Robust response transformations based on optimal prediction

Nonlinear regression problems can often be reduced to linearity by transforming the response variable (e.g., using the Box-Cox family of transformations). The classic estimates of the parameter defining the transformation as well as of the regression coefficients are based on the maximum likelihood criterion, assuming homoscedastic normal errors for the transformed response. These estimates are nonrobust in the presence of outliers and can be inconsistent when the errors are nonnormal or heteroscedastic. This article proposes new robust estimates that are consistent and asymptotically normal for any unimodal and homoscedastic error distribution. For this purpose, a robust version of conditional expectation is introduced for which the prediction mean squared error is replaced with an M scale. This concept is then used to develop a nonparametric criterion to estimate the transformation parameter as well as the regression coefficients. A finite sample estimate of this criterion based on a robust version of smearing is also proposed. Monte Carlo experiments show that the new estimates compare favorably with respect to the available competitors.

Estimation of percentage body fat in 6- to 13-year-old children by skinfold thickness, body mass index and waist circumference.

We evaluated the accuracy of skinfold thicknesses, BMI and waist circumference for the prediction of percentage body fat (PBF) in a representative sample of 372 Swiss children aged 6-13 years. PBF was measured using dual-energy X-ray absorptiometry. On the basis of a preliminary bootstrap selection of predictors, seven regression models were evaluated. All models included sex, age and pubertal stage plus one of the following predictors: (1) log-transformed triceps skinfold (logTSF); (2) logTSF and waist circumference; (3) log-transformed sum of triceps and subscapular skinfolds (logSF2); (4) log-transformed sum of triceps, biceps, subscapular and supra-iliac skinfolds (logSF4); (5) BMI; (6) waist circumference; (7) BMI and waist circumference. The adjusted determination coefficient (R² adj) and the root mean squared error (RMSE; kg) were calculated for each model. LogSF4 (R² adj 0.85; RMSE 2.35) and logSF2 (R² adj 0.82; RMSE 2.54) were similarly accurate at predicting PBF and superior to logTSF (R² adj 0.75; RMSE 3.02), logTSF combined with waist circumference (R² adj 0.78; RMSE 2.85), BMI (R² adj 0.62; RMSE 3.73), waist circumference (R² adj 0.58; RMSE 3.89), and BMI combined with waist circumference (R² adj 0.63; RMSE 3.66) (P < 0.001 for all values of R² adj). The finding that logSF4 was only modestly superior to logSF2 and that logTSF was better than BMI and waist circumference at predicting PBF has important implications for paediatric epidemiological studies aimed at disentangling the effect of body fat on health outcomes.

Statistical Modeling of the Eye for Multimodal Treatment Planning for External Beam Radiation Therapy of Intraocular Tumors.

PURPOSE: Ocular anatomy and radiation-associated toxicities provide unique challenges for external beam radiation therapy. For treatment planning, precise modeling of organs at risk and tumor volume are crucial. Development of a precise eye model and automatic adaptation of this model to patients' anatomy remain problematic because of organ shape variability. This work introduces the application of a 3-dimensional (3D) statistical shape model as a novel method for precise eye modeling for external beam radiation therapy of intraocular tumors. METHODS AND MATERIALS: Manual and automatic segmentations were compared for 17 patients, based on head computed tomography (CT) volume scans. A 3D statistical shape model of the cornea, lens, and sclera as well as of the optic disc position was developed. Furthermore, an active shape model was built to enable automatic fitting of the eye model to CT slice stacks. Cross-validation was performed based on leave-one-out tests for all training shapes by measuring dice coefficients and mean segmentation errors between automatic segmentation and manual segmentation by an expert. RESULTS: Cross-validation revealed a dice similarity of 95% ± 2% for the sclera and cornea and 91% ± 2% for the lens. Overall, mean segmentation error was found to be 0.3 ± 0.1 mm. Average segmentation time was 14 ± 2 s on a standard personal computer. CONCLUSIONS: Our results show that the solution presented outperforms state-of-the-art methods in terms of accuracy, reliability, and robustness. Moreover, the eye model shape as well as its variability is learned from a training set rather than by making shape assumptions (eg, as with the spherical or elliptical model). Therefore, the model appears to be capable of modeling nonspherically and nonelliptically shaped eyes.

Change in cognitive errors and coping over the course of brief psychodynamic intervention.

