This experiment tests the effectiveness of “sandwich” feedback. 91 university students solved 12 mathematical problems from the secondary-school curriculum. After the time limit, we assigned them randomly to one of three possible treatments. One group received corrective computer-administrated feedback, describing the mistakes with their methods and solutions. The second group received sandwich feedback, consisting of the same corrective part presented between two general positive statements unrelated to the participants’ actual performance. The third group did not receive any feedback. Afterwards, the participants had 10 min to prepare for the second set of similar problems. Participants who received sandwich feedback utilized more time on preparation and solved more problems from the second set than the participants from the other two groups. This study provides only partial evidence for the effectiveness of sandwich feedbacks as it tested the effect under one specific condition using computer-mediated written feedback on math test. Further replications are needed to test the effect under various conditions, to test various forms of sandwich feedback, to explain the mechanism of sandwich feedback and to show whether the effect of sandwich feedback is caused by the specific sequence of feedback components or by mere presence of positive statements.

该实验测试了“三明治”反馈的有效性。91名大学生从中学课程中解决了12个数学问题。在时间限制之后，我们将它们随机分配给三种可能的治疗方法之一。一组人收到了计算机管理的纠正性反馈，描述了他们的方法和解决方案中的错误。第二组收到三明治反馈，该反馈包含两个与参与者的实际表现无关的一般正面陈述之间的相同纠正部分。第三组没有收到任何反馈。之后，参与者有10分钟的时间为第二组类似的问题做准备。与其他两组参与者相比，收到三明治反馈的参与者在准备工作上花费了更多时间，并且在第二组解决了更多问题。这项研究仅提供了部分三明治反馈的有效性的部分证据，因为它使用一种计算机介导的数学测试书面反馈在一种特定条件下测试了效果。需要进一步复制以测试在各种条件下的效果，测试各种形式的三明治反馈，解释三明治反馈的机理以及显示三明治反馈的效果是由特定的反馈分量序列还是仅由反馈成分的存在引起的肯定的陈述。