Browsing by Subject "Bayesian"

Now showing 1 - 16 of 16

Application of the bifactor model to computerized adaptive testing.
(2011-01) Seo, Dong Gi
Most CAT has been studied under the framework of unidimensional IRT. However, many psychological variables are multidimensional and might benefit from using a multidimensional approach to CAT. In addition, a number of psychological variables (e.g., quality of life, depression) can be conceptualized as being consistent with a bifactor model (Holzinger & Swineford, 1937) in which there is a general dimension and some number of subdomains with each item loading on only one of those domains. The present study extended the work on the bifactor CAT of Weiss & Gibbons (2007) in comparison to a fully multidimensional bifactor method using multidimensional maximum likelihood estimation and Bayesian estimation for the bifactor model (MBICAT algorithm). Although Weiss and Gibbons applied the bifactor model to CAT (BICAT algorithm), their methods for item selection and scoring were based on unidimensional IRT methods. Therefore, this study investigated a fully multidimensional bifactor CAT algorithm using simulated data. The MBICAT algorithm was compared to the two BICAT algorithms under three different factors: the number of group factors, the group factor discrimination condition, and estimation method. A fixed- test length was used as the termination criterion for the CATs for Study 1. The accuracy of estimates using the BICAT algorithm and the MBICAT algorithm was evaluated with the correlation between true and estimated scores , the root mean square error (RMSE), and the observed standard error (OSE). Two termination criteria (OSE = .50 and .55) were used to investigate efficiency of the MBICAT for Study 2. This study demonstrated that the MBICAT algorithm worked well when latent scores on the secondary dimension were estimated properly. Although the MBICAT algorithm did not improve the accuracy and efficiency for the general factor scores compared to two BICAT algorithms, the MBICAT showed an improvement of the accuracy and efficiency for the group factors. In the two BICAT algorithms, the use of differential entry on the group factors did not make a difference compared to initial item at trait of 0 for both the general factor and group factor scales (Gibbons, et al., 2008) in terms of accuracy and efficiency.
A Bayesian approach to custer sampling
(2012-12) Soma, Michael Blohm
Cluster sampling is a survey design that is commonly used when a simple random sample may be too costly or inefficient to implement for a population. The idea of a cluster sample is that a population can be divided into groups called clusters. The usual cluster sample consists of sampling n of the N clusters within a population. Within these clusters, a further sample may be taken of secondary sampling units. When a further sample is taken within the primary sample, this is called a two-stage cluster sampling design. Design-based estimates of mean and total are well-established for two-stage cluster samples. For certain populations, however, a Bayesian approach may be preferred for ease of interpretation or to estimate a parameter not well-developed in the design-based literature. When a Bayesian approach is used, a commonly desired property of the model is that the sampled and unsampled units are exchangeable. In this way, the data and not the model is influencing population estimates. A Bayesian approach in which the sampled and unsampled units are exchangeable is the Bayes Urn model. In this dissertation, we will show how the Bayes Urn model can provide admissible estimators under any finite parameter space. Using simulation studies for a variety of two-stage cluster samples, we will show how the Bayes Urn estimates are similar to the standard design-based estimates. As an extension, we will use the Bayes Urn model to incorporate auxiliary information for two-stage cluster samples. For both of these cases we will compare the Bayes Urn estimate of the population total to the standard design-based estimates as well as two standard Bayesian estimates. In the case when auxiliary information is available for a two-stage cluster sample, there is no well-developed Bayesian approach in the literature. In addition to population total, we will also consider estimates of cluster median, in which no design-based approach is well-developed in the literature. Through these simulation studies, the admissibility proof, and the flexible properties inherent of the Bayes Urn model, we will see that this approach can provide useful estimations in complex designs beyond the two-stage cluster sample design.
Bayesian approach to Phase II statistical process control for time series
(2013-04) Zhou, Tianyang
In statistical process control (SPC) problems, in-control values of parameters are required by traditional approaches. However this requirement is not realistic. New methods based on the change point model have been developed to avoid this requirement. The existing change-point methods are restricted to independent identically distributed observations, ignoring the numerous settings in which process readings are serially correlated. Furthermore, these frequentist methods are unable to make use of prior imperfect information on the parameters. In my research, I propose a Bayesian approach to the online SPC based on the change point model in an ARMA process. This approach accommodates serially correlated data, and also provides a coherent way of incorporating prior information on parameters.
