TDDD89 Scientific Method

Seminar 5

Seminar 5

Purpose

The purpose of Seminar 5 is to practise writing a more complete thesis plan including a project time plan where you, starting from the main scientific method(s) to be used, outline necessary activities that will go into your work, as well as describe milestones and analyze potential risks.
Also, you will practice reviewing such as thesis plan.

Preparations

First, revise your thesis plan according to the feedback you received at Seminar 3 (and earlier). You should now include at least four references to related publications in your Related Work section, and summarize and discuss them properly.

Now start the Project Plan chapter with a subsection (one paragraph) about the intended main method(s) (cf. Lecture 5) to be followed, which usually results from your research questions.
You will also need to include a few references to method papers and books, i.e., published work describing how studies are generally conducted in your area of research. You might make use of the references given in preparation for Seminar 2, references in Lecture 5, or other publications that describe established best-practices for how to study the kind of phenomenon you are interested in as part of your master's thesis.

Next, you will need to add a project time plan including main tasks and milestones, and a risk analysis to the project plan chapter. Instructions are given below.

At this stage, your extended thesis plan is expected to be close to 10 pages long, including references.
Note that we expect at least 10 pages for the UPG1 submission in january.

Submit your revised plan via LISAM (Seminar5/group) no later than Tuesday before the seminar, i.e., 12 december 2023, and then (individually) read and review the other 3 submissions of your topic group according to the grading rubric before the seminar. Your individual review notes should be available in Lisam at seminar start.

Time plan

Identify the main tasks required in the project towards obtaining results that will finally allow you to answer your research questions. Identify dependences between tasks.

Add a time plan to your thesis plan with weekly activities that are mapped to the above tasks. Complete with mid-thesis review and final presentation.

You should by now have gained some understanding of how long it takes to write different parts of your thesis, given the time it has taken you to produce text in this course and in previous courses. Also, you need to plan time to search for more information and learn the specifics of your topic.

Take that into account as you formulate your plan for the subsequent work with your thesis.

The time plan should be written as a Gantt chart (maybe useful: PGFgantt LaTeX package) where weekly activities are described, along with a separate, more detailed description of each activity/task.

Milestones

A milestone is a decision point at which some important insight or partial result must be available, and on which further steps might depend.

Based on your current understanding of the problem area that you will do your thesis in, define 2-4 milestones for your project.
Shortly describe each milestone and identify how you will know that it has actually been reached (passed).

Risk Analysis

Risks: There are a lot of potential pitfalls during a 20-week project (e.g., important equipment or test data unavailable, technical issues with your development/deployment environment, increased time to learn techniques or implement solutions, project partner low-performing or leaving before completion etc.), and you should try to imagine what those risks may be and how best to address them. Maybe you need to plan a couple of concurrent activities for each week that you can select from, based on what is possible to work with? Maybe you need to have fallback options available if your main course of action proves more difficult than you imagined?

Based on your current understanding of the problem area that you will do your thesis in, identify 3-4 different risks in your thesis plan that you will need to manage during your work.
Avoid trivial or general risks that might apply to any project, try to be specific w.r.t. your project topic and method.
(Hint: Having another look at the Lecture 5 slide on method-specific threats to validity could be helpful here.)

Assess each identified risk by its likelihood (low, medium, high) and severity (low, medium, high), and suggest possible prevention and mitigation strategies.

Other advice

For implementation-heavy projects, make sure to plan sufficient time for the experimental evaluation, its preparation (e.g., developing performance test programs if not yet available) and analysis (e.g., additional/refined experiments for identifying the root causes of observed performance anomalies or bottlenecks).

Make sure to plan writing on your thesis continuously, and to provide weekly, short written summaries of your work. These will be useful at the very end, when you are required to write a reflection on your thesis work (instructions also available in Swedish).

Submission

Submit your thesis plans to the collaborative workspace (Seminar5/group) by Tuesday before Seminar 5.

Reviews

Review (individually) each other's submissions according to the grading rubric and upload your reviews in the collaborative workspace before the seminar.

When assessing the other teams' time plans, check whether you have similar levels of detail in describing different activities, and compare your estimates of how much time you think that you will need to spend on different activities.

Seminar

During the seminar, you will compare your reviews of each other's thesis plans. We will end with a joint discussion on requirements for the final submission.

Page responsible: Christoph Kessler
Last updated: 2023-12-01