为了帮助大家熟悉
分类讨论如果评估的目的是告知而不是评判则是没有问题的，但如果是评判教学质量就有问题了。
The answer to the question depends on the purpose of the evaluation. If the purpose of the evaluation is to inform, rather than to judge, then anybody is qualified to express an opinion which can be considered with respect to the organization's aims. On the other hand, if students' evaluations become a means for judging the quality of the teaching then there could be a number of potential misconceptions and misunderstandings:
"差学生"蓄意给老师打低分，好学生参与度不高。
One of my observations regarding the evaluation by students is, that more often thedisappointed and underperforming students express their dissatisfaction most intensively, while good and content students prefer to be quiet, or they are just like that, and often are reluctant even in participating in such surveys which distorts the real situation in a class and draws a rather bad picture.
数据有问题;领导没有安全感所以给老师打分;老师为了得到高分不负责的先给学生高分。
My personal experience reveals that (1) instruments used in teacher's evaluation by students are poorly developed and the data gathered from the survey have not been properly analyzed; (2) not every student is truthful in the evaluations; (3) insecure leaders of the school use the evaluation as a means to show that they are doing something to improve the learning environments and learning. However, the fact is they are not only insecure but also incompetent; (4) teachers who want to get high evaluations tend to lower their teaching or assessment standards so that students can get higher grades and therefore give more positive evaluations. In short, if the instruments are the traditional ones that we commonly encounter, then I advice against the practice.
In a university where the students are from diverse backgrounds, student evaluations will be less consistent, depending on each student's perceptions of the reasons for being there and expectations of the teaching process based on their previous experiences. Therefore, in addition to looking at the average ratings (which is the most commonly used index), evaluators will need to be trained to look at other characteristics of the data, such as the nature of the distribution, its skewness and its variance.