Bayesian hierarchical modeling for adaptive incorporation of historical information In clinical trials.
(2010-08) Hobbs, Brian Paul
Bayesian clinical trial designs offer the possibility of a substantially reduced sample size, increased statistical power, and reductions in cost and ethical hazard. However when prior and current information conflict, Bayesian methods can lead to higher than expected Type I error, as well as the possibility of a costlier and lengthier trial. We develop several models that allow for the commensurability of the information in the historical and current data to determine how much historical information is used. First, we propose methods for univariate Gaussian data and provide an example analysis of data from two successive colon cancer trials that illustrates a linear models extension of our adaptive borrowing approach. Next, we extend the general method to linear and linear mixed models as well as generalized linear and generalized linear mixed models. We also provide two more sample analyses using the colon cancer data. Finally, we consider the effective historical sample size of our adaptive method for the case when historical data is available only for the concurrent control arm, and propose "optimal" use of new patients in the current trial using an adaptive randomization scheme that is balanced with respect to the amount of incorporated historical information. The approach is then demonstrated using data from a trial comparing antiretroviral strategies in HIV-1-infected persons. Throughout the thesis we present simulation studies that compare frequentist operating characteristics and highlight the advantages of our adaptive borrowing methods.
Bayesian Modeling of Multi-Source Multi-Way Data
(2023-11) Kim, Jonathan
Biomedical research often involves data collected from multiple sources and these sources often have a multi-way (i.e.. multidimensional tensor) structure. Existing methods that can accommodate multi-source or multi-way data have various limitations on the exact structure of the data they are able to accommodate and in the type of predictions, if any, they are able to produce. Furthermore, few of these methods are able to handle data that are simultaneously multi-source and multi-way. We first introduce two such multi-source and multi-way datasets of molecular and hematological data from multiple sources, each measured over multiple developmental time points and in multiple tissues, as predictors of early-life iron deficiency (ID) in a rhesus monkey model. We describe preliminary analyses that were conducted on these datasets using existing methods. We then develop a Bayesian linear model that can perform prediction on a binary or continuous outcome and can accommodate data that are both multi-source and multi-way. We use a linear model with a low-rank structure on the coefficients to capture multi-way dependence and model the variance of the coefficients separately across each source to infer their relative contributions. Conjugate priors facilitate an efficient Gibbs sampling algorithm for posterior inference, assuming a continuous outcome with normal errors or a binary outcome with a probit link. Simulations demonstrate that our model performs as expected in terms of misclassification rates and correlation of estimated coefficients with true coefficients, with large gains in performance by incorporating multi-way structure and modest gains when accounting for differing signal sizes across the different sources. Moreover, it provides robust classification of ID monkeys for one of our motivating datasets. Finally, we propose a flexible method called Bayesian regression on numerous tensors (BRONTe) that can predict a continuous or binary outcome from data that are collected from an arbitrary number of sources with multi-way tensor structures of arbitrary, not necessarily equal, orders. Additionally, BRONTe is able to accommodate data where some sources partially share features within a dimension. Simulations show BRONTe to perform well at prediction when the data sources are of unequal dimensions. In an application to our other motivating dataset on multi-way measures of metabolomics and hematology parameters, BRONTe was capable of robust classification of early-life iron deficiency.
