Skip to content

Error Correction Procedure and Transfer Trial


As much as the Type A perfectionist in me dislikes making errors, they are a natural part of life and an essential part of learning. While very cautiously trying to learn a new language, I have been making all types of errors. The part of this scary new learning project that has made it more accessible is that I have a wonderful mentor where the errors are considered additional teaching opportunities. Instead of just being told I am wrong (i.e., stand there in your wrongness and be wrong), I get an opportunity to practice the correct response with a prompt, as well more chances to see if I have learned from the extra teaching.


What is an Error Correction Procedure?

As the name states, an error correction procedure is when the instructor corrects an incorrect response. In the case of a receptive label program, the error correction would be the instructor pointing to the correct picture.


Instructor SD: “point to car”
Learner response: points to bike
Error Correction: instructor points to car and says “this is car”

What is a Transfer Trial?

A transfer trial would be run after the error correction takes place. DO NOT CHANGE THE PRESENTATION OF THE STIMULI YET. The transfer trial gives the learner an opportunity to practice the response correctly, with support from the instructor. Depending on the learner and the teaching context, support from the instructor can be a gesture prompt, a positional prompt, or another type of prompt.


When running the transfer trial, predict that the learner may err
Instructor represents the SD: “point to car” *with gestural prompt*
Learner responds correctly: points to car

What is an Expanded Trial?

Once the original SD is presented, and the learner errs, the instructor will present the error correction and transfer trial. An expanded trial may be conducted by a more experienced instructor.

An expanded trial is when another instruction is given. Similar to high-probability instruction, choose an instruction that is mastered, giving the learner a high-probability of responding correctly.


Instructor SD: “what’s your name?”
Learner responds correctly

Represent the SD

After you have presented the initial SD (student scores as incorrect), you conduct the error correction procedure, transfer trial, and expanded trial, then it is time to represent the SD. This time with a less intrusive prompt than in the transfer trial, or with no prompt (depending on the learner and their individual learning context).


Instructor SD: “point to car” *no prompt, or less intrusive prompt*
Learner response: points to car

How to Collect Data?

The entire sequence of initial presentation of the SD, error correction, transfer trial, expanded trial, and representation of the SD is all considered one trial – the entire sequence is scored as incorrect as the learner erred on the initial presentation.

If the learner responded correctly during the error correction, transfer trial, expanded trial, and representation of the SD, the initial response is still scored as incorrect. These steps are all teaching and practice trials.

If the learner responds incorrectly during the error correction, transfer trial, and/or the representation of the SD, it may be appropriate to revise the targets, teaching steps, prompt levels, or other. Consult the revision criteria for the program.

What Comes Next?

After you conduct the entire above sequence, you should rearrange the stimuli presented, and present the next SD in the program.


Find more information on the echoic-to-mand transfer, and watch our YouTube video:

3 thoughts on “Error Correction Procedure and Transfer Trial”

  1. Hello! Thank you for explaining the error correction procedure and data collection. Question- Why are the error correction, transfer trial, expanded trial, and representation of the SD not considered separate trials? If you’re re-stating the SD and requiring a response, isn’t that another opportunity to take data too?

    1. It’s because the second trial is really part of the teaching trial – it shouldn’t be counted as an independent response. You had just highly supported the correct answer so it wouldn’t be an accurate representation of their skill but it is good teaching.

Leave a Reply

Your email address will not be published. Required fields are marked *

Captcha loading...