Spring 2024 Seminars
February 12, 2024
Austin Hutchens
Title: Review - Muscular adaptations and insulin-like growth factor-1 responses to resistance training are stretch-mediated
Abstract: Muscular Hypertrophy training is the use of resistance training to cause damage to muscle tissue cells to stimulate the adaption process and make them bigger. The scientific understanding of hypertrophy has increased exponentially over the past few decades. The article Muscular adaptations and insulin-like growth factor-1 responses to resistance training are stretch-mediated shows just how influential technique can be. This article performed a thorough experiment and analysis to provide evidence that indicates loading muscles at a fully lengthened position is significantly more hypertrophic than shortened muscle positions. Given this information, it is clear that utilizing a full range of motion that accentuates the stretched position of a muscle is optimal for maximizing the hypertrophy of that muscle.
Nicholas Siwanowicz
Title: The Statistical Writing Process
Reference: Chapters 3-4 of Statistical Writing: https://statds.github.io/stat-writing/.
Devin Nordlund
Title: Article Review - A qualitative examination of the impacts of financial stress on college students’ well-being
Abstract: Students pursuing higher education often face significant challenges, particularly regarding financial stress, which can profoundly impact their academic performance and overall well-being. This presentation provides a comprehensive review of a recently published article, “A qualitative examination of the impacts of financial stress on college students’ well-being,” authored by {Moore et al. (2021)}. The study, conducted at a large, private institution, inspects the complex relationship between financial stress and various aspects of the college experience, including academic performance, social well-being, and mental health. This presentation aims to show some of the challenges encountered by many college students today due to financial constraints. By comparing these findings with established data from the University of Connecticut, the goal is to gain deeper insights into the repercussions of financial stress on students and their applicability to the local community.
February 19, 2024
Bolun Li
Title: Review - Odds Estimating with Opponent Hand Belief for Texas Hold’em Poker Agents
Abstract: As we all know, Texas Hold’em is currently the most popular poker game, and the most important thing in Texas Hold’em, besides analyzing the psychology of other people at the table, is to understand the probability of your own hand beating the hands of other competitors. The paper I am analyzing today provides three different win rate algorithms based on the analysis of various hand combinations to assist in decision-making. These methods are the expected win rate algorithm with start hand range (EWR-SHR), the expected win rate algorithm with fold rate (EWR-FR), and the expected win rate algorithm with opponent hand distribution (EWR-HD). These methods improve the accuracy of win rate estimation by combining the opponent model and observed actions to predict the opponent’s hand range or distribution. I will explain the models of various win rate algorithms through this article and display the data model. Finally, I will summarize the reliability and possible uses of these algorithms through the data in this article.
Ian Oller
Title: Improving the Methodology of Court Decision Analysis Modeling
Abstract: The legal system is complicated with many factors playing a role within determining court decisions. If we were to use data to perform an analysis on court decisions, what exactly would that look like? Some of the common datasets collected in order to investigate this are chosen based upon answering how the area of the court, judges of the court, and the defendant may impact the verdict, which each might contain variables that are collected for multiple of those (i.e., race, gender) or specific to a particular category (i.e., previous convictions is specific to defendant data). That brings in the question of how to structure a model that can assess these variables correctly, and although some research has already been done where applications, typically a regression analysis, are done to test for any linear relationships, there is a lot of evidence that suggests that the predictors need to be grouped into independent hierarchical levels, since the differences in areas as well as judges should be considered separately, possibly through a multilevel regression.
Reference: Dhami, Mandeep K., and Ian Belton. “Statistical Analyses of Court Decisions: The Example of Multilevel Models of Sentencing.” Law and Method 10 (2016): 247-266. https://www.bjutijdschriften.nl/tijdschrift/lawandmethod/2016/10/lawandmethod-D-15-00011.pdf
Noah Lindsay
Title: Markov Chains and their Applications to Baseball
Abstract: Everyone who watches or pays attention to baseball/most sports in general knows just how critical the underlying numbers are. And with these underlying numbers we can often help predict, understand, and attempt to strategize how future games will go for the team and/or individual players. As seen in the study I will be discussing, the stochastic process known as Markov chains can be strongly and usefully applied to predicting outcomes in baseball games [at least at the collegiate level]. I will begin the presentation with a brief overview of Markov chains and some of their properties; as the class Stochastic Processes is not a mandatory course in the Statistics Coursework. I will then explain how certain types of Markov chains like absorbing Markov chains and their properties can be applied to predict the outcomes of baseball games [the writer of this paper predicted games for their college at the college of Wooster]. These Markov Chains are going to be used to demonstrate how one can predict the amount of runs the college of Wooster will score on average per game; then will be applied to understanding whether or not this works for individual performance as well. The presentation then will conclude how certain Markov chains can be used to implement and advise certain strategical plays by an individual team/player.