Bayesian Models for Analyzing Worker Exposure to Airborne Chemicals During the Deepwater Horizon Oil Spill Cleanup and Response
(2017-07) Groth, Caroline
In April 2010, the Deepwater Horizon oil rig caught fire and sank, sending approximately 5 million barrels of oil into the Gulf of Mexico over the ensuing 3 months. Thousands of workers were involved in the response and cleanup efforts. Many harmful chemicals were released into the air from crude oil, including total hydrocarbons (THC), benzene, toluene, ethylbenzene, xylene, hexane (BTEXH), and volatile organic compounds (VOCs). NIEHS's GuLF STUDY investigators are estimating the exposures the workers experienced related to their response and cleanup work and evaluating associations between the exposures and detrimental health outcomes. My research focuses on developing statistical methods to quantify airborne chemical exposures in response to this event and to other settings in environmental health. Factors complicating the exposure estimation include analytical method and data collection limitations. All analytical methods used to measure chemical concentrations have a limit of detection (LOD), or a threshold below which exposure cannot be detected with the analytical method (measurements below the LOD are called censored measurements). However, even these low exposures must be assessed to provide accurate estimates of exposure. Similarly, due to the scope of this event, it was not possible to take measurements in all scenarios where workers were involved in the response. Therefore, we must develop methods that allow us to estimate exposures under these limitations. I introduce a strategy that uses chemical linear relationships to inform exposure estimates. We describe a Bayesian linear model for quantifying exposure while accounting for censoring in both a chemical predictor and a response. We further expand this model to quantify exposure in multiple EGs. Then, I describe a multivariate Bayesian linear model used to quantify exposures under various amounts of LOD censoring in the chemical response and multiple chemical predictors. We assess our model's performance against simpler models at a variety of censoring levels using WAIC. We apply our model to assess vapor exposures from measurements of volatile substances in crude oil on the Ocean Intervention III taken during the Deepwater Horizon oil spill response and cleanup. Next, I explain how we used a database of over 26 million VOC measurements to supplement information in THC and BTEXH. I discuss the methods we used to convert this large VOC database into a exposure metric that could be compared with THC exposure. Then, I describe how we used the VOC exposure metrics to estimate THC and BTEXH exposure when VOC information was available but THC/BTEXH measurements were unavailable. Finally, I expand the Bayesian linear framework to a spatial setting that allows us to estimate exposure for particular areas in the Gulf of Mexico while accounting for values below LOD in both the response and predictor of interest. We also investigate imputation strategies designed to allow us to estimate exposure to our chemical predictor (providing input to our model) so we can better estimate our chemical response. I conclude with a brief description of our current investigation of environmental exposures during the Deepwater Horizon response and cleanup efforts.
Data and code for "Optimizing early detection strategies: defining the effective attraction radius of attractants for emerald ash borer"
(2021-04-30) Wittman, Jacob T; Silk, Peter; Parker, Katie; Aukema, Brian H; wittm094@umn.edu; Wittman, Jacob T
From the publication: 1. Adult emerald ash borers are attracted to green prism traps baited with the ash host volatile (3Z)-hexenol and the sex pheromone of emerald ash borer (3Z)-dodecen-12-olide [(3Z)-lactone]. Quantifying the heretofore unknown range of attraction of such traps would help optimize deployment strategies for early detection. 2. Examining trap captures of traps deployed in pairs at variable distances offers insight into the range of attraction. Recent work has shown the range of attraction can be estimated as half the intertrap distance at which trap catch begins to decrease, which should occur when proximate traps overlap their respective attractive ranges. 3. We illustrate estimation of the traps' attractive range for emerald ash borer using ninety-eight baited green prism traps deployed in pairs, one trap per tree, in an urban park in Saint Paul, Minnesota, USA in summer 2020. 4. We estimate attractive range by fitting a logistic model to trap catch data using Bayesian inferential methods and describe advantages thereof. 5. The attractive range of these baited traps was estimated to be between 16 -- 73m, with a median of 28m. We recommend that green prism traps baited with these semiochemicals be placed 25 - 35m apart near high-risk entry points.
Estimating a noncompensatory IRT model using a modified metropolis algorithm.
(2009-12) Babcock, Benjamin Grant Eugene
Two classes of dichotomous multidimensional item response theory (MIRT) models, compensatory and noncompensatory, are reviewed. After a review of the literature, it is concluded that relatively little research has been conducted with the noncompensatory class of models. A monte-carlo simulation study was conducted exploring the estimation of a 2-parameter noncompensatory IRT model. The estimation method used was a modification of the Metropolis-Hastings algorithm that used multivariate prior distributions to help determine whether or not a newly sampled value was retained or rejected. Results showed that the noncompensatory model required a sample size of 4,000 people, 6 unidimensional items per dimension, and latent traits that are not highly correlated, for acceptable item parameter estimation using the modified Metropolis method. It is then argued that the noncompensatory model might not warrant further research due to the great requirements for acceptable estimation. The multidimensional interactive IRT model (MIIM) is proposed, which is more flexible than previous multidimensional models and explicitly accounts for correlated latent traits by using an interaction term within the logit. Item response surfaces for the MIIM model can be shaped either like compensatory or noncompensatory IRT model response surfaces.
