Teacher(s)
Language
French
Content
The course mainly concerns the analysis, interpretation and communication of survey results. To this end, the classic sampling methods (quotas, random, stratified, clusters, etc.) used to develop opinion polls are detailed, specifying their characteristics, their interests and limitations. The crucial elements of reliability of studies to which communicators and politicians must be attentive are highlighted. The main tools for analyzing survey data are then discussed with an emphasis on the interpretation and visualization of the results.
Course outline: methods of data collection and analysis
Part I: data collection
- Surveys: general
- Empirical sampling methods
- Probabilistic sampling methods
- Construction of a questionnaire: formulation of questions
- Margins of error: formula and characteristic curve
- Coding, Weighting, re-balancing of data
Part II: Data Analysis and AI
1. New AI tools for facilitating data analysis: analysis of documentation using LLMs and identification of variables related to a given prompt in large datasets.
2. Review of descriptive statistics and data visualization, including elements of factor analysis.
3. Identification of relevant information and relationships in a journalistic context, based on the approaches described in points 1 and 2 above, and construction of relevant summaries.
Course outline: methods of data collection and analysis
Part I: data collection
- Surveys: general
- Empirical sampling methods
- Probabilistic sampling methods
- Construction of a questionnaire: formulation of questions
- Margins of error: formula and characteristic curve
- Coding, Weighting, re-balancing of data
Part II: Data Analysis and AI
1. New AI tools for facilitating data analysis: analysis of documentation using LLMs and identification of variables related to a given prompt in large datasets.
2. Review of descriptive statistics and data visualization, including elements of factor analysis.
3. Identification of relevant information and relationships in a journalistic context, based on the approaches described in points 1 and 2 above, and construction of relevant summaries.
Teaching methods
The course takes place in the form of workshops. A survey database, possibly chosen by the audience, is analyzed throughout the course. Firstly, we criticize the way in which it was constructed, while secondly, we extract the main information using tools described in the content section above.
Course mainly given in person. Some videos can nevertheless be used (and discussed online) to improve understanding or add technical developments.
Course mainly given in person. Some videos can nevertheless be used (and discussed online) to improve understanding or add technical developments.
Evaluation methods
Work on real data with final report (to improve if a second chance evaluation has to be held).
Language of the evaluation: French
Language of the evaluation: French
Other information
Regarding the use of artificial intelligence tools, students are expected to use them responsibly, as defined in the 'Note on the Responsible Use of AI,' approved in July 2024 by the ESPO Faculty Board and available on the Moodle page of the course and the intranet of the faculty in the information for the students section.
Online resources
See course LCOMU2810 on moodle.
Teaching materials
- vidéos et documents (exercices, codes, démonstrations sur logiciel) sur moodle/videos and documents (exercices, codes, software demonstrations) on moodle
Faculty or entity