OBJECTIVE: Cognitive change over the course of psychodynamic psychotherapy has been postulated by several models, but has rarely been studied. Based on the adaptive skills model (Badgio, Halperin, & Barber, 1999), it is reasonable to expect that very brief dynamic psychotherapy may be associated with change in coping patterns and cognitive errors (also known as cognitive distortions) y. METHOD: N = 50 outpatients presenting with various psychiatric disorders and undergoing 4 sessions of Brief Psychodynamic Intervention (BPI; Despland, Drapeau, & de Roten, 2005; Despland, Michel, & de Roten, 2010) were included in this naturalistic study (mean age: 31 years; 56% female; all Caucasian). Cognitive errors and coping strategies were assessed using the Cognitive Errors Rating Scale (Drapeau et al., 2008) and Coping Patterns Rating Scale (Perry et al., 2005). These observer rated methods were applied to the verbatim transcriptions of all 4 therapy sessions completed by each patient. RESULTS: Results indicate change in both cognitive errors and coping patterns over the course of BPI, including an increase in the Overall Coping Functioning and a decrease in unhelpful coping processes, such as isolation, which reflects a shift in participant appraisal towards stress appraised as a challenge at the end of treatment. These changes predicted symptom change at the end of treatment. While cognitive errors also changed over the course of BPI, no predictive effect was found with regard to symptom change. CONCLUSIONS: These results are interpreted within the framework of common change principles in psychotherapy. Implications and future research are discussed.

The truncated mean of an asymmetric distribution

This paper investigates a simple procedure to estimate robustly the mean of an asymmetric distribution. The procedure removes the observations which are larger or smaller than certain limits and takes the arithmetic mean of the remaining observations, the limits being determined with the help of a parametric model, e.g., the Gamma, the Weibull or the Lognormal distribution. The breakdown point, the influence function, the (asymptotic) variance, and the contamination bias of this estimator are explored and compared numerically with those of competing estimates.

Aggregation and errors in management accounting

Il est important pour les entreprises de compresser les informations détaillées dans des sets d'information plus compréhensibles. Au chapitre 1, je résume et structure la littérature sur le sujet « agrégation d'informations » en contrôle de gestion. Je récapitule l'analyse coûts-bénéfices que les comptables internes doivent considérer quand ils décident des niveaux optimaux d'agrégation d'informations. Au-delà de la perspective fondamentale du contenu d'information, les entreprises doivent aussi prendre en considération des perspectives cogni- tives et comportementales. Je développe ces aspects en faisant la part entre la comptabilité analytique, les budgets et plans, et la mesure de la performance. Au chapitre 2, je focalise sur un biais spécifique qui se crée lorsque les informations incertaines sont agrégées. Pour les budgets et plans, des entreprises doivent estimer les espérances des coûts et des durées des projets, car l'espérance est la seule mesure de tendance centrale qui est linéaire. A la différence de l'espérance, des mesures comme le mode ou la médiane ne peuvent pas être simplement additionnés. En considérant la forme spécifique de distributions des coûts et des durées, l'addition des modes ou des médianes résultera en une sous-estimation. Par le biais de deux expériences, je remarque que les participants tendent à estimer le mode au lieu de l'espérance résultant en une distorsion énorme de l'estimati¬on des coûts et des durées des projets. Je présente également une stratégie afin d'atténuer partiellement ce biais. Au chapitre 3, j'effectue une étude expérimentale pour comparer deux approches d'esti¬mation du temps qui sont utilisées en comptabilité analytique, spécifiquement « coûts basés sur les activités (ABC) traditionnelles » et « time driven ABC » (TD-ABC). Au contraire des affirmations soutenues par les défenseurs de l'approche TD-ABC, je constate que cette dernière n'est pas nécessairement appropriée pour les calculs de capacité. Par contre, je démontre que le TD-ABC est plus approprié pour les allocations de coûts que l'approche ABC traditionnelle. - It is essential for organizations to compress detailed sets of information into more comprehensi¬ve sets, thereby, establishing sharp data compression and good decision-making. In chapter 1, I review and structure the literature on information aggregation in management accounting research. I outline the cost-benefit trade-off that management accountants need to consider when they decide on the optimal levels of information aggregation. Beyond the fundamental information content perspective, organizations also have to account for cognitive and behavi¬oral perspectives. I elaborate on these aspects differentiating between research in cost accounti¬ng, budgeting and planning, and performance measurement. In chapter 2, I focus on a specific bias that arises when probabilistic information is aggregated. In budgeting and planning, for example, organizations need to estimate mean costs and durations of projects, as the mean is the only measure of central tendency that is linear. Different from the mean, measures such as the mode or median cannot simply be added up. Given the specific shape of cost and duration distributions, estimating mode or median values will result in underestimations of total project costs and durations. In two experiments, I find that participants tend to estimate mode values rather than mean values resulting in large distortions of estimates for total project costs and durations. I also provide a strategy that partly mitigates this bias. In the third chapter, I conduct an experimental study to compare two approaches to time estimation for cost accounting, i.e., traditional activity-based costing (ABC) and time-driven ABC (TD-ABC). Contrary to claims made by proponents of TD-ABC, I find that TD-ABC is not necessarily suitable for capacity computations. However, I also provide evidence that TD-ABC seems better suitable for cost allocations than traditional ABC.