Gaussian Processes in Semi-Parametric Models
(2023-06) Thompson, Marten
Gaussian processes provide a flexible, non-parametric prior for function estimation. We investigate the applicability of Gaussian processes in semi-parametric models to relax otherwise restrictive assumptions. Our first application of this perspective is climate time series, where we see both the advantage of Gaussian processes in semi-parametric models as well as their computational restrictions. Next, we use Gaussian processes to relax the assumed error distribution of traditional small area models. Finally, we turn our attention to stripping away assumptions on Gaussian processes themselves: can data be used to inform their parameterization? We detail our work on each of these problems and provide software for future researchers.
Incorporation of Covariates in Bayesian Piecewise Growth Mixture Models
(2022-12) Lamm, Rik
The Bayesian Covariate Influenced Piecewise Growth Mixture Model (CI-PGMM) is an extension of the Piecewise Growth Mixture Model (PGMM, Lock et al., 2018) with the incorporation of covariates. This was done by using a piecewise nonlinear trajectory over time, meaning that the slope has a point where the trajectory changes, called a knot. Additionally, the outcome data belong to two or more latent classes with their own mean trajectories, referred to as a mixture model. Covariates were incorporated into the model in two ways. The first was influencing the outcome variable directly, explaining additional random error variance. The second is the influence of the covariates on the class membership directly with the use of multinomial logistic regression. Both uses of covariates can potentially influence the class memberships and along with that, the trajectories and locations of the knot(s). This additional explanation of class memberships and trajectories can provide information on how individuals change, who is likely to belong in certain unknown classes, and how these class memberships can affect when the rapid change of a knot will happen. The model is shown to be appropriate and effective using two steps. First, a real data application using the National Longitudinal Survey of Youth is used to show the motivation for the model. This dataset measures income over time each year for individuals following high school. Covariates of sex and dropout status were used in the class predictive logistic regression model. This resulted in a two-class solution showing effective use of the covariates with the logistic regression coefficients drastically affecting the class memberships. The second step is using a simulation after the motivating real data application. Pilot studies were used to show if the model was suitable for a full simulation using the coefficients from the real data example as a basis for the data generation. Four pilot studies were performed, and reasonable estimates were found for the full simulation. The conditions were set up with a two class model. One class containing one knot, and the second class as a linear slope. Two class predictive covariates and one outcome predictive covariate were used. A full simulation with 200 generated datasets was performed with manipulated conditions being error variance, sample size, model type, and class probability for a 3x3x3x2 model with 54 total conditions. Outcome measures of convergence, average relative bias, RMSE, and coverage rate were used to show suitability of the model. The simulation showed the use for the CI-PGMM was stable and accurate for multiple conditions. Sample size and model type were the most impactful predictors of appropriate model use. All outcome measures were worse for the small sample sizes and became more accurate when the sample sizes were larger. Also, the simpler models showed less bias and better convergence. However, these differences are smaller when the sample size is sufficiently large. These findings were supported with multi-factor ANOVA comparing simulation conditions. Use of the CI-PGMM in the real data example and the full simulation allowed for incorporation of covariates when appropriate. I show that model complexity can lead to issues of lower convergence, thus the model should only be used when appropriate and the sample size is sufficiently large. When used, however, the model can shed light on associations between covariates, class memberships, and locations of knots that were previously unavailable.
