Analyzing Data with R: How Much do my Findings Depend on my Analytical Decisions?
Instructor: Michael Truong, MA (mtruong@yorku.ca)
Course Description: This course is an introduction to the use of the multiverse and targets R packages to address two major challenges faced by scientists. The first challenge is that after collecting the data, there are so many possibilities in how one should clean their data that one might reasonably be nervous about how dependent their eventual findings depend on earlier analytical decisions. For example, does keeping an extreme outlier affect my results? The second challenge relates to automatically cleaning and re-analyzing new data. With an on-going data collection process and changes in how the data are prepared for analysis, is there a way to automate the complete re-analysis of the data? For example, I've just collected 50 more participants, is there a way for R to automatically re-run all the data cleaning and analysis code? This workshop will illustrate each of these challenges and outline the latest R solutions for addressing them. Attendees are encouraged to bring their previous projects, so that they may be retrofitted with these solutions. Retrofitting either a previous project or adjusting a workshop-provided project with these solutions will count towards digital credentialing. Familiarity with R is assumed, but not necessary to appreciate the discussed challenges and current solutions.
Date: Wednesday, November 22, 2023 (11:30 am - 2:30 pm)
Format: Online
Cost: $15
Eligible for a Digital Credential/Badge: Yes
Crash Course in Factor Analysis
Instructor: David Flora, PhD (dflora@yorku.ca)
Course Description: The general purpose of exploratory factor analysis (EFA) is to develop a model which represents the pattern of correlations among a potentially large number of empirically observed variables in terms of a small number of unobserved, or latent, variables, which are referred to as factors. In modern research, EFA is most commonly used to investigate the underlying dimensional structure of a set of items in a test or questionnaire: What are the systematic factors that influence the item responses and how strongly are these factors related to the items? This course will provide a broad overview of the statistical foundation of EFA, including estimation methods, how to determine the optimal number of factors (i.e., model comparison), and factor rotation. The course will also address how these methods are adapted to ordinal, categorial variables such as those elicited from Likert-type questionnaire responses. Several example analyses will be presented using R software.
Dates: November 3 & 10 (Fridays) - 2:30 - 5:00 pm
Format: In-person (159 Behavioural Science Building, York University)
Cost: $25
Eligible for a Digital Credential/Badge: No
Data Management in R
Instructors: Arjunvir Singh, MA (arjun10@yorku.ca); Gabriel Crone, BA (gcrone14@gmail.com)
Course Description: This course is designed to equip you with essential skills in handling data effectively, covering various critical stages of data management. Throughout this program, you will gain proficiency in the following key areas:
• Importing Data in R: We will start by teaching you how to import data into the R programming environment, setting the foundation for your data management journey.
• Understanding Data Formats: You will delve into the fundamental data formats, including scalars, vectors, data frames, and lists. This knowledge forms the bedrock of data manipulation.
• Data Cleaning with Tidyverse: Learn the art of data cleaning using the powerful tidyverse package, ensuring that your data for further exploration.
• String Manipulation with Regex: Explore the intricacies of string manipulation using Regular Expressions (Regex) in R, a crucial skill for data transformation.
In addition to these core topics, and if time permits, this course will also provide you with a brief introduction to data visualization, data sharing techniques, and the utilization of various data storage options, including SQL databases.
By the end of this course, you will have the knowledge and practical skills to confidently manage and manipulate data for your analytical needs.
Dates: October 18 & 25, November 1 (11:30 - 2:30)
Format: Online
Cost: $45
Eligible for a Digital Credential/Badge: TBA
Mixed Effects Models
Instructors: Arjunvir Singh, MA (arjun10@yorku.ca)
Course Description: Mixed Effects Models are the Swiss Army knife of statistical analysis, allowing us to tackle the most intriguing data puzzles. Whether you're dealing with data that behaves differently across various groups, categories, or units, or you simply want to explore the nuanced relationships within your dataset, Mixed Effects Models are here to guide the way. In this course, you will learn everything from the fundamentals to the advanced applications of Mixed Effects Models. You'll gain a deeper understanding of the theory underlying these models and learn how to practically implement them using several packages in R. Our journey begins by exploring the core concepts of Mixed Effects Models, distinguishing between fixed and random effects. Think of fixed effects as the well-rehearsed actors on the main stage, representing the constants you want to study. Meanwhile, random effects are the intriguing backstage personas, influencing the show in subtle yet significant ways. As we progress, you'll discover how to model the distribution of random effects and analyze the conditional distribution of observed data, making sense of the complex interplay between different factors within your dataset.
NOTE: This workshop will take a more interactive approach compared to traditional workshops, the goal is to allow users to interact with several aspects of the mixed effects models and see how things change. All users will gain further access to applications and resources relating to mixed effects models along with detailed code.
Dates: November 14, 21, 28 (Tuesdays) - 11:30 am - 2:30 pm
Format: Hybrid (In-person/Online)
Cost: $45
Eligible for a Digital Credential/Badge: No
