Equitability revisited: why the “equitable threat score” is not equitable

In the forecasting of binary events, verification measures that are “equitable” were defined by Gandin and Murphy to satisfy two requirements: 1) they award all random forecasting systems, including those that always issue the same forecast, the same expected score (typically zero), and 2) they are expressible as the linear weighted sum of the elements of the contingency table, where the weights are independent of the entries in the table, apart from the base rate. The authors demonstrate that the widely used “equitable threat score” (ETS), as well as numerous others, satisfies neither of these requirements and only satisfies the first requirement in the limit of an infinite sample size. Such measures are referred to as “asymptotically equitable.” In the case of ETS, the expected score of a random forecasting system is always positive and only falls below 0.01 when the number of samples is greater than around 30. Two other asymptotically equitable measures are the odds ratio skill score and the symmetric extreme dependency score, which are more strongly inequitable than ETS, particularly for rare events; for example, when the base rate is 2% and the sample size is 1000, random but unbiased forecasting systems yield an expected score of around −0.5, reducing in magnitude to −0.01 or smaller only for sample sizes exceeding 25 000. This presents a problem since these nonlinear measures have other desirable properties, in particular being reliable indicators of skill for rare events (provided that the sample size is large enough). A potential way to reconcile these properties with equitability is to recognize that Gandin and Murphy’s two requirements are independent, and the second can be safely discarded without losing the key advantages of equitability that are embodied in the first. This enables inequitable and asymptotically equitable measures to be scaled to make them equitable, while retaining their nonlinearity and other properties such as being reliable indicators of skill for rare events. It also opens up the possibility of designing new equitable verification measures.

The sample autocorrelation function and the detection of long-memory processes

The detection of long-range dependence in time series analysis is an important task to which this paper contributes by showing that whilst the theoretical definition of a long-memory (or long-range dependent) process is based on the autocorrelation function, it is not possible for long memory to be identified using the sum of the sample autocorrelations, as usually defined. The reason for this is that the sample sum is a predetermined constant for any stationary time series; a result that is independent of the sample size. Diagnostic or estimation procedures, such as those in the frequency domain, that embed this sum are equally open to this criticism. We develop this result in the context of long memory, extending it to the implications for the spectral density function and the variance of partial sums of a stationary stochastic process. The results are further extended to higher order sample autocorrelations and the bispectral density. The corresponding result is that the sum of the third order sample (auto) bicorrelations at lags h,k≥1, is also a predetermined constant, different from that in the second order case, for any stationary time series of arbitrary length.

Lifestyle and health status in a sample of Swedish women four years after pregnancy : a comparison of women with a history of normal pregnancy and women with a history of gestational diabetes mellitus

Background: Despite the recommendations to continue the regime of healthy food and physical activity (PA) postpartum for women with previous gestational diabetes mellitus (GDM), the scientific evidence reveals that these recommendations may not be complied to. This study compared lifestyle and health status in women whose pregnancy was complicated by GDM with women who had a normal pregnancy and delivery. Methods: The inclusion criteria were women with GDM (ICD-10: O24.4 A and O24.4B) and women with uncomplicated pregnancy and delivery in 2005 (ICD-10: O80.0). A random sample of women fulfilling the criteria (n = 882) were identified from the Swedish Medical Birth Register. A questionnaire was sent by mail to eligible women approximately four years after the pregnancy. A total of 444 women (50.8%) agreed to participate, 111 diagnosed with GDM in their pregnancy and 333 with normal pregnancy/ delivery. Results: Women with previous GDM were significantly older, reported higher body weight and less PA before the index pregnancy. No major differences between the groups were noticed regarding lifestyle at the follow-up. Overall, few participants fulfilled the national recommendations of PA and diet. At the follow-up, 19 participants had developed diabetes, all with previous GDM. Women with previous GDM reported significantly poorer self-rated health (SRH), higher level of sick-leave and more often using medication on regular basis. However, a history of GDM or having overt diabetes mellitus showed no association with poorer SRH in the multivariate analysis. Irregular eating habits, no regular PA, overweight/obesity, and regular use of medication were associated with poorer SRH in all participants. Conclusions: Suboptimal levels of PA, and fruit and vegetable consumption were found in a sample of women with a history of GDM as well as for women with normal pregnancy approximately four years after index pregnancy. Women with previous GDM seem to increase their PA after childbirth, but still they perform their PA at lower intensity than women with a history of normal pregnancy. Having GDM at index pregnancy or being diagnosed with overt diabetes mellitus at follow-up did not demonstrate associations with poorer SRH four years after delivery.

