<p>This dissertation is a collection of four essays on nonstationary time series econometrics, which are grouped into four chapters. The first chapter investigates the inference in mildly explosive autoregressions under unconditional heteroskedasticity. The second chapter develops a new approach to forecasting a highly persistent time series that employs feasible generalized least squares (FGLS) estimation of the deterministic components in conjunction with Mallows model averaging. The third chapter proposes new bootstrap procedures for detecting multiple persistence shifts in a time series driven by nonstationary volatility. The last chapter studies the problem of testing partial parameter stability in cointegrated regression models.</p>

Meta-analysis is a popular statistical method that synthesizes evidence from multiple studies. Conventionally, both the hierarchical and bivariate models for meta-analysis of diagnostic test accuracy (DTA) studies assume that the random-effects follow the bivariate normal distribution. However, this assumption is restrictive, and inferences could be misleading when it is violated. On the other hand, subjective methods such as inspection of forest plots are used to identify outlying studies in a meta-analysis of DTA studies. Moreover, inferences made using the well-established bivariate random-effects models, when outlying or influential studies are present, may lead to misleading conclusions. Thus, the aim of this thesis is to address these issues by introducing alternative and robust statistical methods. First, we extend the current bivariate linear mixed model (LMM) by assuming a flexible bivariate skew-normal distribution for the random-effects. The marginal distribution of the proposed model is analytically derived so that parameter estimation can be performed using standard likelihood methods. Overall, the proposed model performs better in terms of confidence interval width of the overall sensitivity and specificity, and with regards to bias and root mean squared error of the between-study (co)variances than the traditional bivariate LMM. Second, we propose objective methods based on solid statistical reasoning for identifying outlying and/or influential studies in a meta-analysis of DTA studies. The performances of the proposed methods are evaluated using a simulation study. The proposed methods outperform and avoid the subjectivity of the currently used ad hoc approaches. Finally, we develop a new robust bivariate random-effects model which accommodates outlying and influential observations and leads to a robust statistical inference by down-weighting the effect of outlying and influential studies. The proposed model produces robust point estimates of sensitivity and specificity compared to the standard models, and also generates a similar point and interval estimates of sensitivity and specificity as the standard models in the absence of outlying or influential studies. / Thesis / Doctor of Philosophy (PhD) / Diagnostic tests vary from the noninvasive rapid strep test used to identify whether a patient has a bacterial sore throat to the much complex and invasive biopsy test used to examine the presence, cause, and extent of a severe condition, say cancer. Meta-analysis is a widely used statistical method that synthesizes evidence from several studies. In this thesis, we develop novel statistical methods extending the traditional methods for meta-analysis of diagnostic test accuracy studies. Our proposed methods address the issue of modelling asymmetrical data, identifying outlier studies, and optimally accommodating these outlying studies in a meta-analysis of diagnostic test accuracy studies. Using both real-life and simulated datasets, we show that our proposed methods perform better than conventional methods in a wide range of scenarios. %Therefore, we believe that our proposed methods are essential for methodologists, clinicians and health policy professionals in the process of making a correct judgment to using the appropriate diagnostic test to diagnose patients.

An efficient reliability technique has been developed based on Response Surface Methodology (RSM) in conjunction with the First Order Second Moment (FOSM) reliability method. The technique is applied when the limit state function cannot be obtained explicitly in terms of the design variables, i.e., when the analysis is performed using numerical techniques such as finite elements. The technique has proven to be efficient because it can handle problems with large numbers of design variables and correlated as well as nonnormal random variables. When compared with analytical results, the method has shown excellent agreement. The technique contains a sensitivity analysis scheme which can be used to reduce the computation time resulting in nearly the same accuracy. This technique allows the extension of most finite element codes to account for probabilistic analysis, where statistical variations can be added to the design variables. An explicit solution for rocket motors consisting of propellant and steel case under environmental temperature variations is compared to the RSM technique. The method is then used for the analysis of rocket motors subjected to mechanical loads for which the stress analysis is performed using the finite element method. The technique is also applied to study the reliability of a laminated composite plate with geometric nonlinearity subjected to static and time dependent loadings. Different failure modes were considered as well as different meshes. Results have shown that when the relative size of the element is introduced into the probabilistic model, the same reliability value is obtained regardless of the number of elements in the mesh. This is good because it allows the technique to be used for problems where the failure region is unknown. / Ph. D.

A common problem encountered in practice is determining when it is appropriate to change the sampling interval for control charts. This thesis examines this problem for Shewhart X̅ charts. Duncan's economic model (1956) is used to develop a relationship between the most appropriate sampling interval and the present rate of"disturbances,” where a disturbance is a shift to an out of control state. A procedure is proposed which switches the interval to convenient values whenever a shift in the rate of disturbances is detected. An example using simulation demonstrates the procedure. / M.S.

