Training for the Future

Evaluating National Teacher Training Programs in Rural China


China's rural students lag far behind their urban counterparts in academic achievement.  If rural students lack necessary knowledge and skills, they will likely struggle to succeed as China's economy continues to develop.

There are many reasons for the persistent urban-rural achievement gap, but one likely contributor is the poor quality of teaching in rural schools. The flow of talent and rewards in China’s education system is almost entirely unidirectional: good teachers move to urban districts where conditions and incentives are better. This trend has left rural areas with a severe shortfall of experienced and qualified teaching staff. Through no fault of their own, rural students are left with less experienced and less qualified teachers. Studies from developed and developing countries, including China, consistently show that teachers are one of the (if not the) most important factors affecting student achievement.


In China's education system, better teaching conditions and incentives push good teachers out of rural areas.

In a number of developed countries, inexperienced or poorly qualified teachers receive high-quality teacher training to upgrade and reinforce their pedagogy. Indeed, the potential for teacher training programs to improve rural schooling outcomes is not lost on China’s education policymakers. Since 2010, the central government has invested close to a billion US dollars annually in the National Teacher Training Program (NTTP) to fund training activities across the country. The policy also specifically dictates that rural teachers should make up at least two-thirds of the training recipients. In fact, as part of this undertaking, each teacher in rural China is required to receive 360 hours of training over the next five years.


Teachers are required to keep track of trainings they have attended, but there is no evidence proving that these trainings are actually effective.

In spite of massive government investment in NTTP since 2010, to date there has been no rigorous assessment of teacher training programs at the provincial level. A national assessment (of teachers, not students) showed that teachers were “satisfied” with NTTP training. In general, teachers said that they believed that the training had had an effect on them. However, it is unclear whether these training programs are having any impact on student learning.  Furthermore, there has been no assessment of the effectiveness of particular types of teacher training programs in improving teaching practices and student performance.


The overall goal of the “Training for the Future” initiative is to assess the effectiveness of the NTTP for rural teachers in China today so as to understand how best to improve the teaching practices of rural teachers and ultimately improve rural students’ learning.  Our specific objectives are to identify what types of training programs improve (a) teaching practices and (b) student achievement in rural schools.

This project seeks to answer the following research questions:

  • Does a 15 day teacher training session provided by NTTP result in changes in teaching practices and student learning?
  • How do the impacts of these programs vary across demographic groups (age, gender, ethnicity, socioeconomic status)?
  • What aspects of the programs are successful and which are less successful (using causal chain analysis)? How can the interventions potentially be improved?


REAP is conducting a pilot study of teacher training programs in Shaanxi province this year, with plans to expand the project across Shaanxi province and into Henan province in the near future.  In this pilot program, primary school math teachers from 35 schools in Shaanxi were selected to participate in NTTP training sessions according to the existing teacher selection mechanism used by the NTTP. We refer to this group of teachers as our “treatment group.”

REAP will incorporate measures of both student and teacher performance into our evaluation of NTTP training programs.

In order to understand the impact of the teacher training intervention, it is vitally important to have a counterfactual. We can only assess whether the NTTP training had an impact on these treatment teachers if we can look at the outcomes of a similar group of teachers who did not attend the training. We therefore selected a group of teachers who were most similar to our treatment teachers based on location, teaching subject, and school characteristics. We refer to these “untrained” teachers as our “comparison group.”

In fall of 2014, REAP enumerators traveled to 70 schools in two prefectures of Shaanxi Province to administer baseline surveys and tests in the classrooms of teachers in our training group and our comparison group. Teachers were evaluated on teaching practices, attitudes, and math knowledge. Students were measured on learning and attitudes.

Next, teachers in the treatment group travelled to a central location in Shaanxi province to receive a 15-day NTTP training for primary school math, implemented according to standard NTTP procedure. Teachers in the comparison group did not attend the training.

Three months after the training, REAP will return to treatment and comparison schools to administer an endline survey. Again, teachers will be evaluated on teaching practices, attitudes, and math knowledge. Again, students will be measured on learning and attitudes. By comparing the change in performance before and after the teacher training session between our treatment group and comparison group, we will be able to assess the impact of attending the NTTP training session.


Results from Phase 1 of this project suggest that although NTTP has the resources and institutional capacity to train teachers effectively, the current model is not leading to improved teacher or student outcomes. NTTP's primary shortcomings are:

  • The vast majority of rural teachers do not have access to NTTP. Official NTTP attendance data show that only 4 percent of rural teachers in Shaanxi Province attended on-site training between 2011 and 2013.
  • Teachers are often not matched to trainings that fit their needs. For example, in the two NTTP mathematics trainings REAP evaluated, only 50.7 percent of the teachers in attendance actually taught math.
  • Although curriculum requirements appear pedagogically sound, implementation of actual training is very poor. Different instructors teach with minimal coordination, there is an overwhelming emphasis on  "moral education" and party principles, and sometimes no emphasis on how to teach.

Overall, there was no significant difference between the test scores of students whose teachers attended an NTTP session and those whose teachers did not. However, we believe that if a stronger NTTP model is developed, the NTTP management can implement it themselves. Therefore, we are working to expand our evaluation of NTTP. In Phase 2, we will conduct a larger, rigorous randomized evaluation of NTTP short-term on-site programs, NTTP online programs, and county-level teacher training programs in Shaanxi and Henan provinces. Results from Phase 2 are expected in summer of 2016. 


Funding for this project is provided by the Xu Family Charitable Foundation and the Ford Foundation.