<div> <div> <div> <p>Billions of birds migrate annually between breeding and wintering habitats, following transient resources. Though a majority of time is spent in breeding and wintering habitats, habitats used during the migratory periods are especially important for migrating birds. Migration and stopovers, where birds rest and refuel before continuing a migratory journey, are critical points in a bird’s annual lifecycle, and are important ecologically, socially, and economically. Populations of migratory birds are declining on a global scale, however, and proper management is vital to their persistence in an urbanizing environment. Indiana in the Midwestern United States is an important area in which to study stopover ecology of migratory birds, as it is a fragmented forest- urban-agricultural matrix almost entirely managed through private ownership. In this dissertation, I studied three questions of stopover ecology within the landscape context of the Midwestern United States, primarily using weather surveillance radar and eBird citizen science data. </p> <p><br></p><p>First, I studied spatiotemporal changes over an 11-year period (autumn 2005-2016 and spring 2006-2017) in densities of nocturnally migrating birds at two radar stations in Indiana. I found that mean density of migratory birds stopping over in Indiana declined by approximately 6.8% annually, but variability in stopover site use increased over the same period. This is consistent with other work completed on continental scales, and highlights the need for further conservation of migratory birds. Second, I studied patterns of stopover site use in Indiana during spring 2016- 2017 and autumn 2015-2016, identifying landscape and local factors associated with those patterns. I used both traditional land cover characteristics and a novel approach using human socioeconomic measures to describe these patterns, and found that socioeconomics, particularly the size of a housing unit, were among the most important predictors of migratory bird density in Indiana. The results from this study suggest that migratory birds are utilizing urban habitats, which are known to contain several novel hazards for birds, but that migratory birds will benefit greatly from interdisciplinary work focusing on urban habitats. Third, I explored a novel method of using weather surveillance radar and eBird citizen science data in combination with each other, to see if both measures provided similar estimates of bird abundances during stopover. Though I found no correlation between the two, I argue that eBird and radar still provide important and complementary insights for the field of migration ecology. Finally, I provide guidelines for private landowners in Indiana on management for declining populations of migratory birds. </p> </div> </div> </div>

Thesis (Ph.D. (Statistics)) -- University of Limpopo, 2019 / This study presents the application of Direct and Indirect methods of Small AreaEstimation(SAE)techniques. Thestudyisaimedatestimatingthetrends and the proportions of households accessing water, sanitation, and electricity for lighting at small areas of the Limpopo Province, South Africa. The study modified Statistics South Africa’s General Household Survey series 2009-2015 and Census 2011 data. The option categories of three variables: Water, Sanitation and Electricity for lighting, were re-coded. Empirical Bayes and Hierarchical Bayes models known as Markov Chain Monte Carlo (MCMC) methods were used to refine estimates in SAS. The Census 2011 data aggregated in ‘Supercross’ was used to validate the results obtained from the models. The SAE methods were applied to account for the census undercoverage counts and rates. It was found that the electricity services were more prioritised than water and sanitation in the Capricorn District of the Limpopo Province. The greatest challenge, however, lies with the poor provision of sanitation services in the country, particularly in the small rural areas. The key point is to suggestpolicyconsiderationstotheSouthAfricangovernmentforfutureequitable provisioning of water, sanitation and electricity services across the country.

No description available.

