Below, Joaquín Maroto, Qiji Xiang, and Ryan Daher tell us a bit more about their project on Cohen’s Kappa Coefficient.
IE University is a place for forward-thinking visionaries to learn in a transformative environment. That’s why we decided to stretch our Bachelor in Data and Business Analytics students, challenging them to design and pitch short presentations to other students, graduates, and academics on various topics such as statistics, data analytics, and forecasting.
Briefly explain what your presentation was about.
Our presentation was based on Cohen’s kappa, a widely used statistic to measure inter-rater reliability for qualitative variables. We began by providing a brief introduction of the kappa coefficient, along with an example scenario which would help our audience to grasp the basic concept.
Next, we explained its benefits and limitations compared to other models, before working through an example to bring everything together and make it all much easier to understand.
What was your biggest challenge?
At the beginning, we faced two main challenges. First of all, we struggled to understand when to use the model and for which data types. Eventually, we solved this by testing it out using real-life examples.
The second, equally taxing challenge was the interpretation of the Cohen’s kappa coefficient. Judging the level of kappa that should be acceptable varies depending on the research paper and field, which makes application of the rule quite confusing. So we made sure to cover this issue in the presentation and included it in the limitations of the model, alongside suggestions on how to overcome it.
What did you enjoy most about working on this presentation?
We relished the complexity of the topic. The content was challenging and thought-provoking, especially when discussing the interpretations of the test results. We kept having to conduct further research to better understand the model and interpret the results. But even when addressing the challenges, which were extremely complex, we still found the work engaging. Finally, working as a team and motivating each other to learn new statistical concepts made the project very exciting.
How does this project relate to what you are learning in class?
The project is related to our probability and statistics class and the overall data and business analytics degree. In the first semester of our first year, we had an introductory course to probability and statistics. The following semester, we learned the basics of probability and statistics for data analysis, and began using our knowledge to solve real-world problems.
This semester, we’re studying more advanced statistical topics and learning a larger variety of tests. An important part of using rater-collected data is ensuring that the collected data is reliable and accurate. But we learned that one way to do this is using Cohen’s kappa coefficient, which is why we decided to base our presentation on it.
What are your conclusions?
We believe that the assignment has provided a great opportunity to improve our presentation skills, especially with regards to presenting new and more complex topics to an older audience. Moreover, we further developed our problem-solving abilities and deepened our understanding of how to ensure test accuracy and data reliability, which will undoubtedly come in handy moving forward with the degree.
Finally, working as a team on a presentation in which only one member is to present was an interesting challenge. It meant we learned how to delegate work fairly and be held responsible for our individual tasks, all with the aim of improving the overall outcome of the presentation.
Projects like these are fundamental in developing well-rounded and capable professionals who are able to tackle today’s challenges head on. Our students continually surprise us with their ability to understand complex issues and convey them in an interesting and accessible format. We love watching their hard work pay off and look forward to seeing them apply it in their professional life!