<p dir="ltr">The passing of the Equity Executive Order by the US government in 2021 and the Equity Action Plan developed by the US Department of Transportation have made addressing equity a priority in transportation planning projects. The Equity Action Plan recognizes socio-economically disadvantaged (SED) and rural communities as overburdened and underserved in transportation services and highlights the need to provide basic equality of opportunities and a fair distribution of burdens in transportation. In response, several transportation planning agencies have proposed their own criteria, methods, metrics, and tools to quantify equity issues and use them in planning. However, transport equity is a multi-faceted phenomenon and its quantification faces challenges due to a lack of standards and a comprehensive assessment framework.</p><p dir="ltr">The first objective of this dissertation is to develop a nuanced understanding in three key aspects of transport equity – (i) accessibility to opportunities, (ii) environmental burden, and (iii) health outcomes – based on the concept of ‘compound disadvantage’. A detailed assessment of prominent accessibility measures reveals a substantial measurement bias in the current measures, which can result in inappropriate conclusions such as an overestimating accessibility to opportunities for SED neighborhoods on average by 16%. Despite this, spatial accessibility is found to be high for compact, urban areas which also tend to have higher concentration of SED communities. However, there are significant modal differences in accessibility that reveal a substantial lack of utilization of infrastructure for alternate modes of travel – public transit, walking, and bicycling.</p><p dir="ltr">Evaluation of inequalities in environmental and health outcomes shows substantial disadvantage faced by SED communities, particularly poor people and people of color. A proposed emission equity index shows that low-income and racial minority neighborhoods of Indiana’s largest cities disproportionately experience vehicular pollution from travelers residing in high-income, White-majority areas passing through their neighborhoods. Similarly, essential workers living in low-income areas are observed to have experienced significantly worse COVID-19 infection rates than in high-income areas in Chicago and New York City, and this effect was mediated by their ‘mobility vulnerability’. Together, these results suggest a strong compounding of disadvantage by the current transportation systems for already disadvantaged communities despite their higher accessibility to opportunities owing to their predominantly urban residences.</p><p dir="ltr">The second objective of this dissertation is to integrate the equity measures in an interactive screening tool for identifying the vulnerable and priority areas for investment. ‘Indiana Equity Atlas’, an equity-screening dashboard tool, is developed to allow transportation planners and analysts to identify priority areas in terms of compound disadvantage of two selected indicators of socio-economic, accessibility, environmental burden, and health disadvantage. With this tool and the associated data and metrics, this work seeks to provide a comprehensive framework for identifying vulnerable regions to formally capture equity issues in transportation and urban planning and analysis.</p>

Yes / This paper investigates the feasibility of using artificial neural networks (NNs) to predict the shear capacity of concrete members reinforced longitudinally with fibre reinforced polymer (FRP) bars, and without any shear reinforcement. An experimental database of 138 test specimens failed in shear is created and used to train and test NNs as well as to assess the accuracy of three existing shear design methods. The created NN predicted to a high level of accuracy the shear capacity of FRP reinforced concrete members. Garson index was employed to identify the relative importance of the influencing parameters on the shear capacity based on the trained NNs weightings. A parametric analysis was also conducted using the trained NN to establish the trend of the main influencing variables on the shear capacity. Many of the assumptions made by the shear design methods are predicted by the NN developed; however, few are inconsistent with the NN predictions.

<p dir="ltr">Extensive tree-cutting operations and prioritization of crop production a century ago caused a significant decline in maple syrup production in Indiana. Today, there is an increasing consumer interest in natural and locally produced food, creating potential for regrowth in the industry. Understanding both production and consumption behaviors is crucial to capitalize on this potential and ensure sustainable sugarbush resources in the state. This thesis comprised two chapters, presenting the findings on consumers and producers, respectively. The first chapter examined the factors influencing consumers' behavior toward maple syrup purchase and satisfaction. Results suggested that suburban residency, education, age, income, and lifestyle factors – such as visiting farmers' markets and purchasing organic food – may influence consumers' attitudes toward purchasing maple syrup. Moreover, consumers who purchase organic food tended to be satisfied with maple syrup and were more likely to recommend it to others. The second chapter explored the factors motivating maple syrup producers to manage their forests and investigated whether crowd-in or crowd-out effects exist among different management practices. I used logistic regressions to identify relationships among the adoption of a variety of forest management practices and a variety of independent variables, using data pooled from two producer surveys conducted in Indiana in 2022 and 2023, respectively. Production as a hobby, owning a sugarbush, and production capacity all positively affected the likelihood of management practice adoption. In addition, crowding-in effects were identified in adopting forest management practices, suggesting that producers were more likely to adopt multiple practices simultaneously. These chapters emphasized the importance of understanding maple syrup consumer socio-demographic characteristics for effective marketing strategies and producer behavior for sustainable production practices to promote industry growth.</p>

