# Hands-on Data Anonymization February 2023 chat :::warning Please do not write personal information on this document. ::: ## Icebreaker 1 Write something about your work and the type of data you work with, mention if you were at the workshop this morning - I use video capture from classroom teaching and learning to study features of teaching quality across Nordic classrooms. For example I am interested in students opportunities to engage in class discussions thus investigating students opportunities to talk and how the other students and the teacher build/ use these utterances drawing on the video data, accompanying audioes and selective transciptions based on these videos/ audios. (I was in the Breaing) - I study what is at stake in children and young peoples everyday lives living in residential care. I use quatlitative methods, especially semistructured interviews and field work. I went at the workshop this morning - I study so-called patriotic education in Russia and in occupied parts of Ukraine, using in part social media sources - giving rise to many ethical and legal challenges. Qualitative analysis. Present at the seminar. - I work with researchers who conduct studies on school students and teachers through video observations and surveys. I was at the morning session. - In my work, we deal with fMRI collected from human subjects, so the data is usually time series of brain activity and also structural MRIs. I was at the seminar this morning - I have done interviews of a small number of people for a scholarly article. - I have conducted language tests, had interviews, collected surveys about linguistic background. I was at the presentation this morning. - I haven't been to the morning workshop - The data I usually work with includes all sorts of information that is used in behavioral research (i.e., questionnaires and test data, video data, biomedical data, sometimes interviews) - In my work I conduct in-depth interviews with refugees that I intend to cite anonymously. I have not been present on morning workshop - I primarily work with interview data related to political views and action, in different contexts and on different scales (e.g. at the level of a specific neighbourhood in relation to a local protest movement, or with broader groups of respondents such as practitioners and young people of different backgrounds within the context of larger cities). - In my work, we deal with sensitive qualitative data, interviews, field conversations and observations, relating to political violence, religion,racism and democracy. - Data on language abilities in children with developmental disorders - I look at the challenge of sharing qualitative (social anthropological) data in a short term project - I work with video recorded lessons, I wasn't at the lesson this morning. - I - Fun fact: I've just been interviewed by master students on two separate occasions. I "auto-anonymised" my data while answering, making sure I could not be identified - I was not in the morning session. - ... ## Icebreaker 2 Spend 5 minutes browsing the [slides from the morning workshop](https://docs.google.com/presentation/d/1dxK-7PrIcl73laNcQu1VkJ3D7F4iV8gr0CX0jXXX8Wc/edit?usp=sharing) (even if you did not join the workshop). Write down something that needs to be clarified, or expanded, or something you want to discuss. - How to pool data to ensure higher k-anonymisation - How to deal with visual data - Part 2 will cover this - Different standards/requirements for anonymization/pseudonymization/minimization in s; data collection, storage, handling, dissemination and academic publishing Document with DPIA matrix: https://ico.org.uk/for-organisations/guide-to-data-protection/guide-to-the-general-data-protection-regulation-gdpr/data-protection-impact-assessments-dpias/how-do-we-do-a-dpia/ - maybe a little bit about the motivations for why we want to or need to anonymize data and when it is or is not relevant - More on the difference between anonymisation and pseudonymisation - Can pseudo-anonymised data be shared? And the question of ethics, what happens if a data subject asks to be removed from a dataset since they can recognize themselves - I would like to discuss how personal data minimalisation in quatliative research relates to Open science. Is it in order to share with other researchers, or is it in order for other persons (also non-researchers) to re-analyse and validate or de-evaluate findings that is at stake for them. I - Enrico to add a link to the trento talk on youtube by Giorgia Bincoletto - When is qualitative data completely anonymized? - Does the removal of personal data from online sources from the web change the way you can use them in research? - Add a link on online research - How to deal with the possibility that what is anonymized today might be open to re-identifician in the future? - What are the legal responsiblities for individual researchers c.f. institutions ### BONUS on "consent": - some clarification on "consent" since in Norway things are a bit different than in finland 1. "ethical Consent" (I wanna be part of your study) 2. "legal consent" (I accept these cookies in my computer) 3. "consent as a basis for GDPR" (I am aware that you process personal data in a lawful way and your way of processing it is that I am consenting (not explictly) for you to do it by being part of this experiment) (versus the "public interest as a basis for GDPR") ## Questions on part 1 - ... - ... - ... Those of you who did not register for the workshop, it would be great if you could still submit your registration (for our records) here: https://nettskjema.no/a/314516 --- ## Part 2 ## Exercise 1 https://docs.google.com/document/d/1AH1JWOTdzEcUBM3G1l394LWPyLmF6tgLwJVat_cLOjQ/edit?usp=sharing Be a robot and anonymize all those identifiers ### Questions - What was "wrong" with this task? - What was difficult? --- # Part 3 ## Exercise 2 1. Open the link with the census data https://docs.google.com/spreadsheets/d/1ftabUDxG3o7T-A0pKYjxH4nGIvoRPWY3evpELycoRaQ/edit?usp=sharing 2. make a copy of that spreadsheet and name it as your group number 3. discuss the following question: ### Which columns would you delete to make the data anonymous? --- ## Exercise 3 You are a peer reviewer and you need to look at the method section of a paper and the data that they attached to the paper. Are the claims from the methods section reflected in the data? What would be your comments as a reviewer? Paper + data : https://docs.google.com/spreadsheets/d/1mHvshwjQiCm2y10aGlkIjnNStzxrCZfcM9gqg2GX3eU/edit --- # Part 4