Contact Us
For more information about the Data Science Methods for Digital Learning Platforms certificate program, contact us at ProfessionalLearning@gse.upenn.edu
Over the last 15 years, the educational data mining and learning analytics communities have developed a range of algorithms tailored to the data and R&D goals of digital learning platforms. The Data Science Methods for Digital Learning Platforms certificate provides learners with a breadth and depth of skills that expand beyond existing courses of its length in data science or education more generally.
Working with the data from digital learning platforms requires a range of data science methods. Do you have the right tools to pursue your education research questions?
Through the Data Science Methods for Digital Learning Platforms certificate program, you will learn to use both algorithms designed specifically for digital learning platforms and how to effectively apply algorithms developed for more general purposes to digital learning platform data.
Spring 2024 Starts: February 5, 2024
Spring 2024 Application Deadline
Fall 2024 Start: September/October
Fall 2024 Application Deadline
Certificate offered: Penn GSE certificate in Data Science Methods for Digital Learning Platforms
Ideal candidates
Prerequisites
In this 16-week program, participants will learn to conduct analysis on real-world data, working directly with authentic student interaction data that has not been cleaned. You’ll compare statistical and psychometric approaches to machine learning and data mining methods. You will also learn how to move beyond the well-structured use cases often utilized in introductory data science and statistics courses, which are often not representative of the data that comes from digital learning platforms.
The program is online and asynchronous, with one optional synchronous and virtual “ask me anything” session with the instructors. Each module includes discussion-based interactions with peers and instructors and a project-based assignment for which fellows will be able to apply the skills they learn using authentic tools and datasets. The examples and assignments corresponding with each module will align with real challenges and scenarios common to digital learning platforms. Emphasis is given to identifying the development of relevant research questions and understanding the limitations and affordances that different types of digital learning platform data may provide in addressing these questions.
This program is designed for individuals with a clear intent to pursue education research, and who have some degree of prior quantitative analysis background and either an intermediate-level understanding of statistics or psychometrics or a background in computer science. Twenty participants will be selected for the first cohort based on a competitive application process.
There will be five cohorts of fellows over three years, the next application will be for the second cohort starting in Fall 2024. There is no cost to participate in this program.
Dates | Topics |
Feb 5 – 9, 2024 | Introduction, Challenges, and Framework |
Feb 12 – 16, 2024 | Prediction Modeling and Metrics |
Feb 19 – 23, 2024 | Feature Extraction and Feature Engineering |
Feb 26 – March 1, 2024 | Neural Networks and Deep Learning |
March 4 – 8, 2024 | Data Visualization |
March 11 – 15, 2024 | Ethics, Equity, and Algorithmic Bias |
March 18 – 22, 2024 | Data Management and Database Access |
March 25 – 29, 2024 | Knowledge Graphs |
April 1 – 5, 2024 | Knowledge Tracing |
April 8 – 12, 2024 | Data and Measurement Validity |
April 15 – 19, 2024 | Cluster Analysis |
April 22 – 26, 2024 | Network Analysis |
April 29 – May 3, 2024 | Sequential Pattern Mining and Temporal Analysis |
May 6 – 10, 2024 | Causal Reasoning |
May 13 – 17, 2024 | Natural Language Processing |
May 20 – 24, 2024 | Transformer and Foundation Models |
This program and certificate are made possible through a partnership between PennGSE, the University of Florida, and Digital Promise.
This program is led by Principal Investigator (PI) Ryan Baker, Penn GSE; along with 5 co-PIs representing the collaborating institutions: Anthony Botelho, University of Florida; Bodong Chen, Penn GSE; Elizabeth Cloude, Tampere University; Stefani Pautz Stephenson, Digital Promise; and Jeremy Roschelle, Digital Promise.
This project is supported by the Institute of Education Sciences, U.S. Department of Education, through Grant R305B230007 to the University of Pennsylvania. The opinions expressed are those of the authors and do not represent views of the Institute or the U.S. Department of Education.
Ryan Baker
University of Pennsylvania
Anthony Botelho
University of Florida
Bodong Chen
University of Pennsylvania
Elizabeth Cloude
Tampere University
Jeremy Roschelle
Digital Promise
Stefani Pautz Stephenson
Digital Promise