Robust Variance Component Models and Powerful Variable Selection Methods for Addressing Missing Heritability
(2018-08) Arbet, Jaron
The development of a complex human disease is an intricate interplay of genetic and environmental factors. Broadly speaking, “heritability” is defined as the proportion of total trait variance due to genetic factors within a given population. Over the past 50 years, studies involving monozygotic and dizygotic twins have estimated the heritability of over 17,800 human traits [1]. Genetic association studies that measure thousands to millions of genetic “markers” have attempted to determine the exact markers that explain a given trait’s heritability. However, often the identified set of “statistically-significant” markers fails to explain more than 10% of the estimated heritability of a trait [2], which has been defined as the “missing heritability” problem [3][4]. “Missing heritability’ implies that many genetic markers that contribute to disease risk are still waiting to be discovered. Identification of the exact genetic markers associated with a disease is important for the development of pharmaceutical drugs that may target these markers (see [5] for recent examples). Additionally, “missing heritability” may imply that we are inaccurately estimating heritability in the first place [3, 4, 6], thus motivating the development of more robust models for estimating heritability. This dissertation focuses on two objectives that attempt to address the missing heritability problem: (1) develop a more robust framework for estimating heritability; and (2) develop powerful association tests in attempt to find more genetic markers associated with a given trait. Specifically: in Chapter 2, robust variance component models are developed for estimating heritability in twin studies using second-order generalized estimating equations (GEE2). We demonstrate that GEE2 can improve coverage rates of the true heritability parameter for non-normally distributed outcomes, and can easily incorporate both mean and variance-level covariate effects (e.g. let heritability vary by sex or age). In Chapter 3, penalized regression is used to jointly model all genetic markers. It is demonstrated that jointly modeling all markers can improve power to detect individual associated markers compared to conventional methods that model each marker “one-at-a-time.” Chapter 4 expands on this work by developing a more flexible nonparametric Bayesian variable selection model that can account for non-linear or non-additive effects, and can also test biologically meaningful groups of markers for an association with the outcome. We demonstrate how the nonparametric Bayesian method can detect markers with complex association structures that more conventional models might miss.
Searching, Clustering and Regression on non-Euclidean Spaces
(2015-08) Wang, Xu
This dissertation considers three common tasks (e.g., searching, clustering, regression) over Riemannian spaces. The first task considers the problem of efficiently deciding which of a database of subspaces is most similar to a given input query. Motivated by applications in recognition, image retrieval and optimization, there has been significant recent interest in this problem. Current approaches to this problem have poor scaling in high dimensions, and may not guarantee sublinear query complexity. We present a new approach to approximate nearest subspace search, based on a simple, new locality sensitive hash for subspaces. For the second task, we advocates a novel framework for segmenting a dataset in a Riemannian manifold into clusters lying around low-dimensional submanifolds. This clustering problem is useful for applications such as action identification, dynamic texture clustering, brain fiber segmentation, and clustering of deformed images. The proposed clustering algorithm constructs an affinity matrix by exploiting the geometry and then applies spectral clustering. Theoretical guarantees are established for a variant of the algorithm. To avoid complication, these guarantees assume that the submanifolds are geodesic. Extensive validation on synthetic and real data demonstrates the resiliency of the proposed method against deviations from the theoretical model as well as its superior performance over state-of-the-art techniques. In the third task, we proposes a novel framework for manifold-valued regression and establishes its consistency as well as its contraction rate for a particular setting. Our setting assumes a predictor with values in the interval [0,1] and response with values in a compact Riemannian manifold. This setting is useful for applications such as modeling dynamic scenes or shape deformations, where the visual scene or the deformed objects can be modeled by a manifold. The proposed framework uses the heat kernel on manifolds as an averaging procedure. It directly generalizes the use of the Gaussian kernel in vector-valued regression problems. In order to avoid explicit dependence on estimates of the heat kernel, we follow a Bayesian setting, where Brownian motion induces a prior distribution on the space of continuous functions. We study the posterior consistency and contraction rate of the discrete and continuous Brownian motion priors.
Some Estimates of Farmers' Utility Functions
(Minnesota Agricultural Experiment Station, 1982) Hildreth, Clifford; Knowles, Glenn J.
Survey sampling and multiple stratifications
(2013-09) Zimmerman, Patrick Lennon Kendall
In survey sampling, stratied random sampling and post-stratification can increase the precision of estimation. In some cases, however, there may be multiple ways to stratify a population. We present a method, based on a non-informative Bayesian approach, that uses a finite mixture model to incorporate information from each stratification into estimation. This approach works well when the response variable is categorical or discrete,and for some non-response types of problems. We provide the theoretical basis for our method, present some simulation results, discuss various extensions, and define some software that implements the method.
Systematics of the family Polycentropodidae (Inseecta:Trichoptera: Psychomyioidea) and taxonomic revisions of New World Polyplectropus Ulmer.