Delun Xu
Title: The Application of Statistics in Daily Life.
Abstract: My presentation delves into the pervasive influence of statistics across various sectors, emphasizing its critical role in data interpretation, decision-making, and research advancement. Through illustrative examples, it elucidates fundamental statistical principles such as descriptive and inferential statistics, hypothesis testing, Bayes’ Rule, and multiple regression analysis. Highlighting the practical application of these concepts in everyday scenarios, the text underscores the indispensable nature of statistical literacy in comprehending and leveraging data to inform decisions in a data-driven world.
February 26, 2024
Neha Jammula
Title: Quantitative Methods and Statistical Analytics in Public Policymaking
Abstract: Statistical and data analytics have recently gained transformative roles in informing evidence-based decision making, including within social research and thereby policymaking. This presentation considers the variables typically captured in big data and examines the quantitative methodologies used in statistical and data analysis, showcasing its capacity to uncover intricate patterns and correlations within society. It explores diverse sources of social data, including surveys, polls, social media, and government records, and how they facilitate the extraction of nuanced insights into societal risks and issues, particularly within demographic differences. This presentation considers the power of statistical data analytics in assessing public sentiments, policy effectiveness, and demographic-specific impacts, emphasizing the benefit of policymakers integrating social scientists’ complex correlations and findings into actionable policies.
- Hossin, M. A., Du, J., Mu, L., & Asante, I. O. (2023). Big Data-Driven Public Policy Decisions: Transformation Toward Smart Governance. SAGE Open, 13(4). https://doi.org/10.1177/21582440231215123.
- Suominen, A., & Hajikhani, A. (2021). Research themes in big data analytics for policymaking: Insights from a mixed-methods systematic literature review. Policy Internet, 13, 464–484. https://doi.org/10.1002/poi3.258.
- Sheng, J., Amankwah-Amoah, J., Khan, Z. and Wang, X. (2021), COVID-19 Pandemic in the New Era of Big Data Analytics: Methodological Innovations and Future Research Directions. Brit J Manage, 32: 1164-1183. https://doi.org/10.1111/1467-8551.12441.
Tyler Lin
Title: How to format your Research Paper
Abstract: With the project proposals coming up it is helpful to know what exactly is demanded by each section of the research paper. This presentation focuses on chapter 4 of the textbook which entails formatting and content suggestions for each one of the sections of the research paper. Additionally, this chapter of the online textbook provides some tips regarding the flow and use of word tenses for research papers as well as rules for how to avoid abusing key terms in statistics. The importance of this chapter is conveyed by its thorough description of how to make professional and presentable publication of your research results. The intention is for viewers to leave to take this advice and apply it not only to their upcoming projects but towards any other publications they may participate in.
Link to slides: https://docs.google.com/presentation/d/1Hjy_eS7NtkWO4j55H20iJQHIcj5VRAas35JJClWLQU8/edit?usp=sharing.
Jingyuan Yao
Title: Prediction of Obesity Level by Using K-Neighbors Classifier and Decision Tree Model
Abstract: This research analyzes a publicly available dataset on Amazon AWS, including gender, age, eating, smoking, and other lifestyle feature. Two methods are selected to predict the obesity level: K-neighbor classifiers can fit the data, train a simple and effective classification model and identify common features of people with different obesity levels; random forest classifier in the decision tree model can handle input samples with high-dimensional features and has good predictive performance, the model parameters can also explore the factors that have big impact on obesity levels. Through these explorations, we can effectively guide the factors that contribute to obesity and assess the possibility of individual obesity.
