In this paper we study if feedback strategies during formative assessment with polling technology have an impact on learning gains. We conduct a randomized experiment in physics class in upper secondary education with the web-based polling system Socrative comparing three conditions. In the cooperative condition students receive a combination of peer discussion and teacher feedback, while in the individual condition they receive only teacher feedback. In the control condition students do not receive any feedback, from either teacher or peers. To measure what individuals learn from teacher feedback, whether or not combined with peer discussions, students in all three conditions individually answer paired isomorphic multiple-choice questions. Per question, we study the probability to answer the second isomorphic question correct and compare this between conditions The analyses show that teacher feedback, regardless of the peer discussions, positively affects learning gains in question pairs of the treatment conditions in comparison with the control condition. However, the cooperative condition shows that the largest learning gains occur when peer discussion is followed by teacher feedback. The findings provide insights into which feedback strategies affect learning gains in question pairs and are important for teachers when implementing feedback strategies during formative assessments.

在本文中，我们研究了使用投票技术的形成性评估过程中的反馈策略是否会对学习收益产生影响。我们使用基于网络的投票系统Socrative在高中物理课上进行了一项随机实验比较三个条件。在合作条件下，学生会收到同伴讨论和教师反馈的组合，而在个人条件下，他们只收到教师反馈。在控制条件下，学生不会收到任何来自老师或同龄人的反馈。为了衡量个人从教师反馈中学到的东西，无论是否与同伴讨论相结合，所有三种情况下的学生都会单独回答成对的同构多项选择题。对于每个问题，我们研究正确回答第二个同构问题的概率并在条件之间进行比较。分析表明，与控制条件相比，教师反馈，无论同伴讨论如何，都会对处理条件的问题对的学习收益产生积极影响。然而，合作条件表明，当同伴讨论之后是教师反馈时，学习收益最大。研究结果提供了关于哪些反馈策略影响问题对学习收益的见解，并且对于教师在形成性评估期间实施反馈策略时很重要。