Abstract: The importance of student evaluation of teaching (SET) in higher education has increased substantially in the past decades. Despite this increase, there is no consensus on the most efficient way of executing these surveys or about the questions. This study analyses the impact of an SET system reform on the distribution of the responses and the response rate. The reform impacted the questions, the scale used for evaluation, the anchor labels of the scale elements, and the incentives for students to fill out the survey. Our results show that the number of extreme responses increased after the reform, which can be due to the elimination of anchor labels of the middle scale points. Insufficient effort bias also increased; however, the new motivation system doubled the response rate, which helped to collect more representative sets of evaluations. Taking into consideration the relatively small increase in insufficient responses, we believe this incentive can be a valid choice for SET surveys.