No description available.

Several standard methods are available to analyze and estimate parameters of count data. None of these methods are designed to account for potential misclassification of the data, where counts are observed or recorded as higher or lower than their actual value. These false counts can result in erroneous conclusions and biased estimates. For this paper, a standard estimation model was modified in several ways in order to incorporate each misclassification mechanism. The probability distribution of the observed data was derived and combined with informative distributions for the misclassification parameters. Once this additional information was taken into account, a distribution of observed data conditional on only the parameter of interest was obtained. By incorporating information about the misclassification mechanisms, the resulting estimation will be more accurate than the standard methods. To demonstrate the flexibility of this approach, data from a count distribution affected by various misclassification mechanisms were simulated. Each dataset was analyzed by several standard estimation methods and an appropriate new method. The results from all simulated data were compared, and the impact of each mechanism in regards to each estimation method was discussed. Data from a colorectal polyp prevention study were also analyzed with all available methods to showcase the incorporation of additional covariates.

This dissertation examines the effects of the Supplemental Nutrition Assistance Program or SNAP, formally known as the Food Stamp Program, on adult weight outcomes. The focus of this work is to uncover the causal effects of the program by applying rigorous identification methods as well as techniques that address data limitations. By understanding the true impact of SNAP on adult obesity, policymakers can pursue appropriate reform measures to avoid unintended consequences of the program while promoting healthy weight outcomes for low-income Americans. The first essay expands on previous work examining the effects of SNAP participation on adult obesity. Previous research provides some evidence that SNAP participation may have a small positive effect on weight gain for women and no significant effect on men. However, additional research has found that misreporting of SNAP participation in surveys is prevalent and that analysis of program effects when participation is misclassified (misreported) can produce estimates that are biased and misleading. Until now, nearly all studies examining the effects of SNAP on adult obesity have ignored the issue of respondent misreporting. This chapter uses state-level policy variables regarding SNAP administration to instrument for SNAP participation for NLSY79 respondents. To address respondent misreporting I adopt an approach based on parametric methods for misclassified binary dependent variables that produces consistent estimates when using instrumental variables. This study is the first to document the considerable rates of SNAP participation under-reporting in the NLSY79 dataset. In addition, this study finds that, although SNAP participation increases adult BMI and the likelihood of being obese, without correcting for misreporting bias the estimates are overstated by nearly 100 percent. The second essay uses the same data but applies a different identification strategy to investigate the intensive margin effects of SNAP on adult obesity. To mitigate the severity of endogenous participation and misreporting biases, I employ a strategy that examines only individuals who report participating in SNAP. I utilize a quasi-experimental variation in SNAP amount per adult due to the timing of school eligibility for children. The identification examines the proportion of school-age children in SNAP households who automatically qualify for in-school nutrition assistance programs. A greater proportion of school-age children eligible for free in-school meals proxies for an exogenous increase in the amount of SNAP benefits available per adult. This study finds that increases in SNAP benefits, as proxied by increases in the proportion of school-age children, reduce BMI and the probability of being severely obese for SNAP adults. Taken together, the results of this dissertation present an intriguing depiction of the effects of SNAP on adult obesity that serves to inform both policymakers and future researchers. On the one hand, the findings indicate that participating in SNAP (extensive margin) leads to weight gain and higher rates of obesity. Yet, on the other hand, increases in the amount of benefits for those who report participating in SNAP (intensive margin) actually leads to reductions in weight and the likelihood of being severely obese.

Recent literature has emphasized the importance of changes in occupation, i.e. occupational mobility, for both personal and aggregate outcomes. In this study I examine the various causes and consequences of that important decision. I begin by developing a new, generalized, model of measurement error which can fully incorporate changes in discrete classification like occupational mobility. I then use this framework to show that occupational mobility has spuriously risen in the monthly Current Population Survey. I then study the consequences of occupational mobility using a high quality 2% sample of the German Social Security Data to study how wages change around occupation and employer transitions. The results are consistent with idiosyncratic matching at the occupation, but not the employer, level. For men, wages increase by 5.5 percent following a voluntary employer transition that does not involve an occupation transition and 10.1 percent following voluntary employer transition that does involve an occupation transition. I build a model where workers differ in their cognitive, manual, and interactive skills, which creates comparative advantage in certain occupations. I estimate this model and show that most of the wage gains for young workers following an occupational transition are due to improved matching of worker skill with occupation tasks, and not simply movements to higher paying occupations. I then use the estimated model to show that the matching of workers to their comparative advantage has worsened in Germany between 1975-2010. Finally, I examine the testable implications of models of search and models of learning to see if they are consistent with the facts I have developed about occupational mobility. I find that while search models can be consistent with a number of empirical facts they, by themselves, are not able to rationalize "back-and-forth" switching that is observed in the data.

