Model Diagnostics in the presence of measurement error

The problem of using information available from one variable X to make inferenceabout another Y is classical in many physical and social sciences. In statistics this isoften done via regression analysis where mean response is used to model the data. Onestipulates the model Y = µ(X) +ɛ. Here µ(X) is the mean response at the predictor variable value X = x, and ɛ = Y - µ(X) is the error. In classical regression analysis, both (X; Y ) are observable and one then proceeds to make inference about the mean response function µ(X). In practice there are numerous examples where X is not available, but a variable Z is observed which provides an estimate of X. As an example, consider the herbicidestudy of Rudemo, et al. [3] in which a nominal measured amount Z of herbicide was applied to a plant but the actual amount absorbed by the plant X is unobservable. As another example, from Wang [5], an epidemiologist studies the severity of a lung disease, Y , among the residents in a city in relation to the amount of certain air pollutants. The amount of the air pollutants Z can be measured at certain observation stations in the city, but the actual exposure of the residents to the pollutants, X, is unobservable and may vary randomly from the Z-values. In both cases X = Z+error: This is the so called Berkson measurement error model.In more classical measurement error model one observes an unbiased estimator W of X and stipulates the relation W = X + error: An example of this model occurs when assessing effect of nutrition X on a disease. Measuring nutrition intake precisely within 24 hours is almost impossible. There are many similar examples in agricultural or medical studies, see e.g., Carroll, Ruppert and Stefanski [1] and Fuller [2], , among others. In this talk we shall address the question of fitting a parametric model to the re-gression function µ(X) in the Berkson measurement error model: Y = µ(X) + ɛ; X = Z + η; where η and ɛ are random errors with E(ɛ) = 0, X and η are d-dimensional, and Z is the observable d-dimensional r.v.

Correcting Measurement Error Bias in Interaction Models with Small Samples

Several methods have been suggested to estimate non-linear models with interaction terms in the presence of measurement error. Structural equation models eliminate measurement error bias, but require large samples. Ordinary least squares regression on summated scales, regression on factor scores and partial least squares are appropriate for small samples but do not correct measurement error bias. Two stage least squares regression does correct measurement error bias but the results strongly depend on the instrumental variable choice. This article discusses the old disattenuated regression method as an alternative for correcting measurement error in small samples. The method is extended to the case of interaction terms and is illustrated on a model that examines the interaction effect of innovation and style of use of budgets on business performance. Alternative reliability estimates that can be used to disattenuate the estimates are discussed. A comparison is made with the alternative methods. Methods that do not correct for measurement error bias perform very similarly and considerably worse than disattenuated regression

Convergence Property of the Measurement Gross Error Correction in Power System State Estimation, Using Geometrical Background

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

Evaluation of risk factors for long term changes in refractive error among adult patients at a Dallas Texas private setting A retrospective chart-review study

Background. Over 39.9% of the adult population forty or older in the United States has refractive error, little is known about the etiology of this condition and associated risk factors and their entailed mechanism due to the paucity of data regarding the changes of refractive error for the adult population over time.^ Aim. To evaluate risk factors over a long term, 5-year period, in refractive error changes among persons 43 or older by testing the hypothesis that age, gender, systemic diseases, nuclear sclerosis and baseline refractive errors are all significantly associated with refractive errors changes in patients at a Dallas, Texas private optometric office.^ Methods. A retrospective chart review of subjective refraction, eye health, and self-report health history was done on patients at a private optometric office who were 43 or older in 2000 who had eye examinations both in 2000 and 2005. Aphakic and pseudophakic eyes were excluded as well as eyes with best corrected Snellen visual acuity of 20/40 and worse. After exclusions, refraction was obtained on 114 right eyes and 114 left eyes. Spherical equivalent (sum of sphere + ½ cylinder) was used as the measure of refractive error.^ Results. Similar changes in refractive error were observed for the two eyes. The 5-year change in spherical power was in a hyperopic direction for younger age groups and in a myopic direction for older subjects, P<0.0001. The gender-adjusted mean change in refractive error in right eyes of persons aged 43 to 54, 55 to 64, 65 to 74, and 75 or older at baseline was +0.43D, +0.46 D, -0.09 D, and -0.23D, respectively. Refractive change was strongly related to baseline nuclear cataract severity; grades 4 to 5 were associated with a myopic shift (-0.38 D, P< 0.0001). The mean age-adjusted change in refraction was +0.27 D for hyperopic eyes, +0.56 D for emmetropic eyes, and +0.26 D for myopic eyes.^ Conclusions. This report has documented refractive error changes in an older population and confirmed reported trends of a hyperopic shift before age 65 and a myopic shift thereafter associated with the development of nuclear cataract.^