Zhiyuan Ping
Title: European Soccer Match Outcome Prediction
Abstract: This research aims to outperform bookmaker odds and achieve a profitable return on investment by using a comprehensive soccer dataset obtained from Kaggle. The dataset includes over 25,000 international matches, player statistics from FIFA, and betting data from 2008 to 2016 across 11 European countries. Our methodology integrates data exploration, visualization and advanced machine learning techniques. We utilized and compared Random Forest and Logistic Regression models to forecast match outcomes, and evaluated models accuracy. Moreover, we also explored the impact of regularization in Logistic Regression and emphasized the significance of cross-validation in model evaluation. This study emphasizes the importance of cross-validation in validating model performance. Through this analysis, we expect to explore the potential of strategic betting decisions, gain a deeper understanding of soccer match dynamics, and demonstrate the practical applications of machine learning in sports analytics.
March 4, 2024
Weijia Wu
Title: Effective Math & Stats Communication in Academic Papers
Abstract: This presentation introduces important techniques for effectively communicating mathematical and statistical ideas in academic writing. It covers understanding the audience, clearly organizing the structure of the paper, using precise language and notation, visualizing data, incorporating examples, and using software tools wisely. These techniques aim to improve the clarity of the writing, engage the reader, and communicate complex analyses clearly and concisely.
Ziyue An
Title: The Impact of Recommender Systems on User Behavior and Information Propagation in Social Networks
Abstract: Recommender systems play an important role in influencing user experiences and information distribution in social networks. This study dives into the complex dynamics of how recommender systems impact user behavior and content dissemination across social media platforms. It investigates the mechanisms by which personalized suggestions impact user perceptions and actions, focusing on the balance between tailored information distribution and maintenance of variety and fairness within social networks. The investigation considers a variety of factors, including information filtering, user engagement, and the privacy concerns of tailored suggestions. By studying these patterns, this study hopes to gain insight into the complex influence of recommender systems on social network dynamics and user interactions.
Natalie Horn
Title: Indirect estimation of reference intervals for thyroid parameters using advia centaur XP analyzer
Abstract: Siemens Healthineers, a global leader in medical technology, offers innovative solutions to enhance the diagnosis and monitoring of thyroid disease. Thyroid function parameters such as thyroid stimulating hormone (TSH), free thyroxine (FT4), and free triiodothyronine (FT3) play a crucial role in accurate diagnosis and management. However, standardization and harmonization of laboratory methods remain challenging, impacting clinical interpretation. To address this, it is essential to establish accurate reference intervals (RIs) tailored to specific populations, considering factors like regional iodine intake and analytical methods. Siemens Healthineers emphasizes the importance of using accurate RIs for informed medical decisions, advocating for both direct and indirect methods for RI calculation. Indirect methods, particularly using data from routine patient samples, offer a cost-effective approach to establishing RIs, aligning with international standards set by organizations like the International Clinical Federation Commission on Chemistry (IFCC). In collaboration with healthcare institutions, Siemens Healthineers aims to improve the quality and efficiency of thyroid disease diagnosis and monitoring, ultimately enhancing patient care outcomes.
Cooper Johnson
Title: Analysis of UCONN Football Analytics Study
Abstract: During the Fall 2023 semester, I had the opportunity to work on a study with the analytics department of the UCONN football team. The primary intention of this study was to use data within the Pro Football Focus Ultimate database to discover any statistical insights that would benefit the team. I had completely free reign within this study as I had the choice of honing in on any specific area of the football team. Within this presentation, I will first be discussing what areas of focus within the team that I selected. Then, I will talk about the database that I used along with some of the essential statistics that I deemed relevant to this study. Additionally, I will reveal my key insights that I found regarding the UCONN football teams and its opponents. In order to truly compare and analyze these insights, I created data visualizations to help the viewer better interpret these statistics. After delivering these insights, I proposed strategic gameplans and recommendations for the team going forward that I personally believe will be beneficial. Lastly, after revealing my insights and recommendations, I will take the audience through a more broad perspective on what I performed within the study and how it can be applied to a real-world statistics related field.
