Gestão do conhecimento na Embrapa Meio Ambiente: análise quantitativa da produção científica entre 2004 e 2007.

A análise da produção científica de uma empresa pública de pesquisa, como a Embrapa, é fundamental para que os responsáveis pela gestão em pesquisa e desenvolvimento (P&D) possam estabelecer objetivos e metas para o futuro, inclusive identificando pontos fortes e eventuais lacunas a serem preenchidas. Para essa tarefa, foi construída uma base de dados com os artigos científicos publicados em periódicos indexados, com pelo menos um dos autores pertencente à Embrapa Meio Ambiente, de 2004 a 2007. A partir desses dados foram construídos quadros e tabelas que forneceram uma visão conjunta da produção e produtividade dos pesquisadores daquela Unidade de Pesquisa. Dos 66 pesquisadores estudados, 57 publicaram como autor ou co-autor pelo menos um artigo nos anos considerados e 30 participaram de pelo menos um artigo por ano em média no período. Foram publicados 202 artigos em 103 periódicos diferentes. Há concentração de 65% dos artigos em 32 revistas e apenas 8 pesquisadores publicaram 10 ou mais artigos no período. Foram publicados 57 artigos em 2007, superando em 23% a produção de 2004. Os dados permitirão que seja feita uma análise multidimensional da produtividade da Embrapa Meio Ambiente, para o que se utilizarão ferramentas como a análise de agrupamento para classificar grupos de pesquisadores quanto ao tipo de produção científica. Foi feita a meta-análise adicionando-se aos dados internos de produção o fator de impacto (impact factor, do Institute for Scientific Information) das revistas onde foram publicados os artigos. O fator de impacto total evoluiu de 4,622 em 2004 para 18,316, 18,517 e 41,179 em 2005, 2006 e 2007, respectivamente.

Employing the five-factor mentoring instrument : analysing mentoring practices for teaching primary science.

Primary science education is a concern around the world and quality mentoring within schools can develop preservice teachers’ practices. A five-factor model for mentoring has been identified, namely, personal attributes, system requirements, pedagogical knowledge, modelling, and feedback. Final-year preservice teachers (mentees, n=211) from three Turkish universities were administered a previously validated instrument to gather perceptions of their mentoring in primary science teaching. ANOVA indicated that each of these five factors was statistically significant (p<.001) with mean scale scores ranging from 3.36 to 4.12. Although mentees perceived their mentors to provide evaluation feedback (95%), model classroom management (88%), guide their preparation (96%), and outline the science curriculum (92%), the majority of mentors were perceived not to assist their mentees in 10 of the 34 survey items. Professional development programmes that target the specific needs of these mentors may further enhance mentoring practices for advancing primary science teaching.

Combination strategies for a factor analysis phone-conditioned speaker verification system

This work aims to take advantage of recent developments in joint factor analysis (JFA) in the context of a phonetically conditioned GMM speaker verification system. Previous work has shown performance advantages through phonetic conditioning, but this has not been shown to date with the JFA framework. Our focus is particularly on strategies for combining the phone-conditioned systems. We show that the classic fusion of the scores is suboptimal when using multiple GMM systems. We investigate several combination strategies in the model space, and demonstrate improvement over score-level combination as well as over a non-phonetic baseline system. This work was conducted during the 2008 CLSP Workshop at Johns Hopkins University.

Confirmatory factor analysis of the Way of Coping Checklist-Revised (WCCL-R) in the Asian context

Industrial employment growth has been one of the most dynamic areas of expansion in Asia; however, current trends in industrialised working environments have resulted in greater employee stress. Despite research showing that cultural values affect the way people cope with stress, there is a dearth of psychometrically established tools for use in non-Western countries to measure these constructs. Studies of the "Way of Coping Checklist-Revised" (WCCL-R) in the West suggest that the WCCL-R has good psychometric properties, but its applicability in the East is still understudied. A confirmatory factor analysis (CFA) is used to validate the WCCL-R constructs in an Asian population. This study used 1,314 participants from Indonesia, Sri Lanka, Singapore, and Thailand. An initial exploratory factor analysis revealed that original structures were not confirmed; however, a subsequent EFA and CFA showed that a 38-item, five-factor structure model was confirmed. The revised WCCL-R in the Asian sample was also found to have good reliability and sound construct and concurrent validity. The 38-item structure of the WCCL-R has considerable potential in future occupational stress-related research in Asian countries.