Collaborative Learning in Data Science Education: a Data Expedition as a Formative Assessment Tool
1 like157 views
The document discusses the use of collaborative learning techniques, specifically data expeditions, as a formative assessment tool in data science education. Data expeditions involve small teams creating data-driven media artifacts over 2-3 weeks, focusing on digital media creation and evaluation criteria such as data citation and reproducibility. The findings also highlight the need for structuring courses to include critical thinking and practical experience while incorporating feedback for future improvements.
Collaborative Learning in Data Science Education: a Data Expedition as a Formative Assessment Tool
- 1. Collaborative Learning in Data Science Education: a Data Expedition as a Formative Assessment Tool ICL-2018, Kos Presenter: Maksimenkova Olga junior research fellow IL ISSA, Faculty of Computer Science, National Research University Higher School of Economics, Russia Β© Maksimenkova O., Neznanov A., Radchenko I. 1
- 2. Our Working Group β’ From left to right β’ Radchenko Irina β’ ITMO University, Saint-Petersburg, Russia β’ Maksimenkova Olga β’ NRU HSE, Moscow, Russia β’ Neznanov Alexey β’ NRU HSE, Moscow, Russia
- 3. Data Science βData Science is an interdisciplinary field that combines machine learning, statistics, advanced analysis, and programming. It is a new form of art that draws out hidden insights and puts data to work in the cognitive eraβ (IBM Analitycs) Β© Maksimenkova O., Neznanov A., Radchenko I. 3 This image license: CC BY-SA
- 4. A Data Scientist Should Β© Maksimenkova O., Neznanov A., Radchenko I. 4 be of an analytical and exploratory mindset have a good understanding of how to do data- based (quantitative) research possess statistical skills and be comfortable handling diverse data be clear, effective communicators with the ability to interact across multiple business levels have a thorough understanding of their business context
- 5. Principles of Teaching Data Science Β© Maksimenkova O., Neznanov A., Radchenko I. 5 This picture license: CC BY-SA-NC organize the course around a set of diverse case studies integrate computing into every aspect of the course teach abstraction, but minimize reliance on mathematic notation structure course activities to realistically mimic a data scientistβs experience demonstrate the importance of critical thinking/skepticism through examples
- 6. What is a Data Expedition? β’ βData expedition (DE) is a learning by doing computer-supported collaborative learning technique, which is applied to Data Science Educationβ β’ Creation a data-driven media artefacts β’ 2-3 weeks β’ Collaboration in small (2-3 persons) teams Β© Maksimenkova O., Neznanov A., Radchenko I. 6 This picture license: CC BY
- 7. A matter of evaluation: How to evaluate a data expedition? β’ What to evaluate? β’ The processes of digital media artefact creation β’ The quality of data citation β’ Correctness and reproducibility β’ The complexity, interactivity, and design of final artefacts Β© Maksimenkova O., Neznanov A., Radchenko I. 7 This picture license: CC BY-NC-ND
- 8. Data Expedition Evaluation Criteria Β© Maksimenkova O., Neznanov A., Radchenko I. 8 Data expedition dairy (experiment journal) for research data expeditions Using scripts written in one of the programming languages Using special software for data preprocessing or analysis Programming code Data tiding Data analysis Data visualization This picture license: CC BY
- 9. Different Learning Environment β’ Integration, communication, and management tools β’ Rich collaborative environments (Microsoft Office 365, Google G Suite) β’ Project management tools (Trello, Microsoft Teams, MeisterTask) β’ Online communication tools (Microsoft Skype, Telegram) β’ Search and external discussions tools β’ Universal search engines (Google Search, Microsoft Bing) β’ Specialized search engines (Wolfram Alpha) β’ Social Media Services (Facebook, Twitter, Instagram) β’ QA services (Data Science Stack Exchange, Public Lab, Quora) β’ Data processing tools β’ Universal cloud data storages (DropBox, Microsoft OneDrive, Google Drive, Yandex Disk) β’ Code repositories (GitHub, BitBucket, GitLab) β’ Data analysis and visualization tools (R Studio, Jupyter Notebook, Orange, Microsoft Excel, Microsoft Machine Learning Studio, Tableau, OpenRefine, Infogram, Plotly) β’ Open data platforms (CKAN, Zenodo) β’ Specialized data extraction tools (Import.IO, Kimono, Scrapy, ABBYY FineReader) β’ Data validation tools (CSVLint, Data Package Validator, Data Hub LOD Validator) β’ Media publishing platforms β’ (Medium, Microsoft Sway, Tilda, GitBook, Wordpress, Blogger) 9Β© Maksimenkova O., Neznanov A., Radchenko I.
