Error Rate for Optimal Learning

Goals

To identify the estimation of the ideal error rate for optimal learning among animals or humans with regard to language learning or any other kind of learning. This will include research papers that support that estimation as well as a summary of the relevant portions of the research paper.

Early Findings

Learning from Errors

  • In January 2017, Janet Metcalfe of Columbia University published a paper in the Annual Review of Psychology entitled 'Learning from Errors.'
  • According to this paper, "Although error avoidance during learning appears to be the rule in American classrooms, laboratory studies suggest that it may be a counterproductive strategy, at least for neurologically typical students."
  • Their experimental studies also "indicate that errorful learning followed by corrective feedback is beneficial to learning."
  • According to Metcalfe, "Interestingly, the beneficial effects are particularly salient when individuals strongly believe that their error is correct: Errors committed with high confidence are corrected more readily than low-confidence errors. Corrective feedback, including analysis of the reasoning leading up to the mistake, is crucial. Aside from the direct benefit to learners, teachers gain valuable information from errors, and error tolerance encourages students’ active, exploratory, generative engagement. If the goal is optimal performance in high-stakes situations, it may be worthwhile to allow and even encourage students to commit and correct errors while they are in low-stakes learning situations rather than to assiduously avoid errors at all costs."

Research on Language Learning

  • In December 2013, a study published by the Linguistic Society of America entitled 'Error and Expectation in Language Learning' reveals that "children’s overregularization errors both arise and resolve themselves as a consequence of the distribution of error in the linguistic environment, and that far from presenting a logical puzzle for learning, they are inevitable consequences of it."

Summary of Findings

  • Our initial research reveals that a study on the error rate for optimal learning of humans and animals has not been conducted yet. The only study available regarding error rate is that of the 15% error rate for artificial neural networks which may potentially be generalized to animal learning, as suggested by researchers.
  • We have included some related researches. Though these research papers do not investigate error rates, they are the closest studies related to learning from errors and error in language learning.

Proposed next steps:

You need to be the project owner to select a next step.