If auditors question a company’s ability to continue existing, they should issue a going concern opinion in the audit report. Whether or not auditors will issue a going concern opinion depends on auditors’ ability to identify going concern problems, as well as their decision whether or not to issue going concern opinions. In Sweden, the going concern accuracy rate has been low compared to other countries. The aim of this study is therefore to analyse whether it is auditors’ lack of ability to identify going concern problems or their decision not to issue a going concern opinion, or perhaps both, that could explain the relatively low accuracy rate. Interviews with four auditors from the Big Four audit firms and four CFOs from middle sized companies were conducted. The results show that there are factors speaking both for and against auditors’ ability to identify going concern problems, why we cannot say for sure whether auditors’ lack of ability to identify going concern problems could be an explanation to the relatively low accuracy rate. The results do however reveal that auditors actively make decisions not to issue going concern opinions to their clients as much as possible, which could explain the relatively low accuracy rate.

This dissertation develops methods to minimize recommendation error costs when inputs to a rule-based expert system are prone to errors. The problem often arises in web-based applications where data are inherently noisy or provided by users who perceive some benefit from falsifying inputs. Prior studies proposed methods that attempted to minimize the probability of recommendation error, but did not take into account the relative costs of different types of errors. In situations where these differences are significant, an approach that minimizes the expected misclassification error costs has advantages over extant methods that ignore these costs. Building on the existing literature, two new techniques - Cost-Based Input Modification (CBIM) and Cost-Based Knowledge-Base Modification (CBKM) were developed and evaluated. Each method takes as inputs (1) the joint probability distribution of a set of rules, (2) the distortion matrix for input noise as characterized by the probability distribution of the observed input vectors conditioned on their true values, and (3) the misclassification cost for each type of recommendation error. Under CBIM, for any observed input vector v, the recommendation is based on a modified input vector v' such that the expected error costs are minimized. Under CBKM the rule base itself is modified to minimize the expected cost of error. The proposed methods were investigated as follows: as a control, in the special case where the costs associated with different types of errors are identical, the recommendations under these methods were compared for consistency with those obtained under extant methods. Next, the relative advantages of CBIM and CBKM were compared as (1) the noise level changed, and (2) the structure of the cost matrix varied. As expected, CBKM and CBIM outperformed the extant Knowledge Base Modification (KM) and Input Modification (IM) methods over a wide range of input distortion and cost matrices, with some restrictions. Under the control, with constant misclassification costs, the new methods performed equally with the extant methods. As misclassification costs increased, CBKM outperformed KM and CBIM outperformed IM. Using different cost matrices to increase misclassification cost asymmetry and order, CBKM and CBIM performance increased. At very low distortion levels, CBKM and CBIM underperformed as error probability became more significant in each method's estimation. Additionally, CBKM outperformed CBIM over a wide range of input distortion as its technique of modifying an original knowledge base outperformed the technique of modifying inputs to an unmodified decision tree.