Slope estimation in a parametric measurement error model: A simulation study

In regression analysis, covariate measurement error occurs in many applications. The error-prone covariates are often referred to as latent variables. In this proposed study, we extended the study of Chan et al. (2008) on recovering latent slope in a simple regression model to that in a multiple regression model. We presented an approach that applied the Monte Carlo method in the Bayesian framework to the parametric regression model with the measurement error in an explanatory variable. The proposed estimator applied the conditional expectation of latent slope given the observed outcome and surrogate variables in the multiple regression models. A simulation study was presented showing that the method produces estimator that is efficient in the multiple regression model, especially when the measurement error variance of surrogate variable is large.^

Mesozooplankton investigations in Bornholm Basin, Baltic Sea

Mesozooplankton production was estimated by using a new sampling technique and two alternative calculation methods. In essence, production estimates are based on significantly higher abundances. The contribution of juvenile stages to copepod and fish dynamics was generally low, so that the omission of juvenile stages in budgets will result in a small error. The situations reported in this study present a unique food web szenario, which in detail, however, was strongly dependent on methodology. Furthermore relations between trophic levels were considered with respect to vertical distribution.

Truncation error estimation in the Discontinuous Galerkin Spectral Element Method

En esta tesis, el método de estimación de error de truncación conocido como restimation ha sido extendido de esquemas de bajo orden a esquemas de alto orden. La mayoría de los trabajos en la bibliografía utilizan soluciones convergidas en mallas de distinto refinamiento para realizar la estimación. En este trabajo se utiliza una solución en una única malla con distintos órdenes polinómicos. Además, no se requiere que esta solución esté completamente convergida, resultando en el método conocido como quasi-a priori T-estimation. La aproximación quasi-a priori estima el error mientras el residuo del método iterativo no es despreciable. En este trabajo se demuestra que algunas de las hipótesis fundamentales sobre el comportamiento del error, establecidas para métodos de bajo orden, dejan de ser válidas en esquemas de alto orden, haciendo necesaria una revisión completa del comportamiento del error antes de redefinir el algoritmo. Para facilitar esta tarea, en una primera etapa se considera el método conocido como Chebyshev Collocation, limitando la aplicación a geometrías simples. La extensión al método Discontinuouos Galerkin Spectral Element Method presenta dificultades adicionales para la definición precisa y la estimación del error, debidos a la formulación débil, la discretización multidominio y la formulación discontinua. En primer lugar, el análisis se enfoca en leyes de conservación escalares para examinar la precisión de la estimación del error de truncación. Después, la validez del análisis se demuestra para las ecuaciones incompresibles y compresibles de Euler y Navier Stokes. El método de aproximación quasi-a priori r-estimation permite desacoplar las contribuciones superficiales y volumétricas del error de truncación, proveyendo información sobre la anisotropía de las soluciones así como su ratio de convergencia con el orden polinómico. Se demuestra que esta aproximación quasi-a priori produce estimaciones del error de truncación con precisión espectral. ABSTRACT In this thesis, the τ-estimation method to estimate the truncation error is extended from low order to spectral methods. While most works in the literature rely on fully time-converged solutions on grids with different spacing to perform the estimation, only one grid with different polynomial orders is used in this work. Furthermore, a non timeconverged solution is used resulting in the quasi-a priori τ-estimation method. The quasi-a priori approach estimates the error when the residual of the time-iterative method is not negligible. It is shown in this work that some of the fundamental assumptions about error tendency, well established for low order methods, are no longer valid in high order schemes, making necessary a complete revision of the error behavior before redefining the algorithm. To facilitate this task, the Chebyshev Collocation Method is considered as a first step, limiting their application to simple geometries. The extension to the Discontinuous Galerkin Spectral Element Method introduces additional features to the accurate definition and estimation of the error due to the weak formulation, multidomain discretization and the discontinuous formulation. First, the analysis focuses on scalar conservation laws to examine the accuracy of the estimation of the truncation error. Then, the validity of the analysis is shown for the incompressible and compressible Euler and Navier Stokes equations. The developed quasi-a priori τ-estimation method permits one to decouple the interfacial and the interior contributions of the truncation error in the Discontinuous Galerkin Spectral Element Method, and provides information about the anisotropy of the solution, as well as its rate of convergence in polynomial order. It is demonstrated here that this quasi-a priori approach yields a spectrally accurate estimate of the truncation error.

