1. médialab Sciences Po
  2. News
  3. MetAt - April 8, 2025 logbook

MetAt - April 8, 2025 logbook

Share our methodological expertise and skills.

Event, Workshop

⇥

Salle K.011, 1 place Saint Thomas d'Aquin 75007 Paris

NOTA BENE



What is METAT?

METAT is a research methods support workshop: every month, a three-hour slot to help you resolve the methodological difficulties you encounter in the course of a scientific project. 

Who is METAT for?

METAT is aimed at anyone needing occasional support in using a research tool or method. All profiles are welcome: students, doctoral students, researchers, research engineering professionals and others, inside and outside Sciences Po, with no restrictions on status or affiliation.

How to register?

Registration is compulsory via the form available on the METAT page.


Session of 8/04/2025

Location: Sciences Po, 1 place Saint Thomas d’Aquin, 75007 Paris. 

Supervisors: Béatrice Mazoyer, Maxime Crépel, Robin de Mourat, Benjamin Ooghe-Tabanou, Carlo Santagiustina, Dimitri Müller, Blazek Patel, Marine Chuberre, Lilla Conte


Learning how to use minet to scrape comments and images on Instagram and X

Support for a PhD student trying to collect comments and images from animal rights association accounts on Instagram and X. The supervisors (re)trained them in the use of minet for X. Their Instagram research account had unfortunately already been blocked, through scraping and the use of Firefox cookies (having failed to retrieve those from Chrome). The images were then collected using minet fetch. The DMI tool Zeeschuimer was also presented, to perform the same type of collection directly in the browser.

Qualitative analysis of an NVivo corpus

Support for a student working on a master's thesis on conflicts in public spaces in the La Chapelle district of Paris. They developped a corpus of official documents, press articles and interview transcripts that they would like to use as part of a qualitative analysis. The materials were deposited and structured using NViVo software. The supervisors and the student reviewed the files and their classification, as well as the codes used. Discussions focused on how to classify and categorise these materials according to different types of research questions (e.g. ‘Who talks about what? Who is talking about whom? Who is doing what (to whom)? They imagined how the partial reorganisation of existing codes and folders could help students to continue and complete their analysis. They also explored the tool's visualisation functions and discussed their relevance to a qualitative approach.

Analysis of TikTok posts and other textual data

Support for a doctoral student and a student in the use of automatic language processing (NLP) tools such as spaCy to analyse textual data. The participants want to analyse unstructured textual data: identifying common themes within 20,000 TikTok comments from 50 videos for one, and identifying themes and interconnections among the work of 50 students for the other. They were given advice on libraries and automatic language processing (NLP) techniques. The session shared examples of practical code for extracting and consolidating Named Entities, Part-of-Speech (POS) tagging and Dependency Relations. The participants then explored how to build ego-networks to visualise and study the relationships between the entities mentioned in the text. Finally, the supervisor shared the code for running a Structural Topic Model (STM) that incorporates both the unigrams and the dependency relations previously extracted as input. The code was adapted to the TikTok dataset that the participant had collected.