Science of Learning Research Centre

Student Feedback

Giving students feedback – getting the balance right

Feedback is one of the most powerful influences in students’ achievement, but it also one of the most variable in its influence. Researchers have been investigating the effect of feedback and reinforcement on learning outcomes.

Professor John Hattie (The University of Melbourne), Professor Annemaree Carroll, Professor Robyn Gilles and Cameron Brooks (all from The University of Queensland) have for a number of years been investigating how students use feedback in the classroom.

Analysis of classroom dialogue revealed that over 75% of feedback within English lessons was directed at the task level which can be equated to promoting surface thinking whilst correspondingly less than 25% of feedback could be attributed to engendering deeper thinking and learning processes.

When students were questioned on their perceptions of feedback, they rated feed forward (feedback that is improvement focused) as the most helpful feedback type. Frequency data however shows this is the least used feedback type in the classroom.

Likewise getting the right amount of feedback is also crucial to learning outcomes.

Speaking to the ABC Radio National’s All in the Mind program, SLRC Director Professor Pankaj Sah said providing feedback on what students are doing, or have achieved, is one of the most critical components in the best learning outcomes.

“For example supposing you had a particular task which you provide feedback on, think about the condition where you do the task five times, and each time you get it right you get feedback on it,” Professor Sah said.

Business quality service customer feedback, rating and survey keys with smiling face symbol and icon on computer keyboard.

“That’s what you might say is 100% feedback; that every time you do a task you get feedback on it, and that results in a particular outcome. Supposing though you only provide feedback 50% of the time – what will be the outcome of that?”

SLRC researchers have used both laboratory animal models, as well as classroom situations, to assess this.

“Well, it turns out in animal models, animals learned just as well with 50% feedback as with 100% feedback. What’s interesting though is that testing them in the long term, those animals getting 50% feedback tend to learn better.

“Now, in the classroom setting, if you’re trying to teach someone a particular thing, suppose you decide not to, say, ‘okay, you’ve done fine, that’s correct’, every single time, but you do it on a partial basis, it’s likely that these kids will learn just as well, but it’s possible they will remember things a bit better or longer, or forget it less.”

Research results showed significant improvements in academic achievement in writing between control and intervention groups. Students perceived the feedback intervention to offer more improvement focused feedback and teachers observed greater self-regulation amongst their students.

In response to these findings, the SLRC has developed and implemented (through teacher coaching) a Feedback for Learning in the classroom matrix. A pilot study of the Feedback for Learning matrix is planned during 2017, with a view of reaching 15 schools by the end of the calendar year with the ultimate aim that the information and feedback tools can be rolled out on a larger scale using a coach-the-coach model.

The ABC Radio National article is available from: