Skip to Main Content
 

Global Search Box

 
 
 

ETD Abstract Container

Abstract Header

Peer Effects: Evidence from the Students in Taiwan

Wu, Shin-Yi, WU
http://rave.ohiolink.edu/etdc/view?acc_num=osu1500150214246876

Abstract Details

2017, Doctor of Philosophy, Ohio State University, Economics.
Since most of the recent research does not consider time lag dependent variables when analyzing peer effects, the estimation of peer effects might be biased. Moreover, whether or not peer effects are persistent is also an issue. The Panel Study of Taipei Youths (PSTY) provides a great opportunity to investigate these issues. I use pooled spatial autoregressive (SAR) models, including time fixed effects, time lag dependent variables and spatial time lag weights matrices to estimate peer effects more precisely. I find that peer effects are not persistent in the sense that students' friendships from previous period do not influence their current behaviors. Furthermore, peer effects might be overestimated without controlling for time lag dependent variables and time fixed effects. Robustness check is verified by the analysis of four specifications and two samples different in sizes. Separate estimation results for each wave from SAR models suggest that peer effects become stronger over time. Empirical evidence also shows that the source of peer effects comes from the change in students' utilities over time by eliminating two other alternatives, which are students sorting more over time and the existence of the lasting effects of students' previous friendships. Luminous Shine and Dark Shadow (LSDS) provides rich information regarding the types of friendship and high frequency friendship network survey data. Such unique feature allows me to investigate peer effects among different types of friendship and the trends of them over time. Results show that the major source of peer effects on students' academic performance is from students' academic friends who are also among their best three friends, with the strongest magnitude among all types of friendship. A non-academic friend of one student does not have effects even if he is one of the student's best three friends. Degree of peer effects starts to increase from 7th to 9th grade but remain constant from 11th to 12th grade. The trend of peer effects of the best three friends fluctuates more than that of academic friends. I continue to focus on the network formation. Empirical evidence of PSTY suggests that one student's friendship formation is significantly influenced by his previous friendship, popularity and religion. Robustness is checked from estimation results by implementing logistic model and fixed effects logistic model on cross-sectional and panel data. The estimation results of LSDS through a multinomial logistic model show that students will like, hate or like and hate simultaneously those classmates they liked, hated or liked and hated simultaneously last period rather than ignore them this period. This implies that the worst thing is to be unknown (not friend) among classmates if a student prefers to make as many good friends as he can.
Lung-Fei Lee (Advisor)
Bruce Weinberg (Committee Member)
Kurt Lavetti (Committee Member)
180 p.
Economics
peer effects; spatial autoregressive; SAR; friendship network; Taiwan

Recommended Citations

Citations

  • Wu, WU, S.-Y. (2017). Peer Effects: Evidence from the Students in Taiwan [Doctoral dissertation, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1500150214246876

    APA Style (7th edition)

  • Wu, WU, Shin-Yi. Peer Effects: Evidence from the Students in Taiwan. 2017. Ohio State University, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu1500150214246876.

    MLA Style (8th edition)

  • Wu, WU, Shin-Yi. "Peer Effects: Evidence from the Students in Taiwan." Doctoral dissertation, Ohio State University, 2017. http://rave.ohiolink.edu/etdc/view?acc_num=osu1500150214246876

    Chicago Manual of Style (17th edition)

osu1500150214246876
368
© 2017, all rights reserved.
This open access ETD is published by The Ohio State University and OhioLINK.