✋ Course Information
LSE ME204 (2026)
This page covers the teaching team, rooms and times, communication, assessment, and our AI policy. For the day-by-day plan, see the Syllabus.
👥 Teaching Team
Dr Jon Cardoso-Silva LECTURER
Assistant Professor (Education)*
LSE Data Science Institute**
AI and Education Fellow
* Becoming Associate Professor from 1 Aug 2026.
** Moving to Dept. Methodology from 1 Sep 2026.
Jonas Weinert CLASS TEACHER
PhD Candidate
Department of Health Policy, LSE
📍 Rooms and Times
Monday to Thursday each week. No class on Fridays.
📟 Communication
Two ways to reach us:
- Moodle forum: post questions here any time, including out of hours.
- Email: for anything specific to you, email directly.
Administrative queries (eduroam access, your LSE e-mail, Student ID, credits) go to the Summer School Office directly, not to us.
📚 Pre-requisites: what we expect from you
This course does not have a formal pre-requisite and is open to all, even complete beginners in computer programming. Beginners have done really well in past editions of this course, but expect a steep first week. As indicated in the official promotional pages, we recommend some familiarity with programming basics: variables, if-else logic, loops, and functions. If you’d like to prepare in advance, focus on Python basics; Chapters 1-5 of Automate the Boring Stuff with Python are a good free starting point.
Perhaps one implicit pre-requisite is the willingness to manage the natural emotional frustration that comes from computer programming! Specially if you are a beginner, you may find coding intimidating at first and you may feel tempted to “offload” the tasks to an AI chatbot for example. We advise that you persist on re-reading and trying things by yourself, consulting with teaching staff or reading official documentation and reserve a mindful use of AI such that you are always in control of your learning. (We teach some techniques on this topic during the course too).
✍️ Assessment Structure
Your grade consists of two components:
| 25% | Problem set | ✏️ Midterm |
Released: Thu 16 Jul 2026 Due: Tue 21 Jul 2026 |
| 75% | Website project | 🌐 Final Project |
Announced: Thu 23 Jul 2026 Due: Fri 31 Jul 2026, 5pm |
The exact time you need to submit the midterm will be confirmed in the assignment instructions. You’ll submit through Moodle, Nuvolos (the cloud platform used in this course), or GitHub, whichever the assignment instructions specify. Teaching staff will mark your submissions shortly after submission and you should receive feedback by Fri 24 Jul. The final project follows the same pattern, with grades due to the Summer School Office by Fri 7 Aug.
You’ll also have daily lab exercises that don’t count towards your grade, but check your understanding and get you ready for the graded assessments.
🤖 Generative AI Policy
LSE gives departments three official stances on generative AI to choose from, and this course adopts the most permissive, Position 3 (full authorised use) (
link to LSE position on Generative AI).
This means you can use generative AI freely for every assignment in this course, including the midterm and the final project.
Being allowed to use it is not the same as knowing how to use it well. Based on my research and other teaching I do, I teach specific techniques during the course for treating generative AI as a catalyst for your learning rather than a crutch: directing it with your own prior understanding, verifying output meaningfully, and modifying what it gives you rather than accepting it as it comes.
How each assessment is marked
We mark for evidence that you understood the work and directed it, not for whether every line of code runs. Work that meets the brief and runs may pass but higher marks are only justified by us seeing a clear reasoning about your data and the choices you made.
When each assessment is released, the brief includes a high-level marking criteria and internally, we have more fine-grained rubrics that helps us assess whether the submissions meet what we provided in teaching material and via oral delivery in lectures and classes.
How we use Generative AI when marking
We use generative AI in our own marking process too, alongside more traditional tools:
- We run a mix of Python scripts and predefined LLM routines, on LSE’s Enterprise Claude, so we don’t miss anything we care about in your submission, though this only produces an inventory, not a grade. This is similar to a practice in Computer Science education that goes back at least into the 1960s well before generative AI where machines are used to gather assessed information on submissions made by students.
- When we mark your submission, a marker reads it first to understand your thought process, taking their own notes. They then check that against the inventory our tools produced, a mix of deterministic checks (did this file exist, did this test pass) and the LLM’s annotation of whether there is evidence of engagement with the learning goals of the lectures and classes prior to the submission. For each criterion (for example, “Analytical Interpretation”), the marker decides a level (“Good,” “Very Good,” and so on) and writes down the evidence they saw. A second marker then verifies and ratifies the grade.
- We may also use LLMs to format that evidence into feedback that’s easier for you to act on.