Below you find a list of proposals for master thesis topics that can be supervised at our chair. Additional topics are available upon request but omitted from the list for confidentiality. Please note that each of the topics is tied to a specific supervisor. Availability of the topics may therefore depend on the availability of the respective supervisor.
Is teaching based on problem solving more effective than lecture style teaching?
Current research stresses the importance of teachers for student learning. However, it has not been settled so far what makes one teacher more successful in enhancing her students’ performance than others. Research by Schwerdt and Wuppermann (2011) indicates that lecture style presentations are not worse than problem solving in class, using TIMMS data.
Simonsohn et al. (2015) criticize that empirical results often hinge on data analytic decisions that are simultaneously defensible, arbitrary, and motivated. They suggest Specification Curve Analysis as a mean to mitigate these problems. Their approach suggests identifying all theoretically justifiable specifications, display their results graphically and then to assess how likely these results are to be driven by chance.
This master thesis project would need to replicate the main findings by Schwerdt and Wuppermann (2011) with a special focus on decisions the authors took when specifying their model. Subsequent Specification Curve Analysis would allow a closer look at the link between problem solving and lecture style teaching. Prior knowledge in Stata or R are essential.
112 emergency calls: Life-saving effects on cardiac arrest patients of a software package that enables 112 response staff to answer emergency calls in a structured way
In case of a medical emergency, people call the emergency telephone code 112. The response staff needs to decide what actions need to be taken and therefore has to receive all relevant information as quickly as possible. In case of a cardiac arrest this is a matter of life and death, since patients have higher survival probabilities if the occurrence of a cardiac arrest is recognized quickly and a reanimation (telephone-led by the response staff) is initiated as soon as possible. A new software package enables response staff to ask and record the important information during a 112 phone call in a structured way. This new software has recently been adopted by several German cities. The proposed master thesis project will evaluate whether this new software improves the capabilities of response staff to recognize the necessity of a reanimation and consequently improves the survival chances of people with a cardiac arrest. Data for the project will be provided.
The long-run effect of the 1918 Influenza Pandemic: intergenerational effects of the Spanish Flu
The 1918 Spanish Flu epidemic was the most severe influenza epidemic in recent history, killing more people than World War 1. It is known that people who experienced the Spanish Flu epidemic in early life had a worse health later in their lives. But does this early-life experience also affect the next generation? I.e. do the children of those who experienced the Spanish Flu in early life also have a worse health? Data to be used in this project are from IPUMS USA.
The long-run effect of the 1918 Influenza Pandemic in the USA: Does the data source matter?
The 1918 Spanish Flu pandemic was the most severe influenza pandemic in recent history, killing more people than World War I. It is known from previous research that people who experienced the Spanish Flu pandemic in early life were worse off later in life. The proposed master thesis project will check whether some of these effects stem from data harmonization efforts (making data comparable across time and countries) and, hence, should not be attributed to the Spanish Flu pandemic. To this end, analyses will be done on unharmonized data from IPUMS International and compared to results from harmonized data.
The Spanish Flu in Spain
The 1918 Spanish Flu epidemic was the most severe influenza epidemic in recent history, killing more people than World War 1 (WW1). Previous research showed that people who experienced the 1918 Spanish Flu epidemic in early life had a worse health later in their lives. But much about this is not clear yet. The proposed master thesis project will investigate long-run effects of the Spanish Flu in Spain. Unlike most previously studied countries, Spain remained neutral during WW1, so that effects of Influenza and effects of WW1 can be separated. The data to be used in this project are from IPUMS International, which contains very detailed data on Spain. (For this thesis, speaking Spanish is not necessary, but may be an advantage.)
More doctor visits after retirement?
After retirement, individuals have more free time to pursue health promoting activities. While there has been some evidence that retirees do exercise more – do they also use the time to go see a doctor more often to get help or to take preventative steps to stay healthy? This master thesis project is will investigate this research question. Data that will be used in this project are from the Survey of Health, Ageing and Retirement in Europe (SHARE). Regressing the number of doctor visits on people’s retirement status may lead to biased results (omitted variables bias). Therefore, an instrumental variable approach with country fixed effects will be used. Instrumental variables are countries’ official normal retirement ages (and perhaps also the official early retirement ages).
Smoking bans and smoking behavior
Different German states (Bundesländer) implemented smoking bans at different points in time. This master thesis project will study the effects of these smoking bans using a difference-in-differences framework. First, it can be studied how effective smoking bans are in reducing smoking. Second – and this is the main topic of the master thesis – it can be studied to what extent these smoking bans improved people’s health. Data to be used in this project are from the German Socioeconomic Panel (GSOEP). Depending on the Master student’s interests, one option could be to look at specific groups (young people who are in the age where you start smoking, pregnant women, poor people, other heterogeneous effects,…)
Mexico’s war on drugs: do elections increase violence?
During the last two decades, the number of drug trafficking organizations (DTOs), the spectrum of illegal activities they engage in, and their proliferation have increased in Mexico. This has led society to be exposed to different forms of violence and high crime rates. In 2007 the Mexican federal government made fighting organized crime a political priority, and started an aggressive military intervention throughout the country. This was led by President Felipe Calderon affiliated to the Partido Accion Nacional (PAN) political party. Some scholars argue that the massive intervention has aggravated the violence in Mexico, by disrupting the balance of power across DTOs. In your master thesis, you will explore the relationship between local mayoral election outcomes and these crimes. Are tight elections associated with an increase in crime rates? Do crime rates, on average, increase/decrease before/after elections? Are changes in local crime rates more pronounced depending on who won the election? Data to be used in this project are from the Centro de Investigacion para el Desarrollo (CIDAC) (data on local mayoral elections) and from the Instituto Nacional de Geografia y Estadistica (INEGI) (data on crime rates per municipality).