Feedback research revisited

Marcus Specht

Over a long time one article has had a lot of impact on my research and discussions with practitioners when looking at how to design online systems and especially feedback in these systems. Edna Holland Mory’s article on “Feedback research revisited" gives really nice overview on the effects of feedback, what the design options are and what might be most efficient for your setting.

Basically feedback can have different functions based on the learning environment and the learning paradigm implemented. Literally the basic definition according to the Webster dictionary are: “a process in which the factors that produce a result are them- selves modified, corrected, strengthened, etc. by that result” and “a response, as one that sets such a process in motion” (p.520). 

According to Mary information presented in feedback might include "answer correctness, precision, timelines, learning guidance, motivational messages, lesson sequence advisement, critical comparisons, and learning focus” (p.745).

From a historical perspective feedback has fulfilled different functions in the different learning paradigms over the years, i.e.:

  • Law of effect, where feedback was a connector between a presented stimulus and the given responses. Feedback in this sense could be seen as a reinforcement to strengthen relation between the stimuli and “correct” response, but also corrective feedback could be used to loosen connections between “incorrect” responses. In this period already, the function of punishment and reinforcement have been discussed.
  • Skinner (1958) moved this one step further in the design of classroom materials following the principle of programmed instruction. In this context feedback was used as reinforcement as also as incentive in designing learning materials in which you progress through the content based on your correct and/or incorrect responses.
  • Reinforcement learning as a further development of the original programmed instruction approach additionally tried to define redundant and small steps in learning materials so that the probability for errors was minimized. A critical component in this approach was that students must respond before seeing the feedback or parts of a correct solution.
  • Feedback as reinforcement vs. feedback as information shifted the understanding of the importance of errors and the idea of feedback to correct and/or to reinforce learner responses.

In a second part of the article the major theoretical models for feedback are described including connectivist model, certitude model, and a five stage model of mindfulness. The interested reader might dive into the article him/herself.

As a main information in this blog I would like to refer to the dimensions and research variables around feedback:

  • Complexity: In different variations feedback can be given either with the idea to just say correct/not correct up to details on why and where an error was made. Variations: No feedback, simple verification feedback, correct response feedback, elaborated feedback, try-again feedback. There is inconclusive evidence on how feedback should be elaborated, nevertheless the main variations research are giving question and correct and/or incorrect answers as also giving students the possibility of exploring further on correct answers in multiple choice tests.
  • Timing: In research on timing of feedback the differentiation between immediate and delayed feedback has been investigated and in most cases immediate feedback has been shown to be more effective, nevertheless there are newer studies in which also delayed feedback can have a positive effects on learning outcome.
  • Error specific feedback: This approach is basically defined on a classification of types of errors. Meyer (1986) differentiated four error types and postulated to design feedback according to the error occurring, i.e. lack-of-information errors, motor or production errors, confusions, and run-application errors.
  • Feedback specific for learning outcomes: While for learning descriptive knowledge it often might be more effective to work with simple correct/incorrect feedback for conceptual learning immediate extended feedback for so well correct as also incorrect responses has been found to be more efficient.

The article describes much more variations and models of feedback as also its important impact and relation to motivational questions. For bringing the above described variations to practice I think it is a good starting point to reflect about your own way of giving feedback.

Reflection questions:

  • How do you give feedback?
  • How do you react to students making errors?
  • What do you think about the role of errors, is your main purpose to give guidance to avoid them or to correct them?


Feedback Research revisited: 

Mory, E. H. (2004). Feedback research revisited. In D. H. Jonassen (Ed.), Handbook of research on educational communications and technology (Vol. 45, pp. 745–784). Lawrence Erlbaum Associates. 

More information
Mory, E. H. (2004). Feedback research revisited. In D. H. Jonassen (Ed.), Handb…