CDS‑DS‑6XX: Algorithms and Evidence

🎓 • Boston University

CDS‑DS‑6XX: Algorithms and Evidence (Proposed)

Course Overview

Graduate-level seminar examining how algorithmic decision-making intersects with epistemology, evidence, and justice. Topics include causal inference, interpretability, and the use of data in high-stakes social contexts.

Proposed Course Design

Notes: Proposed by me for the CJ Master’s curriculum. Not yet offered. Focuses on helping students critically evaluate the structure and implications of algorithmic systems.

Learning Objectives

Students will develop critical understanding of:

  1. Epistemological Foundations: How algorithms shape what counts as evidence
  2. Causal Inference: Methods for establishing causation from observational data
  3. Algorithmic Interpretability: Understanding and explaining automated decisions
  4. Justice and Algorithms: Examining fairness, bias, and accountability in AI systems
  5. High-stakes Contexts: Evaluating algorithmic use in criminal justice, healthcare, and education
  6. Evidence Standards: Different requirements for evidence across domains

Proposed Course Structure

  • Seminar Format: Discussion-based exploration of cutting-edge research
  • Case Study Analysis: Examining real-world algorithmic systems and their impacts
  • Interdisciplinary Reading: Drawing from computer science, philosophy, law, and social science
  • Critical Projects: Students conduct original analysis of algorithmic systems
  • Guest Experts: Practitioners from law, policy, and technology

Key Topics

  • Philosophy of Science: What constitutes valid evidence and knowledge
  • Causal Methods: Experimental design, natural experiments, and causal identification
  • Algorithmic Transparency: Explainable AI and interpretability techniques
  • Bias and Fairness: Technical and philosophical approaches to algorithmic fairness
  • Regulatory Frameworks: Legal and policy responses to algorithmic decision-making

Target Audience

Designed for graduate students in computational journalism, data science, and related fields who need to critically evaluate and report on algorithmic systems in society.

Course Innovation

Unique integration of technical methods with philosophical and legal frameworks for understanding evidence and decision-making in the age of algorithms.