Paper: Benefiting from Mistakes: The Impact of Guided Errors on Learning, Performance, and Self-Efficacy

S. J. Lorenzet, E. Salas, S. I. Tannenbaum. “Benefiting from Mistakes: The Impact of Guided Errors on Learning, Performance, and Self-Efficacy.” Human Resource Development Quarterly, vol. 16, no. 3, 2005. doi:10.1002/hrdq.1141

  • Ninety undergraduate students with no previous experience received either training that guided them to commit common errors or alternatively training that sought to prevent errors from occurring
  • A typology for manipulating errors and a new way of using errors in training are presented
  • Findings revealed superior performance (accuracy and speed) and self-efficacy associated with using guided errors during training
  • Classic reinforcement theories suggest that training should be structured so errors are minimized. The logic behind this “error-avoidance” approach was that errors take away from “on-task” time and therefore reduce the amount of learned information.
  • The logic behind the use of errors is that they serve an informational function and thus provide feedback, giving trainees an opportunity to see the consequences of their mistakes, learn corrective strategies, and take corrective steps
  • There are four approaches to error occurrence: avoiding, allowing, inducing, and guiding errors, all summarized below:
  • The second component of using errors in training is either allowing trainees to work through the errors unaided by the trainers or any other support mechanism or supporting the trainees through trainer intervention or computer-based assistance
  • The dominant model of training evaluation in the research literature for nearly the past half century has been Kirkpatrick’s four levels of evaluation: trainee reactions, learning, behavior, and results
  • Several new evaluation frameworks have recently emerged emphasizing evaluation issues such as learning, performance, change, perception, and return on investment
  • The emergence of these new models may at least in part be a result of research highlighting the limitations of Kirkpatrick’s evaluation model, including the implication that results are the best measure and the practice of many cases studies relying on just one level of training evaluation
  • On the basis of taxonomies developed in other disciplines, Kraiger and colleagues (1993) proposed three categories of learning outcomes: cognitive, skill-based, and affective. In the present study,
    these categories are represented as cognitive learning, performance, and self-efficacy
  • Guided errors should be more effective than error-free training in giving trainees coping strategies when they do encounter mistakes, may aid in the construction of a mental model and remind learners to avoid prior mistakes, and aid in discovering relevant information or generalizing skills acquired during training to new problems
  • Some of the rationale for advocating error-free training has come from concern over negative affective reactions associated with mistakes, which may lead to reduced self-efficacy and lower performance outcomes
  • Trainees who experience guided errors should be less likely to attribute mistakes to internal causes and thus may not experience a reduction in self efficacy
  • Successful performance during error-free training may actually have some negative effects, including overestimation of skill level and a subsequent drop in self-efficacy after a poorer performance than expected
  • Error-free trainees were given “click-by-click” instructions and were led through training without errors, and guided-error trainees were given the same click-by-click instructions but were led into mistakes and then shown how to fix them
  • Fixing mistakes simply included identifying mistakes and performing the operations correctly, not teaching additional software capabilities
  • A potential concern was that guided-error training would take longer than error-free training, which could account for post-training performance differences, but this was controlled for by ensuring that the type took approximately the same amount of time
  • Results did not reveal statistically significant differences between guided error trainees and error-free trainees with regard to self-efficacy immediately following training
  • Analysis revealed higher post-performance self-efficacy for guided-error trainees compared to error-free trainees
  • Trainees who received guided errors performed more accurately and performed faster than those who received error-free training
  • The self-efficacy of guided-error trainees remains virtually unchanged after performance (M=75.34 before, 77.09 after), while error-free trainees show a substantial drop in self-efficacy (M=77.09 before, 63.60 after)

______________

This research, demonstrating that guiding trainees through errors and their solutions leads to better outcomes than guiding through the correct procedure only, seems both plausible from a personal perspective and immediately useful in an instructional context. While the study does not address the origin of the difference in outcomes from the two training methods, it appears that there are two likely candidate causes for the lower performance in the error-free approach: lack of knowledge on how to fix errors and fear of making errors in the first place. More to come in an upcoming blog post…

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s