Search

Research suggests that mathematics achievement is associated with future socioeconomic success and opens the possibility of social mobility for many individuals. However, disparities in mathematics achievement indicators expose inequitable opportunities and inconsistent access to appropriate educational models for a broad range of students. This study used the principles of Improvement Science to explore the problem of underachievement in numeracy for English language learners. This paper builds on Barton et al.'s ecologies of parental engagement framework and Yosso's community of cultural wealth framework to examine how a school district empowered families to use linguistic capital when engaging with their children about mathematics. The convergent mixed methods design examined how participating in Parent University influenced parental beliefs about mathematics, their use of time engaging with their children about mathematics, and their understanding of linguistic capital. Findings indicate that parents' beliefs about the usefulness of mathematics and their capacity to engage in mathematical tasks influenced how they used their time engaging with their children about mathematics. The parents' affirmation of native language capital impacted their use of mathematical discourse with their children at home. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

The three papers which make up this dissertation form part of a larger project, which aims to solve the `access problem' for realism about mathematics by providing a clear and plausible example of what a satisfying explanation of human accuracy about objective mathematical facts could look like. They fit into this project as follows., Philosophy

This paper presents findings on research-based interventions that were employed as part of a semester-long PDSA (Plan-Do-Study-Act) cycle at Davis & Elkins College (D&E) aimed at decreasing math anxiety while increasing student achievement on high-stakes standardized assessments with a focus on growth mindset and productive struggle. Based on data from 2015-2018, the success rate of preservice elementary teachers (K-6) on the mathematics portion of the Praxis Core Academic Skills for Educators exam was approximately 30% on first attempt at D&E. Drivers of this problem of practice include the attitudes, beliefs, and mindsets of test takers. Often, students who have a history of negative experiences in mathematics suffer from a cycle of math anxiety involving a lack of confidence and avoidance of mathematics. Due to the COVID-19 pandemic, students were enrolled in online courses and the interventions included two free online courses: Stanford University's "How to Learn Math: For Students" and Khan Academy's "Praxis Core Math." The interventions aimed to enable students to experience a more positive, productive cycle of math achievement. Results indicate that these methods were more helpful for some groups of students than for others. The characteristics of each group are noteworthy for future study. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

While significant scholarly attention focuses on the development and testing of theoretically grounded models of the college choice process at the undergraduate level, far less research explores the area of graduate school enrollments. Graduate school choice, which is defined for the purposes of this paper as the decision to pursue any post-baccalaureate degree program at the masters, doctoral-research, or doctoral-professional practice level, is shaped and determined by a number of individual and organizational level characteristics. The relative influence and predictive power of these variables in modeling graduate school choice behaviors is of significant theoretical and practical interest, given the role graduate education plays in access to certain career paths, career levels, and lifetime earnings. This paper addresses a gap in the literature by advancing a conceptual framework of graduate school choice derived from the work of Perna (2006), drawing significantly from human capital theory and incorporating the salient concepts of cultural and social capital. The methodology employed is a set of generalized hierarchical linear models in which students are nested within undergraduate institutions. The dependent variables of interest were graduate school aspiration, application, and enrollment. The dataset analyzed was the 2000/01 Baccalaureate & Beyond Longitudinal Study (B&B:00/01). The nationally representative B&B:00/01 study is comprised of approximately 10,000 students who received a baccalaureate degree from one of the over 1,000 institutions sampled between July 1, 1999 and June 30, 2001. In this study, individuals most likely to aspire to, apply for, and enroll in graduate school were dependent students who obtained high undergraduate grade point averages, majored in the humanities, social or behavioral sciences, mathematics, or life and physical sciences, and attended a master's or doctoral institution. This study also found that, when controlling for all other variables in the models, African-America and Hispanic students are more likely to engage in the graduate school choice process than white students. A key variable of interest, undergraduate indebtedness, does not affect graduate school choice, when accounting for all other variables in the model. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