Roof and rib failures in underground coal mines are one of the major problems facing the industry today. In addition to safety considerations, the resulting economic impact of such failures is staggering. Uncovering and replacing buried and damaged equipment and clearing entries can account for a large expenditure in lost man-hours and machinery. Yet, because of the complex nature of their formation, geological variability, and structural characteristics, coal mine roof strata are one of the least controllable of all mine design parameters. This is especially true along the leading (southeastern) edge of the Appalachian coalfields where considerable faulting and movement have contributed to hazardous coal mining roof conditions. For this research, a detailed study of several mines, in the southern Appalachian coalfields, was undertaken to determine the most prominent geomechanical factors affecting roof stability and to evaluate their influence in promoting unstable ground conditions. In order to accomplish this task, the major geological and geomechanical features found to be detrimental to the coal mine roof within the Appalachian basin were identified and mapped in four Virginia mines. Statistical processing by chi-square and linear regression analysis as well as analytical analysis by the finite element method were used to determine the influence of geology, mine-layout, and support methods on roof stability. It was found that some easily determined parameters can be successfully used to predict potentially unstable areas. A simplified roof classification system was developed based on the geomechanical parameters, which can be used to assess the stability of a particular roof type. A Roof Rating Index was also devised capable of expressing the probability of failure under a given set of geomechanical conditions. / Ph. D.

𝐈𝐧𝐭𝐫𝐨𝐝𝐮𝐜𝐭𝐢𝐨𝐧) Tropical forests are the world’s most structurally complex ecosystems, providing key functions like biomass accumulation and harboring biodiversity. Yet climate-change poses a potential threat to the stability of these forests – tropical cyclones in the North Atlantic are projected to increase in intensity, leading to higher forest damage rates, potentially reducing their carbon sequestration and biodiversity potential. Hurricane Maria in 2017 was a possible portent of this dynamic, causing widespread devastation in Puerto Rico. How do forests resist such severe disturbances? Forests ecosystems contain ecological memory – physical and biological legacies from past natural disturbances like fires and windstorms – that can increase their resilience to future disturbances. In fire-prone forests, for example, prior exposure to non-severe fires has been shown to increase resistance to severe wildfires. Does the same mechanism apply in cyclone-prone tropical forests? In this dissertation, I examine how chronic exposure to non-hurricane winds impacts hurricane damage at the tree, stand, and landscape scales in Puerto Rico. Specifically, I ask – 1) Do chronic winds alter tree architecture to reduce their risk of stem-breaks? 2) Do chronic winds reduce forest stand structural complexity? 3) Do chronic winds and lower canopy structural complexity reduce individual tree and forest stand damage from Hurricane Maria? 𝐌𝐞𝐭𝐡𝐨𝐝𝐬) I used a novel combination of remote sensing, fieldwork, and high-resolution Light Detection and Ranging (LiDAR) data collected in 2016 to address the above questions. In Chapter 1, I connected sub-meter resolution GPS data and 30 years of forest inventory with 0.03m resolution airborne LiDAR data to evaluate chronic wind impacts on the tree architecture and wind-risk of 124 forest trees of four key species. In Chapter 2, I used machine learning, remote-sensing and LiDAR data to predict the chronic wind impacts on the canopy height and structural complexity of ~20,000 0.28 ha forested sites across climatic, forest age and topographic gradients. In Chapter 3, I used pre-storm size and damage assessment field data for ~7,000 trees of 160 species across 14, 0.25 ha sites spanning an 800 m elevation gradient, alongside a remote-sensing dataset of ~12,000 forests to evaluate multiscale drivers – including canopy structural complexity – of individual, stand and landscape level hurricane damage. 𝐑𝐞𝐬𝐮𝐥𝐭𝐬 𝐚𝐧𝐝 𝐜𝐨𝐧𝐜𝐥𝐮𝐬𝐢𝐨𝐧𝐬) At the individual tree scale, I found that long-lived species grew ~3.5 m shorter and ~ 4 m2 smaller crowns on average due to chronic wind-exposure, substantially reducing their estimated wind-risk, whereas short-lived species did not respond architecturally to chronic winds. At the stand and landscape scales, I found that chronic winds reduced canopy height by 2.12 m on average, and that structural complexity decreased substantially with forest age. I found that stand-level hurricane damage was primarily a function of increased canopy structural complexity, which in turn decreased with elevation; and that individual tree damage increased with stem size and varied only slightly by species, with short-lived species much more susceptible to damage. My findings suggest that tropical forest resistance to increasingly severe hurricanes depends largely on the physical structure of their canopies, and only then on adapted species-level life-history traits. The physical structure of forest canopies, in turn, changes substantially with exposure to non-hurricane winds. In old-growth forests in Puerto Rico, there is therefore evidence that ecological memory driven by exposure to non-hurricane winds can protect forests from severe wind disturbances. However, younger, more structurally complex forests may be potentially increasingly more vulnerable in a changing climate.

published_or_final_version / Applied Statistics / Master / Master of Social Sciences