References: https://ultimate.pff.com/ncaa/teams/146/offense_stats
March 18, 2024
James Wirchansky
Title: The Impact of Stephen Curry on Basketball and the NBA Landscape
Abstract: In the year 2009, Stephen Curry was drafted by the Golden State Warriors. By the year 2014-2015 he became a star in the league and a league icon. With Curry’s ability to score from anywhere on the court he was able to revolutionize the way the game of basketball has been played for decades. During my research I have found other analytical reviews of Curry’s impact, but I plan referencing and expanding upon their findings into different relevant effects. I will analyze the different impacts Curry has had beyond the rise of three point attempts per game. Using the basketball references database, I will gather league data to perform statistical analyses. During this presentation, I plan to investigate the correlation between 3 point percentage and different important statistics. Additionally, with the change in the way the game is played, front offices, coaches and players had to adjust to the new style. Within this project I plan to dive into specific careers and teams that displayed this change. By conducting statistical analyses and creating data visualizations I will be able to present the scope of Curry’s impact to the audience in an understandable fashion.
Adam Nordmark
Title: Identifying Athletic Potential in High School Football Players
Abstract: This semester, I’ve been working with the UConn Football Program through the Sports Statistics Experiential Learning program in order to aid with their recruiting process. We have been tasked with using data from historical records of prospects in order to find trends that could aid in identifying undervalued recruits that could help boost UConn Football, as well as recruits that have not yet grown to their athletic potential and will flourish in college. In my presentation I will discuss the metrics used to evaluate athleticism in a recruit. I will discuss both physical attributes like height and wingspan as well as performance evaluations like the 40 yard dash and 3 cone drill. I will go over how we identify which past recruits could be seen as successful as well limitations in the datasets.
Braedon Hook
Title: The Advancement of Statistics in European Soccer History
Abstract: The integration of statistical analysis in European soccer history has undergone significant evolution, paralleling the growth of the sport itself. This paper examines the historical trajectory of statistical methods within European soccer, tracing its development from rudimentary data collection to sophisticated analytics. Through a comprehensive review of literature and historical records, this presentation will show the pivotal moments and influential figures that have shaped the landscape of statistical analysis in soccer. Additionally, it explores the impact of technological advancements and data availability on the refinement of statistical methodologies. The analysis also delves into the utilization of statistics in various facets of soccer, including player performance evaluation, tactical analysis, and strategic decision-making by clubs and national teams. Furthermore, this paper discusses the challenges and opportunities associated with the proliferation of statistics in soccer history, including issues of data reliability, interpretation biases, and ethical considerations. By synthesizing historical insights with contemporary trends, this paper offers a comprehensive understanding of the advancement of statistics in European soccer history and its implications for the future of the sport.
Kunpeng Yang
Title: Golden Arches Across Cultures: Understanding McDonald’s Global and Local Consumer Behavior
Abstract: Amidst globalization, the fast-food industry, notably brands like McDonald’s, has attracted significant interest. McDonald’s dominates in the U.S. and has seen remarkable success in markets like China. Yet, despite its global image, differences in marketing and consumer behavior exist between these nations. This study aims to dissect the distinctions between McDonald’s in China and the U.S., focusing on product adaptability, advertising, and consumer behavior. The intent is to grasp how global brands localize in varied cultural environments. Historically, McDonald’s transitioned from a U.S. local brand to a global powerhouse. The U.S. formed its foundational market, while China presents both vast opportunities and challenges. Key questions addressed include McDonald’s product adjustments for diverse cul- tures and whether advertising should be country-specific. To answer these queries, this paper mixes qualitative and quantitative research. Data will be sourced from consumer surveys and interviews, and a thorough analysis of McDonald’s advertising in both nations. Sec- ondary data will also be examined for a holistic view. This approach aims to pinpoint McDonald’s strategic adaptations and potentially guide other global brands in localization endeavors.