Impact of worksite health-promotion programs in hospital admission costs and medical leaves in a sample of brazilian companies

Programas de saúde e bem-estar têm sido adotados por empresas como forma de melhorar a saúde de empregados, e muitos estudos descrevem retornos econômicos positivos sobre os investimentos envolvidos. Entretanto, estudos mais recentes com metodologia melhor têm demonstrado retornos menores. O objetivo deste estudo foi investigar se características de programas de saúde e bem-estar agem como preditores de custos de internação hospitalar (em Reais correntes) e da proporção de funcionários que têm licença médica, entre Abril de 2014 e Maio de 2015, em uma amostra não-aleatória de empresas no Brasil, através de parceria com uma empresa gestora de ‘big data’ para saúde. Um questionário sobre características de programas de saúde no ambiente de trabalho foi respondida por seis grandes empresas brasileiras. Dados retirados destes seis questionários (presença e idade de programa de saúde, suas características – inclusão de atividades de screening, educação sobre saúde, ligação com outros programas da empresa, integração do programa à estrutura da empresa, e ambientes de trabalho voltado para a saúde – e a adoção de incentivos financeiros para aderência de funcionários ao programa), bem como dados individuais de idade, gênero e categoria de plano de saúde de cada empregado , foram usados para construir um banco de dados com mais de 76.000 indivíduos. Através de um modelo de regressão múltipla e seleção ‘stepwise’ de variáveis, a idade do empregado foi positivamente associada e a idade do programa de saúde e a categoria ‘premium’ de plano de saúde do funcionário foram negativamente associadas aos custos de internação hospitalar (como esperado). Inesperadamente, a inclusão de programas de screening e iniciativas de educação de saúde nos programas de saúde e bem-estar nas empresas foram identificados como preditores positivos significativos para custos de admissão hospitalar. Para evitar a inclusão errônea de licenças-maternidade, apenas os dados de licença médica de pacientes do sexo masculino foram analisados (dados disponíveis apenas para duas entre as companhias incluídas, com um total de 18.957 pacientes do sexo masculino). Analisando estes dados através de um teste Z para comparação de proporções, a empresa com programa de saúde que inclui atividades voltadas a cessação de hábitos ruins (como tabagismo e etilismo), controle de diabetes e hipertensão, e que adota incentivos financeiros para a aderência de funcionários ao programa tem menor proporção de empregados com licençca médica no período analisado, quando comparada com a outra empresa que não tem estas características (também conforme esperado). Entretanto, a companhia com menor proporção de funcionários com licença médica também foi aquela que adota programa de screening entre as atividades de seu programa de saúde. Potenciais fontes de ameaça à validade interna e externa destes resultados são discutidas, bem como possíveis explicações para a associação entre programas de screening e educação médica a piores indicadores de saúde nesta amostra de companhias são discutidas. Novos estudos com melhor desenho, com amostras maiores e randômicas são necessários para validar estes resultados e possivelmente melhorar a validade interna e externa destes resultados.