Evaluation of follicular lymphoid depletion in the Bursa of Fabricius: an alternative methodology using digital image analysis and artificial neural networks

Fifty Bursa of Fabricius (BF) were examined by conventional optical microscopy and digital images were acquired and processed using Matlab® 6.5 software. The Artificial Neuronal Network (ANN) was generated using Neuroshell® Classifier software and the optical and digital data were compared. The ANN was able to make a comparable classification of digital and optical scores. The use of ANN was able to classify correctly the majority of the follicles, reaching sensibility and specificity of 89% and 96%, respectively. When the follicles were scored and grouped in a binary fashion the sensibility increased to 90% and obtained the maximum value for the specificity of 92%. These results demonstrate that the use of digital image analysis and ANN is a useful tool for the pathological classification of the BF lymphoid depletion. In addition it provides objective results that allow measuring the dimension of the error in the diagnosis and classification therefore making comparison between databases feasible.

Knowledge acquisition by networks of interacting agents in the presence of observation errors

In this work we investigate knowledge acquisition as performed by multiple agents interacting as they infer, under the presence of observation errors, respective models of a complex system. We focus the specific case in which, at each time step, each agent takes into account its current observation as well as the average of the models of its neighbors. The agents are connected by a network of interaction of Erdos-Renyi or Barabasi-Albert type. First, we investigate situations in which one of the agents has a different probability of observation error (higher or lower). It is shown that the influence of this special agent over the quality of the models inferred by the rest of the network can be substantial, varying linearly with the respective degree of the agent with different estimation error. In case the degree of this agent is taken as a respective fitness parameter, the effect of the different estimation error is even more pronounced, becoming superlinear. To complement our analysis, we provide the analytical solution of the overall performance of the system. We also investigate the knowledge acquisition dynamic when the agents are grouped into communities. We verify that the inclusion of edges between agents (within a community) having higher probability of observation error promotes the loss of quality in the estimation of the agents in the other communities.

The impact of measurement errors in the identification of regulatory networks

Background: There are several studies in the literature depicting measurement error in gene expression data and also, several others about regulatory network models. However, only a little fraction describes a combination of measurement error in mathematical regulatory networks and shows how to identify these networks under different rates of noise. Results: This article investigates the effects of measurement error on the estimation of the parameters in regulatory networks. Simulation studies indicate that, in both time series (dependent) and non-time series (independent) data, the measurement error strongly affects the estimated parameters of the regulatory network models, biasing them as predicted by the theory. Moreover, when testing the parameters of the regulatory network models, p-values computed by ignoring the measurement error are not reliable, since the rate of false positives are not controlled under the null hypothesis. In order to overcome these problems, we present an improved version of the Ordinary Least Square estimator in independent (regression models) and dependent (autoregressive models) data when the variables are subject to noises. Moreover, measurement error estimation procedures for microarrays are also described. Simulation results also show that both corrected methods perform better than the standard ones (i.e., ignoring measurement error). The proposed methodologies are illustrated using microarray data from lung cancer patients and mouse liver time series data. Conclusions: Measurement error dangerously affects the identification of regulatory network models, thus, they must be reduced or taken into account in order to avoid erroneous conclusions. This could be one of the reasons for high biological false positive rates identified in actual regulatory network models.

Photochemistry of anthocyanins and their biological role in plant tissues

Anthocyanins, the major red, purple, and blue pigments of plants, absorb visible as well as UV radiation and are effective antioxidants and scavengers of active oxygen species. In plant leaves, one of the functional roles proposed for anthocyanins is protection of the photosynthetic apparatus from the effects of excess incident visible or UV-B radiation and photooxidative stress. In essence, a photoprotective role requires that the excited singlet states of both complexed and uncomplexed anthocyanins deactivate back to the ground state so quickly that intersystem crossing, photoreaction, and diffusion-controlled quenching processes cannot compete. Studies of the photochemical properties of synthetic analogs of anthocyanins and of several naturally occurring anthocyanins show that this is indeed the case, uncomplexed anthocyanins decaying back to the ground state via fast (subnanosecond) excited-state proton transfer (ESPT) and anthocyanin-copigment complexes by fast (sub-picosecond) charge-transfer-mediated internal conversion.

Friction loss correction in flowing well discharge tests

Artesian confined aquifers do not need pumping energy, and water from the aquifer flows naturally at the wellhead. This study proposes correcting the method for analyzing flowing well tests presented by Jacob and Lohman (1952) by considering the head losses due to friction in the well casing. The application of the proposed correction allowed the determination of a transmissivity (T = 411 m(2)/d) and storage coefficient (S = 3 x 10(-4)) which appear to be representative for the confined Guarani Aquifer in the study area. Ignoring the correction due to head losses in the well casing, the error in transmissivity evaluation is about 18%. For the storage coefficient the error is of 5 orders of magnitude, resulting in physically unacceptable value. The effect of the proposed correction on the calculated radius of the cone of depression and corresponding well interference is also discussed.