With the increasing wealth of information on the Web, information integration is ubiquitous as the same real-world entity may appear in a variety of forms extracted from different sources. This dissertation proposes supervised and unsupervised algorithms that are naturally integrated in a scalable framework to solve the entity resolution problem, which lies at the heart of the information integration process. This dissertation focuses on two incarnations of the entity resolution problem that arise in the data mining and natural language processing areas. First, "name disambiguation" occurs when one is seeking a list of publications of an author in a digital library, who has used different name variations and when there are multiple other authors with the same name. We present an efficient integrative framework that disambiguates the extracted author metadata from paper headers in a divide-and-conquer fashion: based on the metadata records extracted from paper headers, a blocking method retrieves candidate classes of authors with similar names and a density-based clustering method, DBSCAN, clusters the records by author. The distance metric between papers used for clustering is calculated by an online active selection Support Vector Machines algorithm LASVM. We prove that by recasting transitivity as density connectivity in DBSCAN, transitivity is guaranteed for core points. The method achieves high accuracy on a manually labeled dataset and readily disambiguates about a million author metadata records in CiteSeer, which paves the way for the fielded search by author name feature in CiteSeer[superscript X]. Second, as a key step towards document understanding in natural language processing, we investigate the problem of "cross document coreference" (CDC), which aims to decipher the true reference of a named entity across the boundary of documents. This dissertation presents a novel cross document coreference approach that leverages the profiles of entities which are constructed by information extraction tools and reconciled using a within-document coreference module. We propose to match the profiles by using a learned ensemble distance function comprised of a suite of similarity specialists. We develop a kernelized soft relational clustering algorithm that makes use of the learned distance function to partition the entities into fuzzy sets of identities. Evaluation on a large benchmark collection shows that the proposed methods achieve competitive coreference results. We further discuss the details of the implementation of the CDC and web person search system. This dissertation surveys the literature on author name disambiguation in citations and paper headers, citation matching and cross document coreference. Additionally, we explore the social networks of the disambiguated authors, performing a comprehensive study of the network and community level characteristics and proposing a stochastic model to predict collaborations of individuals. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

Link failures in wide area networks are common and cause significant data losses. Mesh-based protection schemes offer high capacity efficiency but they are slow, require complex signaling, and instable. Diversity coding is a proactive coding-based recovery technique which offers near-hitless (sub-ms) restoration with a competitive spare capacity requirement with respect to other techniques. In this thesis, an optimal algorithm is developed for pre-provisioning of the static traffic using both systematic and non-systematic diversity coding. A Mixed Integer Programming (MIP) formulation is developed for systematic diversity coding which requires many fewer integer variables and constraints than similar optimal coding-based formulations. In all scenarios, diversity coding results in smaller restoration time, higher transmission integrity, and much reduced signaling complexity than the existing techniques in the literature at the expense of slightly lower capacity efficiency than Shared Path Protection (SPP) and p-cycle protection. The drawbacks of the earlier implementations of diversity coding are the low scalability and restricted coding structure. To overcome these, a simple column generation-based design algorithm and a novel advanced diversity coding technique to achieve near-hitless recovery over large networks are proposed. The design framework consists of two parts: a main problem and sub problem. Simulation results over large networks with arbitrary topologies suggest that both the novel coding structure and the novel design algorithm lead to higher capacity efficiency for near-hitless recovery. Moreover, this thesis presents a bidirectional coding-based scheme, named Coded Path Protection (CPP). In addition to a systematic approach of building valid coding structures, this paper presents an optimal and simple capacity placement and coding group formation algorithm. The algorithm converts the sharing structure of any solution of an SPP technique into a coding structure with minimum extra capacity. Simulation results confirm that the CPP is significantly faster than both SPP and Shared Link Protection (SLP) and more capacity efficient than SLP. This thesis also investigates the dynamic provisioning of the traffic on-demand. In dynamic provisioning, an Integer Linear Programming (ILP)-based optimal algorithm covers both of the systematic and non-systematic diversity coding. The performance of proposed dynamic provisioning technique is similar to that of pre-provisioning. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