Estimating abundance and spatial distribution patterns of the bubble crab Dotilla fenestrata (Crustacea : Brachyura)

The bubble crab Dotilla fenestrata forms very dense populations on the sand flats of the eastern coast of Inhaca Island, Mozambique, making it an interesting biological model to examine spatial distribution patterns and test the relative efficiency of common sampling methods. Due to its apparent ecological importance within the sandy intertidal community, understanding the factors ruling the dynamics of Dotilla populations is also a key issue. In this study, different techniques of estimating crab density are described, and the trends of spatial distribution of the different population categories are shown. The studied populations are arranged in discrete patches located at the well-drained crests of nearly parallel mega sand ripples. For a given sample size, there was an obvious gain in precision by using a stratified random sampling technique, considering discrete patches as strata, compared to the simple random design. Density average and variance differed considerably among patches since juveniles and ovigerous females were found clumped, with higher densities at the lower and upper shore levels, respectively. Burrow counting was found to be an adequate method for large-scale sampling, although consistently underestimating actual crab density by nearly half. Regression analyses suggested that crabs smaller than 2.9 mm carapace width tend to be undetected in visual burrow counts. A visual survey of sampling plots over several patches of a large Dotilla population showed that crab density varied in an interesting oscillating pattern, apparently following the topography of the sand flat. Patches extending to the lower shore contained higher densities than those mostly covering the higher shore. Within-patch density variability also pointed to the same trend, but the density increment towards the lowest shore level varied greatly among the patches compared.

Responsabilidade social e competitividade :um estudo sobre a influência de práticas sociais na decisão de escolha de um destino turístico

The work objective was to investigate the influence of social practices social in the choise decision of a tourist destination. For in such a way, a survey with the aid of a questionnaire, was used as being the research instrument. The study used a simple random sample without replacement, due to elements of the population had a equal probability different of zero, to be selected for forming part of the sample. The used collection method of data was personal interview. The data was collected at Augusto Severo International Airport at the moment in which tourists were embarking in return to the residence place or another tourist destination. For determination of sample size, it was considered the tourist who had visited Natal in November and December, 2004, supplied by the Secretariat of Tourism in the RN. The sample for the research was of 403 people. Results showed that the interviewed express the existence of high level of competitiveness in the tourism industry. It was observed that 42.5% of the interviewed believes to exist a very aggressive competition, and 47.5% believed that the competition is aggressive in the tourism industry. 10.4% of the interviewed expressed much interest in knowing about social practices in the tourism industry and 2.8% had not presented any interest in knowing about social practices in the tourism industry. For the interviewed, the travel agency image is a significant item in the choice for a tourist package, because of 35.6% believed that this factor is very important in the choice. 5% of the interviewed only find that little important or that sometimes the travel agency image can be seen as s decisive factor in the tourist package choice

Minimum size for the occurrence of vortex matter in a square mesoscopic superconductor

The study of superconducting samples in mesoscopic scale presented a remarkable improvement during the last years. Certainly, such interest is based on the fact that when the size of the samples is close to the order of the temperature dependent coherence length xi(T), and/or the size of the penetration depth lambda(T), there are some significant modifications on the physical properties of the superconducting state. This contribution tests the square cross-section size limit for the occurrence (or not) of vortices in mesoscopic samples of area L-2, where L varies discretely from 1 xi(0) to 8 xi(0).The time dependent Ginzburg-Landau (TDGL) equations approach is used upon taking the order parameter and the local magnetic field invariant along the z-direction. The vortex configurations at the equilibrium can be obtained from the TDGL equations for superconductivity as the system relaxes to the stationary state.The obtained results show that the limit of vortex penetration is for the square sample of size 3 xi(0) x 3 xi(0) in which only a single vortex are allowed into the sample. For smaller specimens, no vortex can be formed and the field entrance into the sample is continuous and the total flux penetration occurs at higher values of H/H-c2(0), where H-c2(T) is the upper critical field. Otherwise, for larger samples different vortices patterns can be observed depending on the sample size. (c) 2007 Elsevier B.V. All rights reserved.

