Effectiveness of reformulating incorrect sentences as feedback?

Question

My Chinese language exchange partner naturally makes errors when speaking German. One form of feedback that I often use is reformulating the sentence he just spoke, so that he gets to hear a correct version of it. He then just nods, appears to try to memorise what I said and moves on. He does not try to repeat what I said, nor does he write it down. Sometimes, he makes exactly the same mistake one week later (e.g. "Ich rufen Sie an ..." instead of "Ich rufe Sie an ...").

This type of feedback is also often used in classrooms (which is probably where I picked it up). However, is there any research on whether this type of feedback is effective compared to other types of feedback? (Regardless whether it's used in the classroom or in tandem language learning.)

Answer 1

Below is a relevant fragment of my Master's thesis. The focus is on written feedback, but oral feedback is mentioned, too. Based on that, I would say that giving the student a reformulated sentence is unlikely to be effective. A better strategy would be to give them cues as to what is wrong and let them do the self-repair.

[The effectiveness of written corrective feedback] is subject to controversy. A recent example is a meta-analysis performed by Truscott (2007), who obtained results that suggest that traditional forms of written corrective feedback (WCF) do not have any positive effect on students’ ability to write accurately. This does not mean, however, that feedback cannot have benefits.

[...]

Hartshorn et al. (2010) tested the empirical effects of a novel form of WCF, that they called dynamic written corrective feedback and found it to be effective. This form of feedback has four distinguishing properties:

• Meaningful. The learner must know what kind of error they made, and how it can be corrected. Meaningful feedback may not give the complete correct answer, but gives enough information so that the learner can correct the error herself.
• Timely and constant. The feedback should be provided as soon as possible, and its form and scope should not change, so as not to confuse the learner.
• Manageable. The feedback cannot be overwhelming. It needs to be provided in manageable chunks, and information overload should be avoided.

Some of these features are related to the scaffolding functions [...]: manageability has to do with reducing degrees of freedom, and meaningfulness – with proper marking of critical features. The above findings are consistent with the meta-analysis by Lyster & Saito (2010), who found that in oral contexts the most effective is prompt feedback, which does not provide an alternative formulation of the incorrect utterance, but provides cues that let the learner self-repair.

Bibliography:

• Hartshorn, K. James, Norman W. Evans, Paul F. Merrill, Richard R. Sudweeks, Diane Strong-Krause & Neil J. Anderson. 2010. Effects of dynamic corrective feedback on ESL writing accuracy. TESOL Quarterly 44(1). 84–109.
• Lyster, Roy & Kazuya Saito. 2010. Oral feedback in classroom SLA. Studies in Second Language Acquisition 32(02). 265–302.
• Truscott, John. 2007. The effect of error correction on learners’ ability to write accurately. Journal of Second Language Writing 16(4). 255–272.

Answer 2

Firstly, feedback itself is a tool that should only be used to "fine tune" learners' understanding. It is not a replacement for instruction. Unfortunately, insufficient effective instruction* occurs in typical EFL/ESL classrooms & I suspect that much of the research into corrective feedback in second & foreign language learning hasn't taken this principle into consideration so the results have been very mixed, & so I'd say that the controversies that surround the debate on corrective feedback are from poor quality evidence more than anything else.

Rod Ellis offers a typography of feedback here: https://academic.oup.com/eltj/article/63/2/97/440749

The type of feedback you've described is known as a "recast." They may be effective but only if your conversation partner (1) understands your intentions (to give corrective feedback), & (2) processes the feedback, which means switching attention from the topic of conversation to attend to the feedback, & (3) he is "ready" to learn/correct that particular linguistic feature in the feedback, i.e. It's at the right level for them to understand.

On the downside, some researchers have argued that corrective feedback can cause learners to simplify their speech in order to avoid making errors & that this may restrict their language development. The argument is that "interlanguage" development (AKA "developmental sequences") includes errors which are unavoidable & a natural part of the process & that they are insensitive to instruction. Giving corrective feedback may discourage learners from persisting in these areas of development (making errors), with negative consequences.

* "effective instruction" is also a controversial topic in language learning research but it is clear that the usual grammar translation, drilling, & pattern practice exercises have little effect on language learning in terms of getting better at communication (developing communicative competence). One promising avenue of research on language instruction is in "Structured Input" (Bill vanPatten), explanation: https://www.youtube.com/watch?v=LIAhI5oEgHE&list=PLCvyUoftI84sWQ6QCFrPweda0Cwf49SJK&index=4