March 25, 2024
Sean McKenna
Title: The Applications of Statistics in Business Decision Making
Abstract: In today’s dynamic business environment, data has emerged as the cornerstone for informed decision-making. Statistics, as a powerful analytical tool, plays a pivotal role in extracting actionable insights from data to drive strategic business decisions. This presentation delves into the diverse applications of statistics in various facets of business decision-making processes. From sales forecasting to risk analysis, quality control to marketing analytics, statistics offers a robust framework for businesses to navigate complexities and optimize performance. Through real-world examples and case studies, this presentation illustrates how statistical techniques empower organizations to unravel patterns, identify trends, and anticipate future outcomes. Furthermore, this presentation explores the challenges and limitations associated with statistical analysis in business contexts, providing valuable insights into the practices for leveraging these techniques effectively.
Sha Jameson
Title: TikTok threat? Real or myth?
Abstract: TikTok is the fastest growing social media platform today. With over a billion users worldwide and 110 million users in the US, Congress has constantly reminded us it is a “Chinese app”. For the past couple of years congress has been adamant on passing a TikTok ban because it consistently claims it is an entity of the Chinese Communist Party and a “Chinese Spy” by collecting our data. However, what kind of data can an app gather? Is this different from any other US based company such as Meta’s Facebook and Instagram or Snapchat data collection? What is the real reason behind this ban? We will use data analytics to answer these questions.
Yanzhao Wang
Title: A study of Car Accidents in New York City
Abstract: Road traffic accidents kill an estimated 1.19 million people each year, making them one of the deadliest accidents each year and a leading cause of death for children and young people. While traffic accidents cause a large number of deaths, 20 million to 50 million people are injured. My research is based on data provided by the New York City Police Department, updated from 2012 to the present, and includes the date, time, location, vehicle model, number of deaths, number of injuries, and so on. I use R for data analysis and modeling to study whether today’s high-tech cars have effectively reduced the probability of injury and death for drivers and pedestrians after the rapid development of science and technology compared with the past, and whether technological innovation has ensured the safety of drivers and pedestrians by enhancing safety measures.
Jyothsna Vijayaraghavendra
Title: Article Review - Quantitative Analysis of Online Shopping Perceptions
Abstract: In an era where digital commerce is reshaping consumer habits and market dynamics, understanding the evolving perceptions of online shopping remains crucial. Despite the various studies done on online shopping, there remains a gap in understanding how the frequency of online shopping influences consumer perceptions of digital service quality, directly affecting their likelihood of repeat online purchases. This article, authored by Ilona Pawełoszeka and Paula Bajdor, offers a comprehensive analysis of the relationship between shopping frequency and perceived service quality among Polish consumers through a multivariate approach. Through this presentation, I aim to provide a clear and informative overview of the study’s methodologies, findings, and subsequent discussions, encouraging a deeper understanding of the complex relationship between online shopping frequency and service quality perceptions.
April 1, 2024
Sakeena Aimandi
Title: Understanding Large Language Models: A High-Level Overview
Abstract: Large Language Models(LLMs) and generative AI are a recent craze in the tech world, offering unprecedented capabilities that seem to mimic human intelligence. Given its complexity, this advanced technology is seldom understood by the average person. In this presentation, I aim to demystify LLMs by offering a high level overview of their structure and function. Beginning with a discussion of the general concept behind large language models, I will detail their architecture, the training process, and statistical foundations that support their operation. Moreover, I will address ethical considerations and real-world applications of LLMs, as well as describe future directions for this novel technology. Through this presentation, I hope to increase transparency regarding large language models in contemporary AI research and applications.
Alex Wu
Title: Article Review: Better Offensive Strategy in Basketball: A Two-Point or a Three-Point Shot?
Abstract: This research delves into the relationship between shooting tendencies, offensive strategy, and game win probability in basketball, focusing on top level competition in the NBA as a leader in basketball strategy. Through analysis of shooting trends and offensive tendencies, we will see how shooting more three pointers as opposed to two pointers can lead to a higher game win probability. By comparing these trends between weak and strong teams, we can see the differences in how playoff teams approach NBA offensive strategies as opposed to lottery teams. Analyzing these trends will reveal the direction at which NBA offensive strategies are trending and how basketball offensive strategy as a whole will trend in the future.
Reference: Gou H, Zhang H. Better Offensive Strategy in Basketball: A Two-Point or a Three-Point Shot? J Hum Kinet. 2022 Sep 8;83:287-295. doi: 10.2478/hukin-2022-0061. PMID: 36157952; PMCID: PMC9465745.