A control chart based on sample ranges for monitoring the covariance matrix of the multivariate processes

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Prevalência e localização espacial dos casos de tracoma detectados em escolares de Botucatu, São Paulo - Brasil

OBJETIVO: Avaliar a prevalência de tracoma em escolares de Botucatu/SP-Brasil e a distribuição espacial dos casos. MÉTODOS: Foi realizado um estudo transversal, em crianças de 7-14 anos, que frequentavam as escolas do ensino fundamental de Botucatu/SP, em novembro/2005. O tamanho da amostra foi estimado em 2.092 crianças, considerando-se a prevalência histórica de 11,2%, aceitando-se erro de estimação de 10% e nível de confiança de 95%. A amostra foi probabilística, ponderada e acrescida de 20%, devido à possível ocorrência de perdas. Examinaram-se 2.692 crianças. O diagnóstico foi clínico, baseado na normatização da Organização Mundial da Saúde (OMS). Para avaliação dos dados espaciais, utilizou-se o programa CartaLinx (v1.2), sendo os setores de demanda escolar digitalizados de acordo com as divisões do planejamento da Secretaria de Educação. Os dados foram analisados estatisticamente, sendo a análise da estrutura espacial dos eventos calculadas usando o programa Geoda. RESULTADOS: A prevalência de tracoma nos escolares de Botucatu foi de 2,9%, tendo sido detectados casos de tracoma folicular. A análise exploratória espacial não permitiu rejeitar a hipótese nula de aleatoriedade (I= -0,45, p>0,05), não havendo setores de demanda significativos. A análise feita para os polígonos de Thiessen também mostrou que o padrão global foi aleatório (I= -0,07; p=0,49). Entretanto, os indicadores locais apontaram um agrupamento do tipo baixo-baixo para um polígono ao norte da área urbana. CONCLUSÃO: A prevalência de tracoma em escolares de Botucatu foi de 2,9%. A análise da distribuição espacial não revelou áreas de maior aglomeração de casos. Embora o padrão global da doença não reproduza as condições socioeconômicas da população, a prevalência mais baixa do tracoma foi encontrada em setores de menor vulnerabilidade social.

Spatial patterns or trees in the Picinguara coastal sandy forest (state of Sao-Paulo, Brazil)

The species of the sandy plains forests (forests of the ''restingas'') have not yet had their spatial patterns studied as aids to the understanding of the diversity found in the different physiognomies along the Brazilian coast. In this paper a 10 x 10 m quadrat framework laid in a hectare of a tree dominant forest in the sandy plains of the Picinguaba area of the Serra do Mar State Park (municipality of Ubatuba, state of São Paulo, Brazil) was used to assess the spatial pattern of distribution for the ten most important species : Pera glabrata, Euterpe edulis, Eugenia brasiliensis, Alchornea triplinervea, Guatteria australis, Myrcia racemosa, Jacaranda semiserrata, Guarea macrophylla, Euplassa cantareirae and Nectandra oppositifolia. The spatial patterns were inferred through the calculations of their T-Square Index (C) and Dispersal Distance Index (I). P. glabrata shows a random pattern, E. edulis aggregate, E. brasiliensis, A. triplinervia, G. australis, E. cantareirae and N. oppositifolia with a tendency between aggregate and uniform and, M. racemosa, J. semiserrata and G. macrophylla between aggregate and random. Although the indexes are dependent of the sample size and of the technique adjustments, the relationship of the pattern with the environmental factors is shown by clustering methods. The results give confirmation of how the spatial patterns bring associations between populations and shape of the vegetation physiognomy.

