Institution: Graduate School at Faculty of Business, Economics and Social Sciences – Universität Hamburg
Lecturer: Prof. Peter Moffatt, PhD (University of East Anglia, GB)
Schedule:
Do., 16.06.16
Fr., 17.06.16
Place: Universität Hamburg, Von-Melle-Park 9, Room A 510
Registration: You can register for the course until 14.04.16 (13:00) via Geventis
Course description:
PART I: Theory Testing, regression, and dependence
1. Introduction
2. Experimental Auctions
2.1. Overview of auction theory
2.2. Carrying out an experimental auction
2.3. The simulated auction data
3. Tests of auction theory
3.1. A test of RNNE in a second price common value auction
4. Tests of comparative static predictions
4.1. Standard treatment tests
4.2. Treatment testing using a regression
4.3. Accounting for dependence: the ultra-conservative test
4.4. Accounting for dependence in a regression
4.5. Accounding for dependence: the block bootstrap
5. Multiple regression with auction data
5.1. Introducing the effect of uncertainty
5.2. Introducing the effect of experience
5.3. Introducing the effect of cash balance
6. Panel data estimators
6.1. Random effects versus fixed effects
7. Multi-level modelling
8. Modelling data from contest experiments
8.1. The Tullock contest
8.2. A contest experiment
8.3. Analysis of data from a contest experiment
9. Meta Analysis
PART II: Dealing with heterogeneity: Finite Mixture Models
10. Introduction
11. Mixture of two normal distributions
12. The fmm command in STATA
13. The level-k model for the “beauty contest game”
14. A public goods experiment
14.1. Background
14.2. Experiment
14.3. data
14.4. The finite mixture 2-limit tobit model with tremble
14.5. Program
14.6. Results
14.7. Posterior Type Probabilties