Modeling sequencing errors by combining Hidden Markov models.

Among the largest resources for biological sequence data is the large amount of expressed sequence tags (ESTs) available in public and proprietary databases. ESTs provide information on transcripts but for technical reasons they often contain sequencing errors. Therefore, when analyzing EST sequences computationally, such errors must be taken into account. Earlier attempts to model error prone coding regions have shown good performance in detecting and predicting these while correcting sequencing errors using codon usage frequencies. In the research presented here, we improve the detection of translation start and stop sites by integrating a more complex mRNA model with codon usage bias based error correction into one hidden Markov model (HMM), thus generalizing this error correction approach to more complex HMMs. We show that our method maintains the performance in detecting coding sequences.

I have faith in my thing, you know what I mean? Mixed method elements on individual investments on religious markets.

In contemporary society, religious signification and secular systems mix and influence each other. Holistic conceptions of a world in which man is integrated harmoniously with nature meet representations of a world run by an immanent God. On the market of the various systems, the individual goes from one system to another, following his immediate needs and expectations without necessarily leaving any marks in a meaningful long term system. This article presents the first results of an ongoing research in Switzerland on contemporary religion focusing on (new) paths of socialization of modern that individuals and the various (non-) belief systems that they simultaneously develop

Female residents experiencing medical errors in general internal medicine: a qualitative study.

BACKGROUND: Doctors, especially doctors-in-training such as residents, make errors. They have to face the consequences even though today's approach to errors emphasizes systemic factors. Doctors' individual characteristics play a role in how medical errors are experienced and dealt with. The role of gender has previously been examined in a few quantitative studies that have yielded conflicting results. In the present study, we sought to qualitatively explore the experience of female residents with respect to medical errors. In particular, we explored the coping mechanisms displayed after an error. This study took place in the internal medicine department of a Swiss university hospital. METHODS: Within a phenomenological framework, semi-structured interviews were conducted with eight female residents in general internal medicine. All interviews were audiotaped, fully transcribed, and thereafter analyzed. RESULTS: Seven main themes emerged from the interviews: (1) A perception that there is an insufficient culture of safety and error; (2) The perceived main causes of errors, which included fatigue, work overload, inadequate level of competences in relation to assigned tasks, and dysfunctional communication; (3) Negative feelings in response to errors, which included different forms of psychological distress; (4) Variable attitudes of the hierarchy toward residents involved in an error; (5) Talking about the error, as the core coping mechanism; (6) Defensive and constructive attitudes toward one's own errors; and (7) Gender-specific experiences in relation to errors. Such experiences consisted in (a) perceptions that male residents were more confident and therefore less affected by errors than their female counterparts and (b) perceptions that sexist attitudes among male supervisors can occur and worsen an already painful experience. CONCLUSIONS: This study offers an in-depth account of how female residents specifically experience and cope with medical errors. Our interviews with female residents convey the sense that gender possibly influences the experience with errors, including the kind of coping mechanisms displayed. However, we acknowledge that the lack of a direct comparison between female and male participants represents a limitation while aiming to explore the role of gender.

Separation of free amino acids in human plasma by capillary electrophoresis with laser induced fluorescence: potential for emergency diagnosis of inborn errors of metabolism.