Dryland salinity is increasing in the upper catchments of central and northern New South Wales, Australia. Consequently, salts may be exported downstream, which could adversely affect cotton irrigated-farming systems. In order to assess the potential threat of salinity a simple salt balance model based on progressively saline water (i.e., ECiw 0.4, 1.5, 4.0 and 9.0 dS/m) was used to simulate the potential impact of salinisation due to the farming systems. The study was carried out in the lower Namoi valley of northern New South Wales, Australia. A comparison has been made of the various non-linear techniques (indicator kriging, multiple indicator kriging and disjunctive kriging) to determine an optimal simulation method for the risk assessment. The simulation results indicate that potential salinisation due to application of the water currently used for irrigation (ECiw) is minimal and may not pose any problems to sustainability of irrigated agriculture. The same results were obtained by simulation based on irrigation using slightly more saline water (ECiw 1.4 dS/m). However, simulations based on irrigation using water of even lower quality (ECiw of 4 and 9.0 dS/m), shows potential high salinisation, which will require management inputs for sustainable cropping systems, especially legumes and wheat, which are used extensively in rotation with cotton. Disjunctive kriging was the best simulation method, as it produced fewer misclassifications in comparison with multiple-indicator kriging and indicator kriging. This study thus demonstrates that we can predict the salinity risk due to application of irrigation water of lower quality than that of the current water used. In addition, the results suggest here problems of excessive deep drainage and inefficient use of water might be a problem. The second part of this thesis deals with soil information required at the field scale for management practices particularly in areas where deep drainage is large. Unfortunately, traditional methods of soil inventory at the field level involve the design and adoption of sampling regimes and laboratory analysis that are time-consuming and costly. Because of this more often than not only limited data are collected. In areas where soil salinity is prevalent, detailed quantitative information for determining its cause is required to prescribe management solutions. This part deals with the description of a Mobile Electromagnetic Sensing System (MESS) and its application in an irrigated-cotton field suspected of exhibiting soil salinity. The field is within the study area of part one of this thesis-located about 2 km south west of Wee Waa. The EM38 and EM31 (ECa) data provide information, which was used in deciding where soil sample sites could be located in the field. The ECa data measured by the EM38 instrument was highly correlated with the effective cation exchange capacity. This relationship can be explained by soil mineralogy. Using different soil chemical properties (i.e. ESP and Ca/Mg ratio) a detailed transect study was undertaken to measure soil salinity adjoining the water storage. It is concluded that the most appropriate management option to remediation of the problem would be to excavate the soil directly beneath the storage floor where leakage is suspected. It is recommended that the dam not be enlarged from its current size owing to the unfavourable soil mineralogy (i.e. kaolin/illite) located in the area where it is located.

Earlier retrospective noise exposure assessments for use in epidemiological research were not adequately characterized because they did not properly account for use of hearing protection devices (HPD) which would result in potential misclassification. Exposure misclassification has been shown to attenuate exposure-outcomes relations. In the case of already subtle relationships such as noise and cardiovascular diseases, this would potentially annihilate any association. We investigated two approaches using Workers’ Compensation Board (WorkSafe BC) audiometric surveillance data to (i) re-assess the noise exposure in a cohort of lumber mill workers in British Columbia using data on the use of HPD and the determinants of their use available through WorkSafe BC, and (ii) test the validity of the new exposure measures by testing their predictions of noise-induced hearing loss, a well-established association. Work history, noise exposure measurements, and audiometric surveillance data were merged together, forming job-exposure-audiometric information for each of 13,147 lumber mill workers. Correction factors specific to each type and class of HPD were determined based on research and standards. HPD-relevant correction factors were created using 1) deterministic methods and self-reported HPD use after filling gaps in the exposure history, or 2) a model of the determinants of use of HPD, then adjusting noise estimates according to the methods’ predictions and attenuation factors. For both methods, the HPD-adjusted and unadjusted noise exposure estimates were cumulated across all jobs each worker held in a cohort-participating lumber mill. Finally, these noise metrics were compared by examining how well each predicted hearing loss. Analyses controlled for gender, age, race as well as medical and non-occupational risk factors. Both methods led to a strengthening of the noise-hearing loss relationships compared to methods using HPD-unadjusted noise estimates. The method based on the modeling of HPD use had the best performance with a four-fold increase in the slope compared to the unadjusted noise-hearing loss slope. Accounting for HPD use in noise exposure assessment is necessary since we have shown that misclassification attenuated the exposure-response relationships. Exposure-response analyses subsequent to exposure reassessment provide predictive validity and gives confidence in the exposure adjustment methods.

