This paper is from Session 7F: Challenges for statistics in the past and in the future
which comes under Topic 7: Statistical literacy in the wider society
Paper 7F3 (Thursday 12th, 11:00-12:30)
Teaching and learning about tree-based methods for exploratory data analysis
- Joachim Engel (Ludwigsburg University of Education, Germany)
- Laura Martignon (Ludwigsburg University of Education, Germany)
- Tim Erickson (Epistemological Engineering, United States)
Quantitative information about important societal topics are increasingly accessible to the general public and to individual citizen. Making sense of these data requires the ability to explore, understand, and reason about complex multivariate data. For such data, we need flexible and robust analytical methods that can deal with nonlinear relationships, mixed type variables, high-order interactions and missing values. Classification and regression trees are a method that is well suited for the analysis of such complex data. Besides reflecting on tree-based methods for teaching understanding about complex social data the purpose of this paper is to introduce the digital tool EasyTREE which is designed to help understanding about construction and interpretation of decision trees and supports critically evaluating the strength and pitfalls of the tree method.