Tuesday, February 16th
8:10am: Class starts in five minutes and Emma is double checking her problem set before handing it in. Worried she might be called on by the teacher, Emma spent over an hour last night working through the equation. She feels confident, her phone chirps to remind her class is starting…
For decades we’ve heard the mantra, “immediate feedback is best.” But is this always the case? We often simplify learning in the classroom as a response to positive and negative feedback that motivates our students to perform. This type of learning can be referred to as “operant conditioning,” and requires that the outcome be delivered soon after the action. But we now know that effective feedback involves more than just punishments and rewards, and can provide information to guide learning.
Beginning in the 1960s psychologists began to find that delaying feedback could improve learning. An early lab experiment involved 3rd graders performing a task we can all remember doing: memorizing state capitols. The students were shown a state, and two possible capitols. One group was given feedback immediately after answering; the other group after a 10 second delay. When all students were tested a week later, those who received delayed feedback had the highest scores. Most early attempts to replicate this delay-of-feedback benefit in schools had flawed designs and largely failed. A review of these findings in the 1980’s led many to conclude that in the classroom immediate feedback is best, and the idea stuck.
8:15am: Emma is the first student to submit her problem set and is eager to find out how she did. How can we have the correct answer on hand, yet still let this teachable moment slip away?
Recent studies are breathing new life into delayed feedback. One such study looks at an undergraduate engineering course at University of Texas, El Paso. Students in the course submitted a weekly homework assignment and either received feedback immediately, or a week later. Several weeks later all students completed a similar problem on the exam. The students who received delayed feedback scored higher on the exam than those who received immediate feedback.
I found this to be a bit counterintuitive, so I reached out to Andrew Butler, a faculty member at UT-Austin and co-author of the study, to see what he thought was going on here. He describes it as “essentially a spacing effect—distributing exposure to or practice with material improves long-term retention.” He also provides the caveat that “delaying feedback will only be beneficial if students fully process the feedback and sometimes they are less motivated to do so after a delay.”
The study authors required the students to click on the feedback to get credit for the homework—a helpful approach regardless of when feedback is delivered.
9:30am: “Thanks to all of you who turned in your homework on-time,” I tell the class. “You’ll receive the graded feedback next Tuesday, and we’ll review the concepts at the beginning of class.” Emma approaches me with a concerned look on her face, “Professor Roberts, would it be possible to find out the answer now? I learn better when I can quickly find out what I did wrong.”
Neuroimaging experiments are providing evidence that learning from delayed feedback may involve a different memory system than learning that occurs from immediate feedback. A study at Columbia University found that subjects receiving immediate feedback showed activation in a brain area that supports implicit learning (memorizing 7 x 7 = 49) called the striatum, while delayed feedback increased activation of the medial temporal lobe, which supports explicit learning (solving for X, where 7x = 31).
Delaying feedback can improve performance across multiple assessment formats and time frames, providing a strategic opportunity for companies developing intelligent and adaptive platforms. Startups often share how many million students use their product, and this is where the true opportunity lies. Small changes in feedback timing in a quiz app may produce troves of valuable data. As highlighted by Digital Promise during a recent Summit I attended, collaborations of this nature between developers, researchers and educators will be essential to continued innovation in edtech.
Tuesday, March 15th
9:30am: It’s been a month since Emma handed in her first problem set of the unit and she’s logging in to see how she did on yesterday’s midterm. She scored 94 percent, her first A on an exam! By simply delaying the return of Emma’s homework by a week, I’ve employed the learning sciences to increase Emma’s chances of success.