Sean Smith
Title: The Impact of Key Match Statistics on a Teams Success in the English Premier League: A Statistical Analysis
Abstract: This study aims to explore the intricacies of football match outcomes in the English Premier League (EPL), particularly from 2011 and 2019. Using an extensive collection of data of 3,327 EPL matches, the study seeks to explain how match statistics can impact a teams likelihood of success. I aim to understand the effects of red and yellow cards, as well as the importance of scoring the first goal in the match, on a team’s chances of winning. The study uses advanced regression analysis, including linear and logistic models, to estimate the impact of these unexplored variables on team success. The hypothesis states that disciplinary actions (red and yellow cards) and the time of the first goal are important elements that can have an inverse effect on match outcomes.
This research is critical for understanding football match dynamics beyond the typical statistics, as it offers findings about league strategic advantages. The study not only improves the scientific field of sports analytics, but it also has real-world implications for clubs, coaches, and analysts who seek success in the competitive atmosphere of the EPL. By emphasizing the importance of these often-overlooked match statistics, this research contributes to a better understanding of the factors which impact team performance and offers new paths for strategic planning in professional football.
April 8, 2024
Nina Longobardi
Title: The Explorations of Social Media Engagement through Statistical Analyses
Abstract: Social media is an ongoing and evolving source of information that communities, businesses, and individuals continue to use. Throughout my research, I wanted to delve deeper into the engagements of social media and the different factors that are associated with it. What is social media engagement? Why is it so important? Understanding the significance in measuring audience interaction and content effectiveness through key metrics such as likes, shares, comments, reposts, and reaches is a crucial part of social media for business owners, influencers, etc. Through surveys and examples, I will highlight the importance of context in social media engagement.
Jeeva Shunmugaraja
Title: Goal-Driven Success: Integrative Statistical Analysis of Soccer Performance Across Competitive Environments
Abstract: In the massive field of professional soccer, statistical analysis has become a pivotal tool used across facets of the sport such as player performance analysis, team strategy, player development, and so on. While there are a number of studies that have been done on the landscape of statistical analysis in soccer, there is little data available that provides a sound analysis of the plethora of variables within the sport and the different ways in which they can be analyzed. This thesis integrates methodologies from various statistical studies to analyze key performance indicators within descriptive, comparative, predictive, and contextual variable groups across a variety of leagues and environments across the globe. We observe that these KPIs —ranging from frequency of player movements to field methodologies and tactics— interplay to shape match outcomes. By taking a closer look into the unique methodologies and analysis methods used, including advanced predictive modeling and regression analysis, this thesis will allow for the examination of the complex relationships between a variety of KPIs and match outcomes across different soccer environments. The outcomes contribute a refined understanding of soccer analytics and their growing effectiveness, while also providing actionable strategies for enhancing team performance in varying league contexts.
Wenhao Ruan
Title: Analysis of NYC’s Merit-based Specialized High Schools: Education, Race and Wealth Discrimination
Abstract: The demographic profile of New York City and Boroughs are highly diversified in types of ethnicity, wealth, and especially education. The state presents a competitive high school structure which allows students to take an optional entrance examination to be admitted into one of eight NYC-supported Specialized High Schools (LaGuardia High School excluded due to audition-based admissions). I will be conducting a data analysis based on NYC Education Department’s School Quality Reports from 2014-2019 and examine trends of discrimination. Overall, the results have shown a significant gap between schools with differences in the Economic Needs Index. Education for a large portion of the students in high school has been increasingly skewed and unfair throughout the years before Covid.
Vincent Xie
Title: Taylor Swift’s Influence on the 2023 NFL Season: Viewership and Merchandise Sales
Abstract: In 2023, the NFL season saw a notable shift in viewership and merchandise sales, particularly Travis Kelce’s jerseys, attributed to pop icon Taylor Swift’s involvement. This study focuses on analyzing viewership data during regular season games and the Super Bowl, along with sales trends, to determine Swift’s impact. Initial findings suggest that there is a small correlation between Swift’s presence and increased viewership as well as a significant rise in Kelce’s jersey sales. These insights highlight the potential of integrating entertainment figures in sports marketing to broaden audience reach and engagement.