(2009-01) Chamorro, Maria Lourdes
The monophyly and phylogenetic relationships of subfamilies and genera traditionally classified in Polycentropodidae Ulmer, 1903, one of the most diverse families in the suborder Annulipalpia, with more than 700 species in 3 subfamilies, were tested. Particular emphasis was placed on testing the monophyly of the cosmopolitan genus Polyplectropus. Larval information is unknown for 46% of the taxa included in this study. To understand the effects of including characters with large sets of missing data, three alternative datasets [TOTAL (all available data for all taxa)= 49 ingroup taxa, 122 characters (including highly incomplete characters); LPA (larval, pupal, adult) = 20 ingroup taxa, 122 characters; ADULT (only adult characters) = 49 ingroup taxa, 86 adult characters] were analyzed under parsimony and Bayesian methods. The five outgroup taxa, representing all four extant families in the Psychomyioidea and the single family in the Hydropsychoidea, remained constant in all datasets. The TOTAL and ADULT datasets included all 20 currently recognized polycentropodid genera placed in 3 subfamilies, and the LPA and TOTAL datasets included characters interpreted from structures of the larvae, pupae, and adults. Results rejected the monophyly of Polycentropodidae, as currently defined; however, the monophyly of the three largest cosmopolitan genera, Polycentropus, Polyplectropus, and Nyctiophylax, could not be rejected nor confirmed. The monophyly of the following taxa was strongly supported in all analyses: Cernotina, Cyrnellus, Kambaitipsyche, Neureclipsis, Paranyctiophylax, New World Polyplectropus sensu stricto, Placocentropus, Neotropical Nyctiophylax, and in the outgroup, Psychomyia + Xiphocentron; while monophyly was strongly supported in some, but not all analyses for the following taxa: Cyrnus, Antillopsyche, Pseudoneureclipsis, Polycentropus sensu stricto, Pseudoneureclipsinae, New Zealand Polyplectropus, Polycentropodinae, Cyrnodes scotti + Pahamunaya jihmita. The implementation of two different analytical methods revealed some areas of conflict which would not have been detected under a single method of analysis. Contradictory results among the datasets were primarily due to either inclusion or exclusion of key sets of characters (i.e., immature characters); and second, missing data negatively affected phylogenetic reconstruction when proportions of characters with missing data were high and characters without missing data were unable to provide adequate phylogenetic signal due to high variation in rates of evolution among characters. Therefore, a combination of few overall characters that have high variation in rates of change, plus an abundance of missing data may be problematic and may lead to poorly resolved trees, thus decreased accuracy. This study also emphasized the importance in phylogenetic reconstruction of including data from all available sources. Several taxonomic changes were necessary in order for classification to properly reflect phylogeny. Three new genera, all from the Neotropical region, will be described in future publications. The redefinition of Paranyctiophylax as a valid genus in Polycentropodinae was confirmed. Additionally, the recommendation was made that North American Polycentropus species previously belonging in Plectrocnemia or Holocentropus be recognized as such (either Plectrocnemia or Holocentropus depending on original designation) and not as belonging in Polycentropus. Furthermore, species described in Polycentropus post-1944 in North America are transferred to either Holocentropus or Plectrocnemia to reflect previously hypothesized sister relationships. The following new or reinstated combinations were proposed: Plectrocnemia albipuncta Banks, 1930 combinatio revivisco; Plectrocnemia aureola Banks, 1930 comb. rev.; Plectrocnemia cinerea (Hagen), 1861 comb. rev.; Plectrocnemia clinei Milne, 1936 comb. rev.; Plectrocnemia crassicornis (Walker), 1852 comb. rev.; Plectrocnemia jenula (Denning), in Denning & Sykora, 1966 combinatio nova; Plectrocnemia icula (Ross), 1941 comb. nov.; Plectrocnemia nascotia (Ross), 1941 comb. nov.; Plectrocnemia remota (Banks), 1911 comb. rev.; Plectrocnemia sabulosa (Leonard & Leonard), 1949 comb. nov.; Plectrocnemia smithae (Denning), 