[pt] Este trabalho propõe a comparação entre os dados do TerraClass 2004 e o Censo Agropecuário 2006 para analisar o uso de um na ausência do outro, para compreender o uso e cobertura do solo pela agropecuária na área de estudo. Através dos dados de uso e cobertura do solo do TerraClass 2004 e do Censo Agropecuário 2006, foi feita uma análise das áreas de cada classe que se enquadram como atividade agropecuária e uma comparação dessas áreas nas duas classificações. A análise utilizou o Sistema de Informação Geográfica (SIG) com imagens de satélite Landsat/TM 5 para auxiliar na análise visual qualitativa das classes e obter respostas sobre o uso e ocupação do solo no município de Lábrea. Foram escolhidos 16 setores censitários sobre os quais se tinha dados de classes referentes a agropecuária, analisando e percebendo uma variação percentual considerável em relação as duas classificações. A metodologia de comparação serviu para verificar que os métodos de aquisição dos dados do TerraClass 2004 e do Censo Agropecuário 2006 apresentam diferenças consideráveis nos resultados. Foram encontradas na classificação do TerraClass, através da análise qualitativa utilizando a imagem de satélite, dúvidas quanto a algumas classes sem vegetação, mas que não foram classificadas como atividade agropecuária, como desflorestamento, e essas consideradas duvidosas. Por fim, o presente trabalho conclui que não é adequado utilizar os dados do TerraClass na ausência do Censo Agropecuário, ou vice-versa, tendo em vista a disparidade encontrada em alguns dos setores censitários analisados. Sugere-se para trabalhos futuros um estudo específico para saber as causas das áreas desmatadas das classes consideradas duvidosas. / [en] This work proposes comparing data provided by 2004 s TerraClass and 2006 s Agribusiness Census, analyzing the validity of using one absent the other to comprehend the land cover in relation to said business within the studied area. The areas of the classes related to agribusiness were compared through the land cover reported in each classification. This analysis took place in a GIS environment with the use of Landsat/TM 5 to aid in the visual qualitative comparison of the data. The region of interest was the municipality of the Labrea district in the southern region of the Amazonas state. Sixteen census sector that showed agribusiness information were chose for the comparison of the percentages of area attributed to that business type. The variation of such areas was noteworthy and it was deemed that one should not use the two sources compared interchangeably as the information does not match. Moreover, some classes of the TerraClass classification stood out and regarded as uncertain – such as generic deflorestation and not forest – as, knowing the regions historical occupation and recent expansion of agribusiness, one would suspect that those are indeed related to the agents studied in this work.

Intro: Public transit offers various advantages for both communities and individuals in terms of mobility and has the potential to serve as a practical alternative to private motorized transportation. Timely and convenient public transit becomes a lifeline for people who cannot access other types of transportation, especially when it comes to emergency care. Method: We built a private network analysis within ArcGIS Pro using coordinate data from hospitals and stand-alone emergency rooms as designated by the North American Industrial Classification System (NAICS) and public transit data from Orlando, FL. Establishing the accessible time as 30 minutes, we computed and visualized the accessible areas where individuals can get to a care facility within the time limit and via public transit. Then, we overlaid the map with demographic distribution and determined the relationship between populations' characteristics and access to care. Result: A total of 4178 active bus stops; among these, there are 1265 within a 5 to 10-minute walk to a care facility, and 68 bus routes, 249 actual care facilities, and 413 census tracts were found in the study area. Sixty-five census tracts showed less than 1% coverage, while 34 displayed complete 100% coverages, with the median service coverage across all tracts around 40%. The t-test series indicated a denser population and higher mobile homeownership in areas without level access. Conclusion: The investigation identified distinct spatial patterns in the distribution of care facilities, concentrated in the Central Business District, and created isolated areas between routes, potentially leaving unwell individuals in residential zones with limited access to timely care.

The research presented in this dissertation touches on all aspects of survey methodology, from questionnaire design to final estimation. We first approach the questionnaire development stage by proposing a method of developing matrix sampling designs, a design where a subset of questions are administered to a respondent in such a way that the administered questions are predictive of the omitted questions. The proposed methodology compares favorably to previous methods when applied to data collected from a household survey conducted in the Nampula province of Mozambique. We approach the data collection stage by proposing a structured procedure of implementing small-scale surveys in such a way that non-sampling error attributed to data collection is minimized. This proposed methodology requires the inclusion of the statistician in the data editing process during data collection. We implemented the structured procedure during the collection of household survey data in the city of Maputo, the capital of Mozambique. We found indications that the data resulting from the structured procedure is of higher quality than the data with no editing. Finally, we approach the estimation phase of sample surveys by proposing a model-based approach to the estimation of the mean squared error associated with synthetic (indirect) estimates. Previous methodology aggregates estimates for stability, while our proposed methodology allows area-specific estimates. We applied the proposed mean squared error estimation methodology and methods found during literature review to simulated data and estimates from 2010 Census Coverage Measurement (CCM). We found that our proposed mean squared error estimation methodology compares favorably to the previous methods, while allowing for area-specific estimates. / Ph. D.

