Essays In Applied Microeconomics

Open Access
Akyol , Saziye Pelin
Graduate Program:
Doctor of Philosophy
Document Type:
Date of Defense:
June 17, 2014
Committee Members:
  • Kala Krishna, Dissertation Advisor
  • Kala Krishna, Committee Chair
  • Sung Jae Jun, Committee Member
  • Paul L E Grieco, Committee Member
  • John Raymond Moran Jr., Committee Member
  • School Choice
  • School Value-Added
  • Multiple Choice Tests
  • Major Choice
  • Gender Gap
CHAPTER 1 (with Kala Krishna) : Preferences, Selection, and Value Added: A Structural Approach This paper investigates two main questions: i) What do applicants take into consideration when choosing a high school? ii) To what extent do schools contribute to their students' academic success? To answer these questions, we model students' preferences and derive demand for each school by taking each student's feasible set of schools into account. We obtain average valuation placed on each school from market clearing conditions. Next, we investigate what drives these valuations by carefully controlling for endogeneity using a set of creative instruments suggested by our model. Finally, controlling for mean reversion bias, we look at each school's value-added. We find that students infer the quality of a school from its selectivity and past performance on the university entrance exam. However, the evidence on the value-added by schools shows that highly valued or selective schools do not have high value-added on their students' academic outcomes. CHAPTER 2 (with James Key and Kala Krishna) : Hit or Miss? Test Taking Behavior in Multiple Choice Exams This paper models and estimates students' decision to guess/attempt or skip the question in a multiple choice test in order to understand the role that student characteristics play. We do this using data from the Turkish University Entrance Exam, a highly competitive, high stakes exam. In particular, we investigate students' behavior according to their gender, predicted score and experience in the exam. Our results show that students' attitudes towards risk differ according to their gender, predicted score and exam experience: female students behave in a more risk averse manner relative to male students, and high scoring students are more risk averse. However, our counterfactual analysis suggests that although different testing regimes can lead to different score distributions, the relationships between exam score percentiles and student characteristics are relatively invariant. CHAPTER 3 (with Kala Krishna and Sergey Lychagin) : Gender Differences in College Major Choice: Evidence from a Centralized System This paper investigates factors affecting college major choice decisions of students. In an environment where college admission depends solely on students' college entrance exam score and their preferences, we found that gender difference in exam performance is an important factor, especially performance in math and science fields. Our results suggest that policies that can increase performance of female students to the level of male students can reduce the gender gap in engineering and technical science majors by half.