The ANU Online Courses in Political Analysis, an online program for graduate students offered by the Australian National University, was held in July following February 2021. This time, two Waseda graduate students from the Graduate School of Political Science and the Graduate School of Commerce participated in the program. Like last time, it was a very valuable opportunity for students to learn data analysis directly from world-class lecturers. The Center for Positive/Empirical Analysis of Political Economy subsidizes the ANU online program tuition fee, providing graduate students with even more international opportunities.
Affiliation/Grade: Graduate School of Political Science, M2
Research Interests: History of Political Thought
Course Attended: Foundations of Statistical Analysis in Political Science
The course Foundations of Statistical Analysis in Political Science is suitable for students with a solid foundation in mathematics and programming. The course time was limited to only 14 hours, but the teacher’s explanations were very clear. This course seeks an understanding of the published correlations in the research literature on political science. In this course, students had the opportunity to write their own data to learn how to use various types of data. The professor used R-Studio and set up R tasks. In this class, I always read the code with my professor and group members.
I belonged to a group with a student from India, who not only showed me his code but also introduced his very beautiful hometown in India with the sea and beaches.
At the same time, I learned about linear regression, which means to predict trends in certain circumstances, find laws, and find suitable formulas. But, code is completely meaningless if there is too little data to research. In addition, this course is aimed at students who have the ability to process a lot of data. Since many machine-learning toolboxes now provide linear regression functions, it was used in this course as well, so I think one needs to learn R language before starting this class.
This course is very useful for my future research plans. For instance, I am interested in the history of digital political thought. In the future, I would like to use software to build a database (DB) and clarify the origins of East Asian policy ideas and understand the transitions of British and American policy ideas for East Asian waters. In this context, I will analyze products that have been distributed throughout the East Asian region after completing the data collection, organization, and examination. Then, I would complete a diachronic analysis of East Asian policy ideas in East Asia. Thus, I am glad that I joined this course because I learned how to better process and analyze data.
Affiliation/Grade: Graduate School of Commerce, M2
Research Interests: Accounting Information
Course Attended: Introduction to Political Text Analysis
My achievements from the ANU online course are as follow.
For one thing, since this course basically uses R and a text analysis package called quanteda, I had a great opportunity to learn and practice the coding by R. The lecturer, Prof Beniot, showed extremely specific process to apply all the text analysis methods to the political article or speech, including installing all packages, importing dictionary and modifying the dictionary, applying lexicons (dictionaries) to texts to measure sentiment or topics and so on. These contents gave me a clear structure and workflow to develop a text analysis even though I only had limited knowledge and experience in R and text analysis.
Furthermore, due to the material used in this course which is mainly speeches used in political area or elections, I learned both how to measure text data in a quantitative way and how to use appropriate methods to measure text data in different areas. As for my own project, I want to use text data about corporations’ risk management in annual reports to evaluate the quality of the risk management applied in the companies. However, these data are basically in text form used to be ignored or not been seen as data. From my perspective, the sentiment analysis and similarity analysis could be used in my own project to detect what different between firm with highly performed risk management and firm with poor risk management quality. I also noticed that the measurement about using sentiment analysis and keyword-in-contexts analysis through a long period could be used in my project to bring an unexpectable result which could be interesting and meaningful in my future research.
The whole course is very informative and operational, and the lecturer is extremely skillful in both teaching and programming. I wish that we could do the assignment by ourselves first and then receive the lecture’s instruction; the analysis detail and the code would make more sense to me.