<p dir="ltr">Machine learning (ML) algorithms are increasingly being used in a variety of applications and are heavily relied upon to make decisions that impact people’s lives. ML models are often praised for their precision, yet they can discriminate against certain groups due to biased data. These biases, rooted in historical inequities, pose significant challenges in developing fair and unbiased models. Central to addressing this issue is the mitigation of biases inherent in the training data, as their presence can yield unfair and unjust outcomes when models are deployed in real-world scenarios. This study investigates the efficacy of data acquisition, i.e., one of the stages of data preparation, akin to the pre-processing bias mitigation technique. Through experimental evaluation, we showcase the effectiveness of data acquisition, where the data is acquired using data valuation techniques to enhance the fairness of machine learning models.</p>

Cette étude utilise la base de données à 100% du recensement de 1901, rendue disponible par le projet Le peuple canadien, pour faire une analyse descriptive des arrangements résidentiels des Autochtones du Canada au début du 20e siècle. La province du Manitoba a été sélectionnée pour une étude de cas. Le premier objectif de cette étude est d’évaluer la capacité du recensement de 1901 à représenter la réalité des ménages autochtones et son deuxième objectif est de faire une analyse descriptive des arrangements résidentiels des ménages autochtones, pour l’ensemble de la province ainsi que selon le district de résidence et l’âge et le sexe des individus. Une comparaison est aussi faite avec les ménages non-autochtones. Les principaux résultats sont que le recensement de 1901 permet aux chercheurs de représenter des ménages autochtones, mais représente seulement partiellement la population autochtone et peut tendre à moins bien représenter les ménages complexes. De plus, le ménage nucléaire est le plus important arrangement résidentiel, pour les Autochtones comme pour les non-Autochtones. Cependant, parmi les autres ménages, ceux comprenant de la famille étendue restent plus communs pour les Autochtones. Plusieurs facteurs laissent tout de même penser que le ménage nucléaire est surreprésenté chez les Autochtones, en raison notamment de biais de la source de données dans sa couverture de la population autochtone et dans sa représentation de leurs ménages. / This research exploits the 1901 Census 100% database, made available by the project The Canadian Peoples, to do a descriptive analysis of the living arrangements of the Indigenous Peoples of Canada at the beginning of the 20th century. The province of Manitoba has been selected for a case study. This research first aims to assess the capacity of the 1901 census to represent the reality of Indigenous households and also aims to do a descriptive analysis of the living arrangements of Indigenous households, for the whole province as well as for the district of residence and the age and sex of the individuals. A comparison is also made with non-Indigenous households. The main results are that the 1901 census allows researchers to represent Indigenous households, but only partially covers the Indigenous population and can tend to represent less accurately complex households. Moreover, the nuclear family household is the most common living arrangement, both for Indigenous and non-Indigenous people. Nonetheless, among other households thosethat include extended family are more common for Indigenous people. Other factors do, however, seem to indicate that the nuclear family household is overrepresented for Indigenous people, notably because of the bias of the data source in its coverage of the Indigenous population as well as its representation of their households.

