75), symptoms of post-traumatic stress disorder (d = −0.69), symptom distress (d = −0.62), social role challenges (d = −0.58), quality of life (d = 0.44), interpersonal relations (d = −0.40), and receiving emotional support (d = 0.23). Conclusion: Participation in this program appeared to improve subjective ratings of Canadian firefighter and police psychological health and social functioning. The results are promising and require further exploration with a randomized trial and longer-term follow-up.</p