arxiv.org web signal

Doğucu, Medina, Lewis map data choices for stats classrooms

By Published Updated

TL;DR

  • A 14-page arxiv guide by Mine Doğucu, Catalina Medina and Crystal Lewis, posted October 30, tackles how to pick data for statistics and data-science classrooms.
  • The paper gives definitions and distinctions for data variously called real, raw, made-up, artificial, fictitious, synthetic, simulated, toy, and messy.
  • The authors say their recommendations on classroom data use are both empirically and theoretically supported, and compile resources for instructors.

Three statistics educators have published a 14-page guide to a question stats and data-science instructors keep asking each other: when you bring data into a classroom, what kind of data should it be. The paper, posted to arxiv on October 30 by Mine Doğucu, Catalina Medina and Crystal Lewis, works through "definitions of different types, sources, formats, and shapes of data" and tries to put labels on a muddle that educators already use.

The authors frame it plainly. "For many educators, the question is not whether to use data or not," they write. "They often want to know whether they should give clean or messy data, and whether they should spend the time to find a real dataset or simulate it." A list of the words classroom data gets described by runs through the abstract: "real, raw, made-up, artificial, fictitious, synthetic, simulated, toy, and messy".

The guide pitches its recommendations as "empirically and theoretically supported" and closes by sharing resources for instructors. The abstract does not list which datasets the authors reach for, or which classifications they prefer in which teaching context.

Shared on Bluesky by 2 AI experts

Originally reported by arxiv.org

Read the original article →

Original headline: Guide to Teaching with Data in Statistics and Data Science Classrooms