1949 comb. nov.; Plectrocnemia vigilatrix Navás, 1933 comb. rev.; Plectrocnemia weedi (Blickle & Morse), 1955 comb. nov.; Holocentropus chellus (Denning), 1964 comb. nov.; Holocentropus flavus Banks, 1908 comb. rev.; Holocentropus glacialis Ross, 1938 comb. rev.; Holocentropus grellus Milne, 1936 comb. rev.; Holocentropus interruptus Banks, 1914 comb. rev.; Holocentropus melanae Ross, 1938 comb. rev.; Holocentropus milaca (Etnier), 1968 comb. nov.; Holocentropus picicornis (Stephens), 1836 comb. rev. Additional taxonomic changes proposed based on current findings were: 1) the elevation of Pseudoneureclipsinae to family status: Pseudoneureclipsidae Ulmer status novus; and 2) the resurrection of Placocentropus Schmid, nomen revivisco, to include the following species: Placocentropus aspinosus (Schmid), 1964 comb. nov.; Placocentropus chilensis (Yamamoto), 1966 comb. nov.; Placocentropus obtusus Schmid, 1955 comb. rev.; Placocentropus quadriappendiculatus (Schmid), 1964 comb. nov.; Placocentropus quadrispinosus (Schmid), 1964 comb. nov.; Placocentropus tuberculatus (Flint), 1983 comb. nov.; Placocentropus valdiviensis (Flint), 1983 comb. nov. A phylogeny of New World Polyplectropus species was inferred. Characters were interpreted from structures of the male and female genitalia as well as the fore- and hind wings. Parsimony and Bayesian phylogenetic analyses of 89 ingroup taxa (97% of the known New World diversity in the genus), two outgroup taxa, and 59 morphological characters were performed. Results of the parsimony and Bayesian analyses were similar, although the Bayesian tree was less resolved. Monophyly of the panamensis and charlesi Groups, as currently defined, was rejected. A total of 10 lineages, with varying amounts of support, were recognized. These groups are the alienus Group (2 species), annulicornis Group (11 species, 8 new), bredini Group (19, 7), charlesi Group (3), fuscatus Group (3, 2), guyanae Group (2, 2), manuensis Group (3, 3), narifer Group (5, 3), santiago Group (25, 6), and thilus Group (15, 7). Four species remain unassigned to any species-group: P. beccus, P. beutelspacheri, P. kanukarum, and P. nayaritensis. The distribution of the genus is mostly restricted to the Mexican and Brazilian subregions of the Neotropics. Most of the species and species-groups are regional endemics. The taxonomy of New World species of Polyplectropus Ulmer, 1905 was revised to include detailed male and female diagnoses, descriptions, illustrations, distribution records, and keys to males of all species and species-groups. A key to genera of New World Polycentropodidae, including a redescription of Polyplectropus, was provided. The homology of the male genitalia of species in the genus was discussed, as well as reassessment and diagnoses of 10 species groups, 6 newly established. A total of 92 species were treated, 39 described as new: Polyplectropus adamsae, sp. nov. (Peru), P. alatespinus, sp. nov. (Brazil), P. amazonicus, sp. nov. (Brazil), P. andinensis, sp. nov. (Argentina, Bolivia), P. blahniki, sp. nov. (Venezuela), P. bolivianus, sp. nov. (Bolivia), P. brasilensis, sp. nov. (Brazil), P. vii brborichorum, sp. nov. (Ecuador), P. cressae, sp. nov. (Venezuela), P. colombianus, sp. nov. (Colombia), P. corniculatus, sp. nov. (Peru), P. cuzcoensis, sp. nov. (Peru), P. ecuadoriensis, sp. nov. (Ecuador), P. flintorum, sp. nov. (Venezuela), P. gaesum, sp. nov. (Brazil), P. guyanae, sp. nov. (Guyana, Venezuela), P. holzenthali, sp. nov. (Brazil), P. hystricosus, sp. nov. (Brazil), P. insularis, sp. nov. (Panama), P. julitoi, sp. nov. (Brazil), P. kanukarum, sp. nov. (Guyana), P. maculatus, sp. nov. (Venezuela), P. manuensis, sp. nov. (Peru), P. matatlanticus, sp. nov. (Brazil), P. minensium, sp. nov. (Brazil), P. novafriburgensis, sp. nov. (Brazil), P. peruvianus, sp. nov. (Peru), P. petrae, sp. nov. (Brazil), P. pratherae, sp. nov. (Brazil), P. puyoensis, sp. nov. (Ecuador), P. robertsonae, sp. nov. (Bolivia), P. rodmani, sp. nov. (Brazil), P. rondoniensis, sp. nov. (Brazil), P. tragularius, sp. nov. (Brazil), P. tripunctatum, sp. nov. (Peru), P. venezolanus, sp. nov. (Venezuela), P. woldai, sp. nov. (Panama), P. zamoranoensis, sp. nov. (Honduras), and P. zuliae, sp. nov. (Venezuela). Polyplectropus buchwaldi is designated as a nomen dubium.
