Some scientific research units think that scientific research workers are of high quality and profound knowledge, so they don't need training. They think that whether researchers can create achievements depends on their own research, so they refuse to train their employees. In addition, some scientific research units are eager for success, hoping to get immediate results. After training the employees, I found that it didn't achieve good results in a short time, so I gave up continuing training for the employees. However, they ignore that employees must keep growing, and growth also takes a long time. 2. The first problem of training evaluation mechanism is the fuzziness of evaluation mechanism. There is no clear regulation on whether it is necessary for employees to attend training. That is to say, in some scientific research units, it is impossible to know what kind of employees need to participate in training, when to conduct training, what kind of training to receive, etc. These are uncertain and lack of a strong system to regulate. The second is the arbitrariness of the training evaluation mechanism. Whether employees participate in training is often decided by leaders, which is directly related to the relationship between individuals and leaders. What may develop after "cronyism" is "cultivating cronyism" Cultivate people who have a good relationship with leaders and promote them more.