- 10. Implementation in Universities Β© Maksimenkova O., Neznanov A., Radchenko I. 10 2016-2017 academic year β’ Data Journalism, first-year master students, National Research University Higher School of Economics, Moscow β’ Students and free-listeners European University, Saint-Petersburg Two different stories
- 11. Post-survey (HSE) Β© Maksimenkova O., Neznanov A., Radchenko I. 11 Question 33% 67% Was your participation to the data expedition useful for you? Yes Probably, yes How was your time allocated during the data expedition? Little by little, but regularly I worked not regularly. Sometimes for a long, sometimes there were no time at all Give a characteristic of your interaction with the other participants? I prefer working alone I had enough interaction, I am satisfied
- 12. Post-survey (HSE) Β© Maksimenkova O., Neznanov A., Radchenko I. 12 β’βI have created universal code which I am going to use in the other projectsβ β’βData expedition is a completely new for me type of a project. So, to accomplish the task the time had to be distributed in a separate way. But, each project goes to βExperienceβ storage and this is goodβ β’βPractical experience in writing data-driven articleβ What were useful for you in data expedition? β’βOn the one hand, we were free in the selection of topic but on the other the result would have been limited stricter. I mean clear requirements to our work. Moreover, it would be better to work with new not previously publicized data (not only links to open data portals which are we familiar with and have no interest). But the freedom in topic selection should be keptβ β’βI did not like that we immediately got down to independent work. At the lessons before the data expeditions we only clean data. It would be better to pass through all the stages with teachersβ controlβ β’βI wish more practice before a data expeditionβ What were not satisfactory for you in data expedition?
- 13. Post-Survey (EU) 13Β© Maksimenkova O., Neznanov A., Radchenko I. Questions Yes, % No, % Have you previously worked with open data? 28,6 71,7 Have you previously worked with online services for open data visualization? 7,1 92,9 Have you previously written scientific papers in your native language? 92,9 7,1 Have you previously written scientific papers in English? 35,7 64,3 Is it interesting for you to learn about Open Science and its implementation for research? 92,2 7,1
- 14. Recommendations β’ Evaluation should take into consideration two aspects of a DE: β’ orientation to the predefined goal β’ multi-stage nature of main Data Expeditionβs processes β’ Predefined goal allows one to check correspondence between task description and final digital media artefacts β’ Multi-stage nature of DE processes allows one to assess involvement and impact of participants based on DE diary and logs β’ The features of digital media artefact which is created during a DE should be reflected in grading rules β’ Proposed rubrics should be discussed with Data Expeditionβs participants β’ Grading of data citation, reproducibility and provenance should follow available guidelines 14Β© Maksimenkova O., Neznanov A., Radchenko I.
- 15. Future Work Assessment in form of Data Expedition should be implemented after a short supportive block which aims to refresh relevant knowledge and skills Working groups should be completed according to rubrics Β© Maksimenkova O., Neznanov A., Radchenko I. 15 This picture, license: CC BY
- 16. Contacts β’ Contacts: β’ Irina Radchenko β’ E-mail: iradche@gmail.com β’ Web-site: http://iradche.ru/english/ β’ Alexey Neznanov β’ E-mail: aneznanov@hse.ru β’ Web-site: https://www.hse.ru/en/staff/aneznanov β’ Olga Maksimenkova β’ E-mail: omaksimenkova@hse.ru β’ Web-site: https://www.hse.ru/en/staff/maksimenkova Β© Maksimenkova O., Neznanov A., Radchenko I. 16