The modern age is overwhelmed by vast amounts of new and often sensible data, containing both economic and scientific perspectives and social risks. This led to data literacy being a key competence for many departments of our university.
The goal of our cross-faculty data literacy initiative is to enable our students to work with modern data in a skillful, critical and responsible way; making our alumni more competitive and sustaining East Bavaria as a center of education, research and industry. Our motivation is simple: more data literacy for more students. No faculty is excluded. Our strategy is simple: data literacy needs a good soil to flourish. This is why we want to invest in substrates like:- Mundane contact to modern data-driven science
- Strengthening of the key competences computer programming and statistics
- Optimal coordination between courses
- Boosting cross-faculty Cooperation
- Informed responsibility for handling sensible data
- Proactive students
- Fun
Data Literacy Education
Data literacy affects everyone: no department of our university can afford to ignore the challenges of digitization and its flood of data. On the contrary, every faculty is able to contribute, from epmirical sciences over mathematics and low to ethics and even theology. Even if questions and issues strongly differ, methods are often indentical no matter what they are used for. Also if you want to evaluate them in an ethical, sociological, psychological or juridical way instead of using them you still have to understand how they work. This is why we want to bundle data science competences in all their facets.Motivation
To raise awareness of students for the benefits of utilizing data science methods we are going to incorporate teasers (2-4 hours) in the introductory lectures in the respective fields of study. Additionaly, a lecture series of teachers presenting their disciplines and use of data science is created to show the importance of cross-faculty exchange.
At the moment we are working on expanding this list with contributions from economical informatics, marketing, media informatics, and psychology as well as units on law and ethics.
The teasers and the lecture series are also used to recruit students for the incubator.
Up til now the following topics are confirmed:
- Stylometry (literary studies)
- Pedestrian navigation system of the Universität Regensburg
- Alexa
- Search behaviour on the internet
- Behaviour of students in class (lectureship)
- Structural data of genomes
- Personalized therapeutic decisions (Med school)
- Structural data of proteins
- High performance computing and artificial intelligence
- Brain-Imaging (Neuro-Science)
- Ecological models
Data 101
This is the only course that's mandatory for both certificates. It is based on the Swirl environment "Learning R in R" and teaches students to deal with data science issues in R. To accomplish this they are analysing real-world datasets from football (soccer) scores via the Clinton emails to intestinal microbiota constituents.
The priority lies on working with data on the computer and not on statistic procedures and their mathematical principles. It is a main goal to lead students to use programming languages like R or Python, as they are more versatile and powerful than Excel and allow unambiguos, reproducible and well documented analyses. Case studies are deliberately designed to be not managable in Excel. A prototype of this class was developed in conjunction with mature students and already field-tested with bachelor students of several faculties.
Key competences
The key competences are computer programming and statistical thinking, enhancing the existing course portfolio. Students are taught fundamental concepts and modes of thought of informatics and statistics. The used programming languages are Python and R, beginner classes are already established and getting tuned to the needs of consecutive courses, which instead of being advanced classes focus on permanent use of the languages in data analysing projects. At this point connectors–seminars linking data science to a field of application (see connectors)–play an important part. In most fields of study classes in statistics are adjusted to the respective necessities and traditions. We will open at least one of these lectures to sutdents of al faculties.
Advanced methods
While the basic data literacy certificate focuses on key competences, Data 101 and connectors, the advanced data literacy certificate is designed to be more profound and mathematic. Current employment ads use tags like machine learning, artificial intelligence, deep learning, Big Data, data mining, image analysis, time series analysis, regression, spatial models, Bayesian methods or simulation of complex systems. All these are advanced techniques rely on basic knowledge in data management, programming and statistics, but still require additional training. This is why we also want to estalish an attractive range of courses in advanced methods, including modules in data visualisation, machine learning and utilization of the machine learning software TensorFlow and practical courses in super computing. Many of these classes already exist at the different faculties and only need to be optimized for the previous knowledge of the certificate candidates.
The key to advanced data science methods is mathematical knowledge, especially in linear algebra, probability theory, numerical analysis and optimatlization. It is planned to install an individual course for data science education coupling mathematics directly to data science issues, thus creating access to the advanced data literacy certificate for non-mathematicians.
Connectors
These modules are rooted in the fields of study, take up data science methods and use them. As with the methods classes there already are several courses at the faculties that may double as connectors, again giving the chance to refine them based on a common education in key competences and advanced methods.
Connectors require at least the attendance of Data 101, in most cases students would benefit from the completion of additional modules like key competences. Current connectors are described on the website of information science (german language).
Social context
The sixth column is a cross-cutting issue that will be adressed in an innovative concept incorporating external actors instead of relying on classic teaching methods. Details can be found in the incubator section.
Data science certificate
We are developing a data science training for students of all faculties in which they learn do utilize modern data science methods in their respective fields of study. They can obtain to levels:Basic Data Literacy Certificate
For students without previous knowledge
Contents
- Lecture series: Data across the sciences
- Data 101
- 2 courses in key competences
- 1 connector
Advanced Data Literacy Certificate
For students with basic knowledge in computer programming, statistics or machine learning
Contents
- Lecture series: Data across the sciences
- Data 101
- Mathematics for data science
- 1 course in advanced methods
- 1 connector