Although variability is of fundamental concern and interest to statisticians, often this does not get communicated to students who are taught instead to view variability as a nuisance parameter. A brief survey of a few case studies, as well as a recounting of some history, shows that variability is worthy of study in its own right, and examination of variability leads to insights that might have been missed had we focused all of our attention on the "trend" of the data. As one of the key components of statistical thinking, variability deserves more prominence in the classroom.

Understanding student learning is an open problem in the teaching of introductory statistical programming. Formative learning is observed when analyzing the

interaction between the learning environment and the student. This can be operationalized by viewing how students respond to the error messages they receive

while programming. Current California Common Core State Standards: Mathematics

(CA CCSSM) highlight perseverance as an overarching habit of a productive

mathematical thinker. Perseverance can be measured as a type of formative

learning by measuring the time and attempts that students use to correct errors.

The MOBILIZE project promotes statistical programming at the high school level

in the Los Angeles Unied School District while following the CA CCSSM. Logs

of high school student statistical programming during the 2013-2014 school year

were collected along with the errors that occurred. Using these logs, error blocks

were formed that follow a student's interaction with an error message. With the

error blocks we were able to observe perseverance by students on multiple days

of curriculum. Our findings suggest that there was an increase in perseverance,

because students increased in time spent and attempts made to correct an error

as the number of days of programming curriculum increased. Additionally, students who showed more perseverance were more likely to eventually fix an error.

Descriptive variables were explored to provide background for the variation in

student programming errors.

With ever improving streaming technologies and accessibility to video games, it comes as

no surprise that competitive gaming or eSports have blown up in recent time. League of

Legends, former gaming startup Riot Games' sole intellectual property, has the title most

popular eSport in the world with a thriving competitive scene and international competition

that rivals traditional sports leagues such as the MLB, the NBA and the NFL [Sta, 2013].

With the high stakes involved in the burgeoning eSports industry, it is imperative that these

organizations develop methods that can dierentiate players based on their skill through

their in-game performance metrics and determine potential acquisitions. Additionally, we

want to leverage the data within Riot Games' databases on how the general playerbase

approaches the game to determine what how in game performance metrics change as player

skill increases. The end goal of this analysis is to create a method to gauge team performance

and assess weak links in strategy.

With the rapid advancement of technologies in the recent few decades, more and more students are entering the Science, Technology, Engineering, and Mathematics (STEM) field in college. As one of the leading universities in the world, University of California, Los Angeles (UCLA) has a strong group of undergraduate programs in STEM. With the size of the STEM community expanding rapidly, it is important that we examine the sentiment of the community through some statistical analyses. Among various approaches, sentiment analysis of the course reviews could help us understand the feedback from the student community, which could also provide us with many interesting insights. In this paper, we will look at the methodologies and results from applying the sentiment analysis pipeline on a corpus with around 7000 course reviews collected from the UCLA Bruinwalk website, with the goal of analyzing if students were happy (positive sentiment) or unhappy (negative sentiment) towards the STEM courses they have taken. As we shall see, because the reviews obtained from the Bruinwalk website were not initially annotated, the sentiment analysis pipeline consisted of two main components, review annotation and sentiment analysis modeling. Throughout the study, various data visualization techniques were utilized to help us obtain a better understanding of structure of corpus, including the features and the sentiment annotation. Multiple NLP model architectures, such as CNN+LSTM, Transformer, and the state-of-the-art BERT and DistilBERT architecture, were established and compared to optimize the sentiment prediction performance. The results from the sentiment analysis showed that around 60\% of the reviews collected contained a positive sentiment, the sentiment of the reviews was positively associated with the student grades, together with several other interesting findings.

With higher education becoming more competitive year by year, along with the job andhousing market, it is no surprise there is so much discourse about issues with the college admissions process. Furthermore, the Supreme Court’s decision to end race considerations within the admissions process makes it evermore important to understand the implications race and other demographic factors have for students’ educational outcomes. Due to the ever-expanding STEM fields and their reputation for having high-paying jobs, this thesis examines the role demographic background plays in predicting if a student is considering a STEM major.

In the past decade, marijuana laws have come under scrutiny with Los Angeles serving as an epicenter and testing ground for regulations concerning the use and distribution of marijuana. The topic is polarizing, acting as a divide amongst the inhabitants of Los Angeles. In this paper, an analysis of crimes near legal marijuana dispensaries is pursued through a few modern statistical tools. This paper is motivated and built upon the 2011 research conducted by RAND authors Chang and Jacobson which sparked a commotion amongst supporters of anti and pro marijuana laws. An alternative view of Chang and Jacobson's result is pur- sued through the quantification of temporal sources of variation in the sampling distribution of a key regression parameter as well as classical spatial statistical tools. Further, we will make use of Voronoi tessellations in two ways. First, we will provide an alternative spatial regularization when interpreting crimes clus- tered around the marijuana dispensaries. Lastly, an application of the Voronoi Estimator, which estimates the intensity of an inhomogeneous point processes, will be implemented.

The Los Angeles Police Department (LAPD) often faces the task of predicting crime before it happens to better safeguard the community. The LAPD has a historical record of over a million incidences. Statistical analysis of some of these incidences can aid the LAPD in crime prevention by identifying patterns and trends. The LAPD can then allocate resources accordingly. This thesis looks at crimes in Los Angeles from 2005 to 2009 though focuses on crimes in the most recent year. It uses clustering and regression techniques in addition to contingency tables and Chi-squared tests to find clear crime trends. Historical weather data and demographic data from the U.S. 2010 Census are included as part of the analysis in order to determine some external factors that may affect crime. This paper looks particularly at trends over the course of the year as well as victim characteristics.