A stochastic shift model for economically designed charts constrained by the process stage configuration

An economic model including the labor resource and the process stage configuration is proposed to design g charts allowing for all the design parameters to be varied in an adaptive way. A random shift size is considered during the economic design selection. The results obtained for a benchmark of 64 process stage scenarios show that the activities configuration and some process operating parameters influence the selection of the best control chart strategy: to model the random shift size, its exact distribution can be approximately fitted by a discrete distribution obtained from a relatively small sample of historical data. However, an accurate estimation of the inspection costs associated to the SPC activities is far from being achieved. An illustrative example shows the implementation of the proposed economic model in a real industrial case. (C) 2011 Elsevier B.V. All rights reserved.

A hierarchy of adaptive (X)over-bar control charts

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Joint X̄ and R Charts with Variable Sample Sizes and Sampling Intervals

Recent studies have shown that the X̄ chart with variable sampling intervals (VSI) and/or with variable sample sizes (VSS) detects process shifts faster than the traditional X̄ chart. This article extends these studies for processes that are monitored by both the X̄ and R charts. A Markov chain model is used to determine the properties of the joint X and R charts with variable sample sizes and sampling intervals (VSSI). The VSSI scheme improves the joint X̄ and R control chart performance in terms of the speed with which shifts in the process mean and/or variance are detected.

Tendência e determinantes do aleitamento materno exclusivo em crianças menores de 6 meses no município de Bauru-SP

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Caracterização dos genes codificadores da hemaglutinina e PB2 do vírus Influenza A (H1N1) pandêmico isolado na mesorregião metropolitana de Belém

A recente pandemia de gripe de 2009/2010 causada pelo vírus A (H1N1) pandêmico mostrou um perfil de gravidade diferente da gripe sazonal, pois um percentual considerável de casos graves e fatais ocorreu em indivíduos adultos jovens, sem comorbidade. A virulência dos vírus Influenza A (H1N1) pandêmico resulta de interações protéicas complexas e depende essencialmente de alguns genes virais. O objetivo deste estudo foi caracterizar os genes codificadores da hemaglutinina (H1) e polimerase básica 2 (PB2) do vírus Influenza A (H1N1) pandêmico mediante a obtenção de cepas provenientes de pacientes com gripe procedente da mesorregião metropolitana de Belém-PA. O tamanho amostral foi constituído de 87 amostras aleatórias de ambos os sexos de 0 a 96 anos, com síndrome respiratória aguda grave (SRAG) sem nenhuma comorbidade relatada, no período de maio de 2009 a agosto de 2010. As amostras foram isoladas em cultura de célula MDCK e analisadas por técnicas de biologia molecular que compreenderam três etapas principais: a) extração do RNA viral (RNAv) a partir do sobrenadante celular; b) amplificação do RNAv pela técnica de Reação em Cadeia mediada pela Polimerase precedida de Transcrição Reversa (RT-PCR); c) sequenciamento completo dos genes codificadores da H1 e PB2. Das 87 cepas amplificadas pelo RT-PCR, em 82 tornou-se possível a obtenção e análise de sequências para o gene HA, enquanto que de 81 amostras virais obteve-se sequências para o gene PB2. A análise comparativa das sequências obtidas com a sequência da cepa vacinal (A/California/07/2009(H1N1)) revelou substituições aminoacídicas na HA (P83S; D97N; S203T; D222G; Q293H e I321V) e na PB2 (K340N; K526R e M631L), no entanto sem associação a hospitalização. Ao nível de substituição na HA, a D97N isolada ou associada com a S203T, foi detectada com mais frequência na primeira onda. Já ao nível da PB2 a substituição K526R foi mais encontrada em cepas que circularam na primeira onda, enquanto que, a M631L foi mais evidenciada na segunda. A substituição D222G na HA só foi encontrada em casos de óbitos. Por fim, observou-se uma tendência de alterações nos sítios antigênicos da HA. Sendo assim, a contínua vigilância genética e antigênica do vírus Influenza A (H1N1) pdm em circulação, bem como o compartilhamento de informações é de extrema importância para a melhor recomendação possível para os vírus que entram na composição vacinal evitando assim maior risco de epidemias severas no futuro.