Earlier retrospective noise exposure assessments for use in epidemiological research were not adequately characterized because they did not properly account for use of hearing protection devices (HPD) which would result in potential misclassification. Exposure misclassification has been shown to attenuate exposure-outcomes relations. In the case of already subtle relationships such as noise and cardiovascular diseases, this would potentially annihilate any association. We investigated two approaches using Workers’ Compensation Board (WorkSafe BC) audiometric surveillance data to (i) re-assess the noise exposure in a cohort of lumber mill workers in British Columbia using data on the use of HPD and the determinants of their use available through WorkSafe BC, and (ii) test the validity of the new exposure measures by testing their predictions of noise-induced hearing loss, a well-established association. Work history, noise exposure measurements, and audiometric surveillance data were merged together, forming job-exposure-audiometric information for each of 13,147 lumber mill workers. Correction factors specific to each type and class of HPD were determined based on research and standards. HPD-relevant correction factors were created using 1) deterministic methods and self-reported HPD use after filling gaps in the exposure history, or 2) a model of the determinants of use of HPD, then adjusting noise estimates according to the methods’ predictions and attenuation factors. For both methods, the HPD-adjusted and unadjusted noise exposure estimates were cumulated across all jobs each worker held in a cohort-participating lumber mill. Finally, these noise metrics were compared by examining how well each predicted hearing loss. Analyses controlled for gender, age, race as well as medical and non-occupational risk factors. Both methods led to a strengthening of the noise-hearing loss relationships compared to methods using HPD-unadjusted noise estimates. The method based on the modeling of HPD use had the best performance with a four-fold increase in the slope compared to the unadjusted noise-hearing loss slope. Accounting for HPD use in noise exposure assessment is necessary since we have shown that misclassification attenuated the exposure-response relationships. Exposure-response analyses subsequent to exposure reassessment provide predictive validity and gives confidence in the exposure adjustment methods.

Dryland salinity is increasing in the upper catchments of central and northern New South Wales, Australia. Consequently, salts may be exported downstream, which could adversely affect cotton irrigated-farming systems. In order to assess the potential threat of salinity a simple salt balance model based on progressively saline water (i.e., ECiw 0.4, 1.5, 4.0 and 9.0 dS/m) was used to simulate the potential impact of salinisation due to the farming systems. The study was carried out in the lower Namoi valley of northern New South Wales, Australia. A comparison has been made of the various non-linear techniques (indicator kriging, multiple indicator kriging and disjunctive kriging) to determine an optimal simulation method for the risk assessment. The simulation results indicate that potential salinisation due to application of the water currently used for irrigation (ECiw) is minimal and may not pose any problems to sustainability of irrigated agriculture. The same results were obtained by simulation based on irrigation using slightly more saline water (ECiw 1.4 dS/m). However, simulations based on irrigation using water of even lower quality (ECiw of 4 and 9.0 dS/m), shows potential high salinisation, which will require management inputs for sustainable cropping systems, especially legumes and wheat, which are used extensively in rotation with cotton. Disjunctive kriging was the best simulation method, as it produced fewer misclassifications in comparison with multiple-indicator kriging and indicator kriging. This study thus demonstrates that we can predict the salinity risk due to application of irrigation water of lower quality than that of the current water used. In addition, the results suggest here problems of excessive deep drainage and inefficient use of water might be a problem. The second part of this thesis deals with soil information required at the field scale for management practices particularly in areas where deep drainage is large. Unfortunately, traditional methods of soil inventory at the field level involve the design and adoption of sampling regimes and laboratory analysis that are time-consuming and costly. Because of this more often than not only limited data are collected. In areas where soil salinity is prevalent, detailed quantitative information for determining its cause is required to prescribe management solutions. This part deals with the description of a Mobile Electromagnetic Sensing System (MESS) and its application in an irrigated-cotton field suspected of exhibiting soil salinity. The field is within the study area of part one of this thesis-located about 2 km south west of Wee Waa. The EM38 and EM31 (ECa) data provide information, which was used in deciding where soil sample sites could be located in the field. The ECa data measured by the EM38 instrument was highly correlated with the effective cation exchange capacity. This relationship can be explained by soil mineralogy. Using different soil chemical properties (i.e. ESP and Ca/Mg ratio) a detailed transect study was undertaken to measure soil salinity adjoining the water storage. It is concluded that the most appropriate management option to remediation of the problem would be to excavate the soil directly beneath the storage floor where leakage is suspected. It is recommended that the dam not be enlarged from its current size owing to the unfavourable soil mineralogy (i.e. kaolin/illite) located in the area where it is located.