Data files for Automated tailored feedback does not help university students overcome mathematics misconceptions

Contains the main data file, Rmd script, and html output of the Rmd script. Also contains the figures from the manuscript.

Article abstract:

Background: Programming automated tailored feedback to help students overcome mathematical errors is time-consuming, and given the mixed results of the feedback literature might not be worth the effort.

Aims: We conducted an experiment to test whether automated tailored feedback helped students overcome a well-known misconception involving calculating the area bounded by a curve and the x-axis.

Sample: Participants were 461 university students drawn from three continents.

Methods: Participants were randomly assigned to receive no feedback, or accuracy feedback, or accuracy with elaborated feedback. The experiment involved a pre-test for participants to revise and practise integral calculus techniques, priming items to cue the misconception, intervention items that provided feedback or none according to condition, and post-test items similar to the intervention items.

Results: Counter to our expectations, the vast majority of participants scored 0 in the post-test. Participants in the accuracy with elaborated feedback condition scored more highly than those in the accuracy feedback condition, who scored more highly than those in the no feedback condition, but these differences were not significant. In addition, participants in the accuracy with elaborated feedback condition gave fewer misconception responses than those in the accuracy feedback condition, who gave fewer misconception responses than those in the no feedback condition, but again these differences were not significant.

Conclusions: Programming automated tailored feedback that responds to common mathematics misconceptions in university mathematics may not be worth the effort.