In this dissertation, I explore the relationship between race and ethnic conflict, demographic composition, and the provision of local public goods across U.S. cities. Specifically, I aim to understand how the preferences of neighboring, often conflicting, racial and ethnic groups are reflected in local political outcomes. To do this, I examine how urban segregation has evolved in recent decades, whether local public finance allocations correlate with differences in racial and ethnic composition, and how animosity between groups influences distributive bargaining decisions and municipal decentralization. Chapter 1 addresses a significant challenge in analyzing demographic composition and municipal finance: the lack of an open-source, robust dataset for evaluating changes in segregation over time. To fill this gap, I introduce the Multi-Level Segregation Database (MLSD), a comprehensive resource that spans the 48 contiguous U.S. states and covers four Census years—1980, 1990, 2000, and 2010. The MLSD provides detailed entropy-based measures of total, within-place, and between-place segregation, offering a rich dataset for analysis. Specifically, the database includes 30,846 Census place observations representing 9,561 unique places in 481 Urbanized Areas (UAs) and 49,778 Census place observations representing 15,986 unique places in 363 Metropolitan Statistical Areas (MSAs). A key contribution of this chapter is the application of the Kullback-Leibler (KL) Divergence metric to quantify the distance between a place’s racial group population distribution and that of its broader region. The utility of the KL Divergence measure is demonstrated through a mathematical proof and a case study of Birmingham, AL. Using the MLSD, I conduct a descriptive analysis that reveals, contrary to previous findings, that segregation within and between places has generally declined from 1980 to 2010 in both UAs and MSAs. These findings challenge the prevailing notion that segregation patterns have predominantly shifted in recent decades from within cities to between cities. Instead, they reveal both forms of segregation have decreased over the four-decade period. This suggests a need for new theoretical approaches to understand what declining segregation truly signifies for group dynamics and political power at the local level. On one hand, lower levels of segregation might indicate improved intergroup relations and a greater willingness among groups to live in closer proximity. On the other hand, these trends could reflect ongoing processes of reshuffling and re-stratification—such as gentrification or the suburbanization of poverty—that perpetuate new forms of inequity. In Chapter 2, I reexamine the relationship between demographic composition and public expenditures using a dataset encompassing over 9,000 cities across six 5-year intervals between 1982 and 2007, resulting in more than 45,000 observations on city-level demographics and public finance outcomes. I test hypotheses suggesting that measures of demographic composition—used as proxies for race and ethnic conflict—explain differences in aggregate spending behaviors of U.S. cities. Specifically, I examine whether increasing diversity and/or segregation at the local level is associated with lower public expenditures. To do this, I extend the MLSD to cover the intercensal years between 1980 and 2010, allowing for the pairing of robust diversity, segregation, and divergence measures with municipal financial data collected by the U.S. Census in years ending in 2 and 7. The Extended Multi-Level Segregation Database (EMLSD) enables the testing of various hypotheses using segregation metrics that are less prone to measurement error and account for changes in municipal boundaries over time. The analysis challenges the prevailing consensus that diverse or segregated communities necessarily experience lower public expenditures. Instead, the findings suggest that the relationship between demographic composition and public spending is more complex than previously understood, with no consistent evidence linking diversity or segregation to significant changes in public goods provision. This study therefore contributes to the literature by questioning long-held assumptions about the impact of demographic heterogeneity on aggregate public finance measures and by emphasizing the importance of considering alternative explanations, such as institutional design and inequitable distributions within cities, in understanding public budgeting outcomes. Finally, in Chapter 3, I develop a two-stage model of legislative bargaining that examines how conflict between groups shapes distributive outcomes and municipal decentralization. The project begins with a single-stage game that recognizes that city council are often comprised of representatives from groups with intense animosities toward each other. Despite strong partisan divisions or cleavages along racial, ethnic, or religious lines, legislatures must still make collective decisions about the distribution of resources. In this chapter, I present a Baron and Ferejohn (1989) style model that captures how animosity shapes distributive bargaining payoffs and the prospects for legislation advanced through cross-party coalitions. I consider a two-party legislature with three members, where each member experiences some level of disutility from allocations to opposing party members. In equilibrium, I find that the expected payoff for majority (minority) party members increases (decreases) as animosity between groups rises. However, only a majority party proposer accrues the increased ex-post payoffs associated with higher animosity, as the proposer can use the threat of cross-party coalition formation to extract additional surplus from co-party members. I then embed the new animosity bargaining framework in a two-stage model of municipal decentralization. The model indicates that the threat of decentralization in stage two can lead to a more equitable—though still unequal—distribution of resources across districts in stage one. When animosity between opposing groups is relatively low, the power to threaten decentralization compels players to form unanimous coalitions and allocate resources to every district, rather than excluding one legislator as seen in the single-stage model. Moreover, the option to decentralize diminishes the proposer’s ability to extract additional resources from other players, ensuring that all players receive the same payoff regardless of whether they are the proposer or a coalition member. Still, like the single-period game, the majority party players are better off than the minority player in equilibrium. Finally, at higher levels of conflict, the ability to initiate a decentralization effort does provide the minority party player with the opportunity to improve its payoffs in a decentralized setting.