Understanding and strengthening exposure judgments using Bayesian integrated exposure assessment strategies.
(2010-12) Logan, Perry William
Accurate exposure assessments are critical for ensuring that potentially hazardous exposures are properly identified and controlled. The availability and accuracy of exposure assessments can determine whether resources are appropriately allocated to engineering and administrative controls, medical surveillance, personal protective equipment and other programs designed to protect workers. A desktop study was performed using videos, task information and sampling data to evaluate the accuracy and potential bias of participants' exposure judgments. Desktop exposure judgments were obtained from occupational hygienists for material handling jobs with small air sampling data sets (0-8 samples) and without the aid of computers. In addition, data interpretation tests were administered to participants where they were asked to estimate the 95th percentile of an underlying lognormal exposure distribution from small data sets. Participants were presented with an exposure data interpretation or rule-of-thumb training which included a simple set of rules for estimating 95th percentiles for small data sets from a lognormal population. Results of each data interpretation test and qualitative and quantitative exposure judgments were compared with a reference judgment obtained through a Bayesian probabilistic analysis of the sampling data to investigate overall judgment accuracy and bias. There were a total of 4,386 participant-task-chemical judgments for all data collections: 552 qualitative judgments made without sampling data and 3,834 quantitative judgments with sampling data. The data interpretation tests and quantitative judgments were significantly better than random chance and much improved by the rule of thumb training. In addition, the rule of thumb training reduced the amount of bias in the data interpretation tests and quantitative judgments. The mean data interpretation test % correct scores increased from 47% to 64% after the rule-of-thumb training (p<0.001). The accuracy for quantitative desktop judgments increased from 43% to 63% correct after the rule-of-thumb training (p<0.001). The rule of thumb training did not significantly impact accuracy for qualitative desktop judgments. The finding that even some simple statistical rules of thumb improve judgment accuracy significantly suggests that hygienists need to routinely use statistical tools while making exposure judgments using monitoring data. Logistic regression analysis indicated "years of exposure assessment experience" (p<0.05), "highest EHS degree" (p<0.05) and a participant's "data interpretation test score" (p<0.05) directly impacted qualitative exposure judgment accuracy. Logistic regression models of quantitative judgment accuracy showed positive correlation with "greater than 10 years of exposure assessment experience" (p<0.05), "highest EHS degree" (p<0.05), a participant's "data interpretation test score" (p<0.001), rule-of-thumb data interpretation training (p<0.001), and the number of sample data points available for a judgment (p<0.005). Analyzing judgments in subsets for participants with fewer or more than 10 years experience indicated additional correlations with Certified Industrial Hygienist and Certified Safety Professional certifications, total number of task exposure assessments, and career number of air surveys. The correlation of qualitative and quantitative exposure judgment accuracy with "greater than 10 years experience" supports similar research findings from other fields. The results of this study indicate that several determinants of experience, education and training, in addition to the availability of sampling data, significantly impact the accuracy of exposure assessments for the set of exposure tasks and agents used in this study. The findings also suggest methods for enhancing exposure judgment accuracy through statistical tools and specific training. Simulations were designed to evaluate the performance of several quantitative exposure assessments strategies for different exposure distributions. Bayesian tools are becoming popular and have been included in the simulations for this study along with simple comparison, point estimate and upper confidence limit strategies using minimum sample sizes less than 7 samples. The decision statistic selected for the simulations was the 95th percentile which defines acceptable exceedance fractions by 0.01%, 0.1%, 1%, and unacceptable defined by 10%, 20%, 30% and 50%. Bayesian strategies with using professional judgment were also included to illustrate the impact of an incorrect prior judgment. For acceptable exposure distributions, simple comparison and professional judgment integrated Bayesian strategies showed the highest probability for detecting an acceptable exposure. Bayesian strategies without professional judgment followed by upper confidence limit strategies were least likely to incorrectly define unacceptable exposure distributions as acceptable. Reviewing the different minimum sampling numbers for strategies indicate that Bayesian integrated methods most often arrive a correct decisions with less samples than other strategies. The results of this study can help design more effective and efficient exposure assessment and management strategies which will hopefully provide a transparent mechanism to strengthen accuracy and bias of exposure judgments.

University Digital Conservancy

Browse by Subject

Browsing by Subject "Bayesian"