Outcomes of transient evoked otoacoustic emission testing in 6-year-old school children: A comparison with pure tone screening and tympanometry

Objectives: (1) To establish test performance measures for Transient Evoked Otoacoustic Emission testing of 6-year-old children in a school setting; (2) To investigate whether Transient Evoked Otoacoustic Emission testing provides a more accurate and effective alternative to a pure tone screening plus tympanometry protocol. Methods: Pure tone screening, tympanometry and transient evoked otoacoustic emission data were collected from 940 subjects (1880 ears), with a mean age of 6.2 years. Subjects were tested in non-sound-treated rooms within 22 schools. Receiver operating characteristics curves along with specificity, sensitivity, accuracy and efficiency values were determined for a variety of transient evoked otoacoustic emission/pure tone screening/tympanometry comparisons. Results: The Transient Evoked Otoacoustic Emission failure rate for the group was 20.3%. The failure rate for pure tone screening was found to be 8.9%, whilst 18.6% of subjects failed a protocol consisting of combined pure tone screening and tympanometry results. In essence, findings from the comparison of overall Transient Evoked Otoacoustic Emission pass/fail with overall pure tone screening pass/fail suggested that use of a modified Rhode Island Hearing Assessment Project criterion would result in a very high probability that a child with a pass result has normal hearing (true negative). However, the hit rate was only moderate. Selection of a signal-to-noise ratio (SNR) criterion set at greater than or equal to 1 dB appeared to provide the best test performance measures for the range of SNR values investigated. Test performance measures generally declined when tympanometry results were included, with the exception of lower false alarm rates and higher positive predictive values. The exclusion of low frequency data from the Transient Evoked Otoacoustic Emission SNR versus pure tone screening analysis resulted in improved performance measures. Conclusions: The present study poses several implications for the clinical implementation of Transient Evoked Otoacoustic Emission screening for entry level school children. Transient Evoked Otoacoustic Emission pass/fail criteria will require revision. The findings of the current investigation offer support to the possible replacement of pure tone screening with Transient Evoked Otoacoustic Emission testing for 6-year-old children. However, they do not suggest the replacement of the pure tone screening plus tympanometry battery. (C) 2001 Elsevier Science Ireland Ltd. All rights reserved.

Recent advances in the preparation and utilization of carbon nanotubes for hydrogen storage

Recent progress in the production, purification, and experimental and theoretical investigations of carbon nanotubes for hydrogen storage are reviewed. From the industrial point of view, the chemical vapor deposition process has shown advantages over laser ablation and electric-arc-discharge methods. The ultimate goal in nanotube synthesis should be to gain control over geometrical aspects of nanotubes, such as location and orientation, and the atomic structure of nanotubes, including helicity and diameter. There is currently no effective and simple purification procedure that fulfills all requirements for processing carbon nanotubes. Purification is still the bottleneck for technical applications, especially where large amounts of material are required. Although the alkali-metal-doped carbon nanotubes showed high H-2 Weight uptake, further investigations indicated that some of this uptake was due to water rather than hydrogen. This discovery indicates a potential source of error in evaluation of the storage capacity of doped carbon nanotubes. Nevertheless, currently available single-wall nanotubes yield a hydrogen uptake value near 4 wt% under moderate pressure and room temperature. A further 50% increase is needed to meet U.S. Department of Energy targets for commercial exploitation. Meeting this target will require combining experimental and theoretical efforts to achieve a full understanding of the adsorption process, so that the uptake can be rationally optimized to commercially attractive levels. Large-scale production and purification of carbon nanotubes and remarkable improvement of H-2 storage capacity in carbon nanotubes represent significant technological and theoretical challenges in the years to come.

Genotype and environment effects on dynamics of harvest index during grain filling in sorghum

An approach based on a linear rate of increase in harvest index (141) with time after anthesis has been used as a simple means-to predict grain growth and yield in many crop simulation models. When applied to diverse situations, however, this approach has been found to introduce significant error in grain yield predictions. Accordingly, this study was undertaken to examine the stability of the HI approach for yield prediction in sorghum [Sorghum bicolor (L.) Moench]. Four field experiments were conducted under nonlimiting water. and N conditions. The experiments were sown at times that ensured a broad range in temperature and radiation conditions. Treatments consisted of two population densities and three genotypes varying in maturity. Frequent sequential harvests were used to monitor crop growth, yield, and the dynamics of 111. Experiments varied greatly in yield and final HI. There was also a tendency for lower HI with later maturity. Harvest index dynamics also varied among experiments and, to a lesser extent, among treatments within experiments. The variation was associated mostly with the linear rate of increase in HI and timing of cessation of that increase. The average rate of HI increase was 0.0198 d(-1), but this was reduced considerably (0.0147) in one experiment that matured in cool conditions. The variations found in IN dynamics could be largely explained by differences in assimilation during grain filling and remobilization of preanthesis assimilate. We concluded that this level of variation in HI dynamics limited the general applicability of the HI approach in yield prediction and suggested a potential alternative for testing.