I have been skeptical about this way of evaluation for a long time. To begin with, these evaluations are often of little use to the teacher/trainer. A relatively low average score could be an indication that the the teacher did not do a very good job but could also be caused by other factors. For example, students may not yet see the usefulness of the topic. Perhaps their enthusiasm is low to moderate at the end of the course but increases later as they have gained more knowledge and experience.
Evaluation can put students in a consumer roleAsking students to evaluate teachers, I fear, also has the disadvantage that it puts them in the role of a consumer. The student is the customer and the teacher the supplier. The implicit deal is: you (teacher) have to satisfy me (student) and I will tell you afterwards if you have succeeded. This consumer role, I am afraid, will undermine student learning. Learning in a classroom or in a training course requires not a judgemental role but active participation. Whether something is learned depends to a large degree on whether the student is actively exploring the topic matter. Putting students in the role of a judge undermines this, I think.
Uttl et al. (2016): no relationship between student evaluations and teacher effectivenessEmpiric studies have thus far shown an inconsistent picture of the use of student evaluations. Previous studies have shown low to moderate correlations between students' evaluations and student learning. A new study by Uttl et al. (2016) shows a different finding. In their paper the authors explain that previous studies have given a skewed picture of the value of student evaluations due to small samples and publication bias.
Uttl et al. have done an up to date meta-analysis which shows that there is no relationship between student evaluations and quality of learning. They argue for abandoning student evaluations as a measure of faculty's teaching effectiveness.