Thesis (MTh)--Stellenbosch University, 2000 / ENGLISH ABSTRACT: In South Africa and the rest of the world research is being conducted on the growth and decline of the Christian church and specific denominations. Most congregations keep record of their membership. These data are very useful. However national census data on religion, when available, bring a neutral dimension to the statistics. Census data in most cases provide the only reliable means of establishing religious affiliations of the population at a certain point in time or over a certain period. Since the 1911 national census respondents are being asked to state their religious affiliation. It is essential that the data of these censuses are available in a database. This basic data can then be used to make different combinations and do certain calculations. Different denominations can be combined into families to compare their percentage markets hare of the total population and of the Christian church. If all census data are made compatible with each other it is possible to establish trends over a longer period. When the Census '96 data was made available by Stats SA it was important to prepare the data so that it could be studied en compared with previous census data. This study attempts to do this. It seems that the biggest determining factor in the religious topography of South Africa is the religious shifts that happen amongst the Black population. The biggest percentage Christians belong to the mainline denominations while the Africa Independent Churches have the second most. To compare different denominations with one another over a longer period of time is a difficult task. Inconsistencies with which data are treated as well as the political instability in South Africa are two important factors that hamper the research. The fact that the question regarding religious affiliation was made optional since 1991 complicates the issue even further. The Christian church in South Africa grew from 1911 to 1980 and then started to decline. Membership of both the African Independent Churches and Pentecostal/Charismatic churches are growing. Since 1991 respondents who indicate that they have no religion are growing amongst all population groups. / AFRIKAANSE OPSOMMING: Wereldwyd en spesifiek in Suid-Afrika word navorsing gedoen oor die groei- en kwyntendense van die kerk in die algemeen asook van verskillende denominasies. Meeste denominasies hou rekord van hulle eie lidmaatskap. Hierdie data is baie waardevol. Nasionale sensus data ten opsigte van geloof, aan die ander kant, bring 'n neutrale dimensie aan die statistiek. Sensus data verskaf dikwels die enigste betroubare inligting oor 'n bevolking se geloofsorientasie op 'n spesifieke stadium of oor 'n langer periode. In Suid-Afrika word sedert die 1911 sensus aan respondente gevra om, as deel van die vraelys, hulle geloofsaffiliasie te verskaf. Dit is dus essensieel dat hierdie data van die nasionale sensusse in 'n databasis beskikbaar sal wees. Die basiese data kan gebruik word om verskillende groeperings en berekenings te maak. Verskillende denominasies kan saam gegroepeer word en hulle persentasie markaandeel van die totale bevolking sowel as van die Christelike kerk kan met mekaar vergelyk word. Indien al die vorige sensusse se data op dieselfde manier verwerk word, kan vergelykings oor 'n langer termyn gemaak word ten einde tendense vas te stel. Toe die data van Sensus '96 beskikbaar word, moes dit verwerk word sodat dit enersyds opsigself bestudeer kan word en andersyds met vorige sensusse se data vergelyk kon word. Dit is wat hierdie studie poog om te doen: Om Sensus '96 se data ten opsigte van geloof te verwerk en te vergelyk met vorige nasionale sensusse se data. Dit blyk dat die grootste bepalende faktor in die godsdienstige topografie van Suid-Afrika die godsdienstige verskuiwinge van die Swart bevolking is. Die grootste persentasie Christene behoort aan die Gevestigde Kerke terwyl die Onafhanklike Afrika Kerke die tweede grootste aanhang geniet. Om verskillende denominasies met mekaar oor 'n langer peri ode te vergelyk, is geen rnaklike taak nie. Data wat nie konsekwent hanteer word nie en politieke onstabiliteit sedert die 1960's in Suid-Afrika is maar twee van die problerne. Die feit dat die vraag sedert 1991 opsioneel is, kornpliseer navorsing verder. Die Christelike Kerk het 'n periode van groei van 1911 tot 1980 beleef en daarna begin kwyn. Die lidmate van die Onafhanklike Afrika Kerke sowel as van die Pinkster/Charismatiese Kerke toon 'n stygende tendens. Sedert 1991 styg die respondente wat aangedui het dat hulle aan geen geloof behoort nie, onder al die bevolkingsgroepe.