Conditional cash transfer (CCT) programs are increasingly being used through the developing world to reduce inequality, break the intergenerational poverty cycle, and build human capital. These programs vary by country but typically make cash transfers conditional upon children meeting certain healthcare and educational standards. While previous research finds that these programs are effective at increasing school enrollment and attendance, there is little evidence that it actually increases academic achievement. This paper examines the impact of Familias en Acción, a CCT in Colombia, on the academic achievement of students from households receiving the subsidies, and the impact of the program on school quality. Program evaluation data is used to conduct a logistic regression to measure the programs effects on literacy and grade retention, and national standardized test results are used to conduct an Ordinary Least Squares Regression on academic performance at the school level. The results show that Familias en Acción increases literacy and decreases grade retention amongst the students receiving the subsidy. While the students participating in the program are not found to have no impact on schools, the overall increase in enrollment does have a negative impact on school level math scores. These findings support the growing use of CCT strategies to increase human capital and alleviate poverty. However, the program still allows for geographic and socio-economic differentials in academic attainment to persist. Additional program options should be investigated to address these concerns. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

The fast growth of global routing table size has been causing concerns that the Forwarding Information Base (FIB) will not be able to fit in existing routers' expensive line-card memory, and upgrades will lead to a higher cost for network operators and customers. FIB Aggregation, a technique that merges multiple FIB entries into one, is probably the most practical solution since it is a software solution local to a router, and does not require any changes to routing protocols or network operations. While previous work on FIB aggregation mostly focuses on reducing table size, this work focuses on algorithms that can update compressed FIBs quickly and incrementally. Quick updates are critical to routers because they have very limited time to process routing updates without impacting packet delivery performance. We have designed three algorithms: FIFA-S for the smallest table size, FIFA-T for the shortest running time, and FIFA-H for both small tables and short running time, and operators can use the one best suited to their needs. These algorithms significantly improve over existing work in terms of reducing routers' computation overhead and limiting impact on the forwarding plane while maintaining a good compression ratio. Another potential solution is to install only the most popular FIB entries into the fast memory (i.e., an "FIB cache"), while storing the complete FIB in slow memory. In this paper, we propose an effective FIB caching scheme that achieves a considerably higher hit ratio than previous approaches while preventing the "cache-hiding" problem. Our experimental results using data traffic from a regional network show that with only 20K prefixes in the cache (5.36% of the actual FIB size), the hit ratio of our scheme is higher than 99.95%. Our scheme can also efficiently handle cache misses, cache replacement and routing updates. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

The purpose of this study was to determine if there were significant statistical differences between scores on constructed response and computer-scorable questions on an accelerated middle school math placement test in a large urban school district in Ohio, and to ensure that all students have an opportunity to take the test. Five questions on a placement test given in the spring of 2018 were changed into questions that could be graded by computer for students taking the test in the spring of 2021. The data was examined for each test first using whole group data. Significant statistical differences were found in each of the five questions, with four questions having a higher mean as a constructed response item and one having a higher mean as a computer-scorable item. No significant differences were found when scores were analyzed by gender, and only one question had statistical difference when analyzed by ethnicity. Correlation was calculated between each student’s placement test score and their Math 8 state test score. Similar correlations were found between the two assessments during both testing years. To ensure that all students are provided an opportunity to succeed in an accelerated math pathway, all must need to be given access to helpful pre-testing materials and encouraged to take the placement test. Recommendations were made to continue using computer-scorable questions to provide an opportunity to identify as many qualified candidates for acceleration as possible.

A study was conducted at Pima Community College to determine the relationship between the final grade received by students in an introductory, algebra-based physics course (PHY 121) and their scores on the reading, writing, and mathematics portions of the college's nonmandatory assessment test. Between 1983 and 1988, 639 students obtained a final grade in PHY 121. Of these, 558 took the Nelson-Denny Reading Test (NDRT), 280 took a locally developed writing test, and 431 took a locally developed math test. When correlation coefficients were examined, the test result most highly correlated with successful completion of PHY 121 was the NDRT score, followed by the Pima College Math Test score. Other study findings included the following: (1) on average, students who took the NDRT read at a 13.8 grade level; (2) 84.2% of the students entered PHY 121 without the prerequisite math skills, indicating that either the prerequisite was inappropriate for the course or the math portion of the test was an inaccurate measure of student ability; and (3) 48.8% of the students who took the writing test were placed in a developmental writing course, compared to 17.8% who were placed in Freshman Composition. (AJL)