Free amino acids (AAs) in human plasma are derivatized with 3-(4-carboxybenzoyl)quinoline-2-carboxaldehyde (CBQCA) and analyzed by capillary electrophoresis (CE) with laser induced fluorescence (LIF) detection. The labeling procedure is significantly improved over results reported previously. Derivatization can be completed in 40 min, with concentrations as low as 4 x 10(-8) M successfully labeled in favourable cases. Twenty-nine AAs (including 2 internal standards) are identified and can be reproducibly separated in 70 min. Migration time RSD values for 23 of these AAs were calculated and found in the range from 0.5 to 4%. The rapid derivatization procedure and the resolution obtained in the separation are sufficient for a semi-quantitative, emergency diagnosis of several inborn errors of metabolism (IEM). Amino acid profiles for both normal donor plasma samples and plasma samples of patients suffering from phenylketonuria, tyrosinemia, maple syrup urinary disease, hyperornithinemia, and citrullinemia are studied.

Mean platelet volume in the early phase of acute ischemic stroke is not associated with severity or functional outcome.

Background: Previous studies reported an increase of mean platelet volume (MPV) in patients with acute ischemic stroke. However, its correlation with stroke severity has not been investigated. Moreover, studies on the association of MPV with functional outcome yielded inconsistent results. Methods: We included all consecutive ischemic stroke patients admitted to CHUV (Centre Hospitalier Universitaire Vaudois) Neurology Service within 24 h after stroke onset who had MPV measured on admission. The association of MPV with stroke severity (NIHSS score at admission and at 24 h) and outcome (Rankin Scale score at 3 and 12 months) was analyzed in univariate analysis. The chi(2) test was performed to compare the frequency of minor strokes (NIHSS score </=4) and good functional outcome (Rankin Scale score </=2) across MPV quartiles. The ANOVA test was used to compare MPV between stroke subtypes according to the TOAST classification. Student's two-tailed unpaired t test was performed to compare MPV between lacunar and nonlacunar strokes. MPV was generated at admission by the Sysmex XE-2100 automated cell counter (Sysmex Corporation, Kobe, Japan) from EDTA blood samples. Results: There was no significant difference in the frequency of minor strokes (p = 0.46) and good functional outcome (p = 0.06) across MPV quartiles. MPV was not associated with stroke severity or outcome in univariate analysis. There was no significant difference in MPV between stroke subtypes according to the TOAST classification (p = 0.173) or between lacunar and nonlacunar strokes (10.50 +/- 0.91 vs. 10.40 +/- 0.81 fl, p = 0.322). Conclusions: MPV, assessed within 24 h after ischemic stroke onset, is not associated with stroke severity or functional outcome.

Evaluating the influence of spatial uncertainty in locality points for species distributional modeling

1. Species distribution modelling is used increasingly in both applied and theoretical research to predict how species are distributed and to understand attributes of species' environmental requirements. In species distribution modelling, various statistical methods are used that combine species occurrence data with environmental spatial data layers to predict the suitability of any site for that species. While the number of data sharing initiatives involving species' occurrences in the scientific community has increased dramatically over the past few years, various data quality and methodological concerns related to using these data for species distribution modelling have not been addressed adequately. 2. We evaluated how uncertainty in georeferences and associated locational error in occurrences influence species distribution modelling using two treatments: (1) a control treatment where models were calibrated with original, accurate data and (2) an error treatment where data were first degraded spatially to simulate locational error. To incorporate error into the coordinates, we moved each coordinate with a random number drawn from the normal distribution with a mean of zero and a standard deviation of 5 km. We evaluated the influence of error on the performance of 10 commonly used distributional modelling techniques applied to 40 species in four distinct geographical regions. 3. Locational error in occurrences reduced model performance in three of these regions; relatively accurate predictions of species distributions were possible for most species, even with degraded occurrences. Two species distribution modelling techniques, boosted regression trees and maximum entropy, were the best performing models in the face of locational errors. The results obtained with boosted regression trees were only slightly degraded by errors in location, and the results obtained with the maximum entropy approach were not affected by such errors. 4. Synthesis and applications. To use the vast array of occurrence data that exists currently for research and management relating to the geographical ranges of species, modellers need to know the influence of locational error on model quality and whether some modelling techniques are particularly robust to error. We show that certain modelling techniques are particularly robust to a moderate level of locational error and that useful predictions of species distributions can be made even when occurrence data include some error.

Dukes B colorectal cancer: distinct genetic categories and clinical outcome based on proximal or distal tumor location.

