Affordable Access

Access to the full text

How to improve statistical thinking: Choose the task representation wisely and learn by doing

Authors
  • Sedlmeier, Peter1
  • 1 Universität-GH Paderborn, FB 2 – Psychologie, Paderborn, 33095, Germany , Paderborn
Type
Published Article
Journal
Instructional Science
Publisher
Kluwer Academic Publishers
Publication Date
May 01, 2000
Volume
28
Issue
3
Pages
227–262
Identifiers
DOI: 10.1023/A:1003802232617
Source
Springer Nature
Keywords
License
Yellow

Abstract

Attempts to teach statistical thinking usingcorrective feedback or a “rule-training” approach havebeen only moderately successful. A new trainingapproach is proposed which relies on the assumptionthat the human mind is naturally equipped to solvemany statistical tasks in which the relevantinformation is presented in terms of absolutefrequencies instead of probabilities. In aninvestigation of this approach, people were trained tosolve tasks involving conjunctive and conditionalprobabilities using a frequency grid to representprobability information. It is suggested that learningby doing, whose importance was largely neglected inprior training studies, has played a major role in thecurrent training. Study 1 showed that training thatcombines external pictorial representations andlearning by doing has a large and lasting effect onhow well people can solve conjunctive probabilitytasks. A ceiling effect prevented comparison of thefrequency grid and a conventional pictorialrepresentation (Venn diagrams) with respect toeffectiveness. However, the grid representation wasfound to be more effective in Study 2, which dealtwith the more difficult topic of conditionalprobabilities. These results suggest methods tooptimize the teaching of statistical thinking and thepresentation of statistical information in themedia.

Report this publication

Statistics

Seen <100 times