April 15, 2024
Jing Wang
Title: Investigation on imbalanced Cox regression based on simulations
Abstract: Data imbalanceness brings challenges for statisics and machine learning. Although, researches have studied a lot on data balancing for classification. Imbalanced survival data is a topic that has been rarely toughed. In this project, we first introduced some recent works on highly imbalanced binary data and then turn to imbalanced survival data. For survival data, a potential connection between imbalanced binary data and survival data has been noticed. We thus propose to borrow ideas from balancing binary data to handle survival data. A serial simulation studies for survival data have been done to investigate potential issues caused by data imbalanceness and possible approaches to handle challenges due to data imbalanceness.
References: Wang, HaiYing. “Logistic regression for massive data with rare events.” In International Conference on Machine Learning, pp. 9829-9836. PMLR, 2020.
Heeju Lim
Title: Heckman Selection – Contaminated Normal Model
Abstract: The Heckman selection model is widely used to address sample selection in econometric analysis. Traditionally, it relies on the assumption of normal error terms. However, real data diverge from this assumption in the presence of heavy tails. Recent advancements have introduced more adaptable frameworks based on the Student’s-t distribution. This presentation introduces a novel Heckman selection model using a bivariate contaminated normal distribution for the error terms. We present an efficient ECM algorithm for parameter estimation with closed-form expressions at the E-step based on truncated multinormal distribution formulas. Through simulation studies and real data analysis, we compare our proposed model with the normal and Student’s-t counterparts and showcase the performance of our model. The proposed algorithms are implemented in the R package HeckmanEM.
April 22, 2024
Ella Stanwood
Title: How Bayesian Hierarchical Modeling Can Be Employed to Analyze Discrimination Within Health Care Accessibility
Abstract: This presentation explores the multifaceted interactions between socioeconomic status, demographics, and health outcomes using Bayesian hierarchical modeling. By leveraging a nationally representative data source, the National Health and Nutrition Examination Survey. Bayesian hierarchical models are employed to uncover intricate relationships that may be obscured by traditional regression techniques. Bayesian methods allow for the incorporation of prior information and uncertainty quantification, offering more reliable estimates of model parameters. Through this comprehensive analysis, we can provide insights into the complex interplay between socioeconomic factors and health outcomes to allow for more accessible health care.
Chorok Kang
Title: “Mapping the Economic Landscape”: A statistical analysis of shifts triggered by Remote work
Abstract: This analysis provides a comprehensive statistical review of the economic transformations instigated by the widespread adoption of remote work globally, focusing on commercial real estate, urban transportation, and small businesses. We use basic statistical tools to analyze trends in office space demand, public transportation usage, and changes in small business revenues. By applying regression analysis and time series analysis, we identify the economic effects of a more distributed workforce and forecast potential future shifts. The goal is to provide clear, actionable insights for decision-makers looking to adapt to new economic realities driven by remote work. This simplified analysis aims to offer a straightforward understanding of how remote work is transforming these sectors and what that means for future planning.
Nadia Alsubai
Title: An Introduction To The Meta-Analysis: A Brief Summary of “Doing Meta-Analysis with R: A Hands-On Guide” by Mathias Harrer, Pim Cuijpers, Toshi Furukawa, and David Ebert
Abstract: This presentation will provide a basic introduction to the meta-analysis using a brief summary of “Doing Meta-Analysis with R: A Hands-On Guide” by Mathias Harrer, Pim Cuijpers, Toshi Furukawa, and David Ebert. The aim of this presentation is to define a meta-analysis, establish why a meta-analysis may be valuable to a researcher, highlight common components of a meta-analysis, and discuss the pitfalls and criticisms that a researcher may encounter when completing a meta-analysis. Please note that this presentation does not cover R programming, and it does not exhaust the topic of meta-analyses. Excluded concepts and statistical methods may be mentioned for future discussion, if time allows.
Reference: Harrer, M., Cuijpers, P., Furukawa, T.A., & Ebert, D.D. (2021). Doing Meta-Analysis with R: A Hands-On Guide. Boca Raton, FL and London: Chapman & Hall/CRC Press. ISBN 978-0-367-61007-4.