PURPOSE: The aim of this study was to determine whether tumor location proximal or distal to the splenic flexure is associated with distinct molecular patterns and can predict clinical outcome in a homogeneous group of patients with Dukes B (T3-T4, N0, M0) colorectal cancer. It has been hypothesized that proximal and distal colorectal cancer may arise through different pathogenetic mechanisms. Although p53 and Ki-ras gene mutations occur frequently in distal tumors, another form of genomic instability associated with defective DNA mismatch repair has been predominantly identified in the proximal colon. To date, however, the clinical usefulness of these molecular characteristics remains unproven. METHODS: A total of 126 patients with a lymph node-negative sporadic colon or rectum adenocarcinoma were prospectively assessed with the endpoint of death by cancer. No patient received either radiotherapy or chemotherapy. p53 protein was studied by immunohistochemistry using DO-7 monoclonal antibody, and p53 and Ki-ras gene mutations were detected by single strand conformation polymorphism assay. RESULTS: During a mean follow-up of 67 months, the overall five-year survival was 70 percent. Nuclear p53 staining was found in 57 tumors (47 percent), and was more frequent in distal than in proximal tumors (55 vs. 21 percent; chi-squared test, P < 0.001). For the whole group, p53 protein expression correlated with poor survival in univariate and multivariate analysis (log-rank test, P = 0.01; hazard ratio = 2.16; 95 percent confidence interval = 1.12-4.11, P = 0.02). Distal colon tumors and rectal tumors exhibited similar molecular patterns and showed no difference in clinical outcome. In comparison with distal colorectal cancer, proximal tumors were found to be statistically significantly different on the following factors: mucinous content (P = 0.008), degree of histologic differentiation (P = 0.012), p53 protein expression, and gene mutation (P = 0.001 and 0.01 respectively). Finally, patients with proximal tumors had a marginally better survival than those with distal colon or rectal cancers (log-rank test, P = 0.045). CONCLUSION: In this series of Dukes B colorectal cancers, p53 protein expression was an independent factor for survival, which also correlated with tumor location. Eighty-six percent of p53-positive tumors were located in the distal colon and rectum. Distal colon and rectum tumors had similar molecular and clinical characteristics. In contrast, proximal neoplasms seem to represent a distinct entity, with specific histopathologic characteristics, molecular patterns, and clinical outcome. Location of the neoplasm in reference to the splenic flexure should be considered before group stratification in future trials of adjuvant chemotherapy in patients with Dukes B tumors.

Understanding the low-temperature limitations to forest growth through calibration of a forest dynamics model with tree-ring data

The sensitivity of altitudinal and latitudinal tree-line ecotones to climate change, particularly that of temperature, has received much attention. To improve our understanding of the factors affecting tree-line position, we used the spatially explicit dynamic forest model TreeMig. Although well-suited because of its landscape dynamics functions, TreeMig features a parabolic temperature growth response curve, which has recently been questioned. and the species parameters are not specifically calibrated for cold temperatures. Our main goals were to improve the theoretical basis of the temperature growth response curve in the model and develop a method for deriving that curve's parameters from tree-ring data. We replaced the parabola with an asymptotic curve, calibrated for the main species at the subalpine (Swiss Alps: Pinus cembra, Larix decidua, Picea abies) and boreal (Fennoscandia: Pinus sylvestris, Betula pubescens, P. abies) tree-lines. After fitting new parameters, the growth curve matched observed tree-ring widths better. For the subalpine species, the minimum degree-day sum allowing, growth (kDDMin) was lowered by around 100 degree-days; in the case of Larix, the maximum potential ring-width was increased to 5.19 mm. At the boreal tree-line, the kDDMin for P. sylvestris was lowered by 210 degree-days and its maximum ring-width increased to 2.943 mm; for Betula (new in the model) kDDMin was set to 325 degree-days and the maximum ring-width to 2.51 mm; the values from the only boreal sample site for Picea were similar to the subalpine ones, so the same parameters were used. However, adjusting the growth response alone did not improve the model's output concerning species' distributions and their relative importance at tree-line. Minimum winter temperature (MinWiT, mean of the coldest winter month), which controls seedling establishment in TreeMig, proved more important for determining distribution. Picea, P. sylvestris and Betula did not previously have minimum winter temperature limits, so these values were set to the 95th percentile of each species' coldest MinWiT site (respectively -7, -11, -13). In a case study for the Alps, the original and newly calibrated versions of TreeMig were compared with biomass data from the National Forest Inventor), (NFI). Both models gave similar, reasonably realistic results. In conclusion, this method of deriving temperature responses from tree-rings works well. However, regeneration and its underlying factors seem more important for controlling species' distributions than previously thought. More research on regeneration ecology, especially at the upper limit of forests. is needed to improve predictions of tree-line responses to climate change further.