Count Your Text: Digital Text analysis

In order for students to learn a more quantitative approach to analysis of literary texts students were introduced to text analysis with Python. Through 13 teaching sessions, the students first had to present on exemplerary case-studies, and later had to do simple coding tasks in Jupyter Notebooks. In this way, the students acquired coding skills and strengthened their analytical thinking skills, and finally presented their own hypotheses and data analyses to their peers.

Motivation

Digital methods have a great potential in the Nordic studies research field, where all parts of the subject (language, literature and media) have texts that are central to this type of analysis.

By implementing these as an elective, it makes it possible for the teachers to try out different ideas in terms of the content and didactics that relates to the integration of digital methods.

Use of technology

  • Python with a focus on Pandas- & SpaCy-modules
  • Github (sharing Notebooks and data)
  • Jupyter Notebooks (used for groupwork and the students’ own projects)
  • Guthenberg Projects (access to public domain literature)
  • Youtube (tutorials)

The students were asked to download the Python “Anaconda” suit. This package gives them access to all the necessary tools (Python, Jupyter, Spyder & Terminal). They then used Github to easily distribute Notebooks and text-data.

Please see the attached Links and materials for an example of Jupytor Notebooks in use, as well as teaching materials.

Outcome

Most students were optimistic about the elective, and some students made it further than initially anticipated.

In the evaluation of the course most students expressed that they felt equipped to work with this subject on their own, and in general, the students were engaged in the course and in concrete text analysis. Therefore, the elective exceeded all expectations.

Activities

The course was split in two parts based on what the students had to do as preparation and during class.

Preparation:

  • Over the course of the semester, the students had to read academic texts about digital humanities, case examples, theoretical and scientific issues as preparation.
  • Furthermore, they watched a curated list of videos on YouTube that introduced them to Python-programming. This included subjects such as: working with strings, lists, regular expressions or the functionality behind the Spacy-module.
  • Early in the course the students had to make group presentations on case analysis of DigHum-projects and later on, the presentation focused on their own Python-scripts and projects.
  • Finally, the students also had to perform reflection tasks as part of exam preparation.

During class

Every teaching session had some version of the following flow:

  • The teacher made a presentation concerning the texts that the students had read as preparation for the given class.
  • Each week, the student presentations were discussed (case analysis or Python-script).
  • The students worked together in smaller groups with the programming assignments for the given week. These assignments were prepared in Jupyter Notebooks by the teacher.

The teacher was available and walked around between the groups and participated in relevant discussions on codes, subjects, etc.

Ressources and support

Resources

  • Texts (case studies, theoretical texts etc.)
  • Jupyter notebooks with assignments
  • links for resources on YouTube.

Support

  • For the students’ presentations in class, they were provided with work questions in Notebooks.
  • Peer feedback and class discussions on their presentations and the reflection assignments.
  • The teacher helped with technical questions and issues.

Challenges and advice

  • One of the biggest challenges was getting access to data that the students were interested in. There were some ideas for exam projects, that could not be done as the texts were not available to us. It is therefore important to make it clear what is possible and impossible with the data available.
  • If students are interested in media texts and social media, it is important to be aware of legal issues and the GDPR legislation.

  • It is essential to find ways to promote the good aspects of using digital methods when you also have to say no to great ideas that may come from the students.

  • The development of an entire course like this can be very time consuming. Developing new teaching plans demands time, even if it is "only" a pilot project. As a teacher, you need to be aware of this and stay patient, as well as allocating more time than usual for preparing and implementing this type of course.

Ulf Dalvad Berthelsen on his teaching case:

Basic information:

Teacher: Ulf Dalvad Berthelsen

Faculty: Arts, Aarhus University

Discipline: Nordic studies

Course: Elective: Intoduction to digital textanalysis

Level of study: MA

Teaching method: Small class teaching, group work, supervision

Number of students: 15

Duration: Whole course

Academic objective

The course Count your text is meant to introduce the students to research on digital humanities with a focus on the use of digital methods in text analysis. Furthermore, the students are introduced to theoretical approaches and the methodological and scientific questions that are connected to this field of research. This course works with both literary and non-literary texts and is done mainly through case analysis.