Selected Research Projects
as Principal Investigator
Experimental Evidence of Systematic Phonics Instruction on
Low-Income Children's Early Reading Development
(2020 - 2021)
We examine the impact of systematic phonics instruction on 5,120 low-income 7th grade early readers’ English reading ability in rural China using a teacher-level randomized experiment. English teachers were randomly assigned to the treatment group to receive systematic phonics training. Students in classrooms with the treatment teachers scored 0.112 standard deviations higher than the students learning with the control teachers two months after the training courses ended. The effects persisted into the next semester. There were substantial shifts in teaching and learning attitudes and behaviors among teachers and students after the intervention. Our results indicate that systematic phonics instruction might play a pivotal role in the reading development of low-SES early readers.
The Earlier, The Better? Experimental Evidence on
the Timing of Systematic Phonics Instruction
(2021 - 2022)
While reading experts often suggest that systematic phonics instruction should be added to each literacy teacher’s teaching routine, little is known about the causal impacts of the timing of systematic phonics instruction on reading achievement. We provide experimental evidence on the impact of the timing of systematic phonics instruction for non-native early readers. In a teacher-level randomized experiment with 68 English teachers and 5,579 8th grade students in a low-income school district in Southwestern China, we compare students’ English achievement and learning attitudes in English classes taught by teachers who were randomly assigned to receive systematic phonics instruction training one year later than the other teachers who received the training the year prior.
Survey Experimental Evidence on the Strategies for
Teaching Early Readers to Read
(2021)
We carried out two survey experiments in the same RCT setting with the first and second experiment on the cohort of seventh graders (n=1760) who did not have experience with IPA nor with word decoding. 1760 students were randomly assigned to the treatment and control group by computer programming. The computer screen of the treatment students (n=880) only showed the decoded words, while the computer screen of the control students (n=880) showed the decoded words attached with IPA symbols. We used this estimation result to explore the effects of IPA symbols on spelling and word meaning.