Course content
Data Gathering and Organization:
Students will learn about surveys, representativity, response rates, and visualizing survey results. Key statistical concepts such as means, medians, standard deviations, and interquartile ranges (IQR) will be covered, alongside covariances and correlations. Practical work will include scatter plots, box plots, histograms, and kernel density plots.
Data in Firms:
A guest lecturer from the business sector will discuss data gathering from customers and transactions, current uses, and ambitions for data analysis as well as challenges related to GDPR. A supporting session will provide guidance on GDPR considerations for students starting their master’s thesis or gathering data in other contexts.
Types of Data:
The course will cover cross-sectional data, time-series data (mostly with attention to pitfalls when encountering time-series, not time-series modelling), panel data, and unstructured data.
Causality and Experimental Methods:
Students will be introduced to potential outcomes, randomized controlled trials (AB-tests), and the decomposition of sample means into selection effects and causal effects. Practical exercises will include analyzing data from marketing AB-tests and performing standard t-tests for comparing two population means.
Multiple Regression:
Students will learn about multiple regression models, OLS estimation, omitted variable bias, and regression anatomy. Practical sessions will involve modeling trends in business-relevant datasets, visualizing multiple linear regression models including interaction effects between dummy variables and continuous variables.
Difference-in-Differences Design:
The course will cover the difference-in-differences design as both a graphical and regression-based method, with practical applications in finance, where this method is widely used.
Forecasting and Time-Series:
Key concepts such as stationarity, differencing, and using linear regression for forecasting will be discussed. Students will also explore machine learning and AI algorithms for predictive analytics, with practical cases utilizing financial data.