Ph.D. coursework

Research Methodology

Batch Aug – Dec 2026

Fill book and research paper details Check filled details

Check your enrolment number

If you did not fill in your enrolment number, please fill the form again.

Presentation submission

Upload your presentation for the Research Methodology course:

Upload PPT or PDF

Guidelines

  • The presentation must not exceed 5 minutes.
  • Accepted formats: .pptx or .pdf only.
  • No external or embedded links.
  • Name the file EnrolmentNo_Name.

Submissions that do not follow these rules may be rejected.

Submitting your assignments

  • You may use any AI tools to assist your work. You remain responsible for the accuracy, integrity and formatting of what you submit.
  • Submit everything in a two-hole binder file. Last date: 27 April 2026.
  • Include a table of contents at the front, with assignment numbers and page numbers.
  • Put your enrolment number and name on the cover page.

Assignments

Eight in total.

Assignment 1

Suggest research topics on these subjects

  • Animal rights
  • Education
  • English
  • Gambling
  • Health
  • Voting
Assignment 2

Annual Status of Education Report 2024

Read the Annual Status of Education Report 2024 and understand its context, method and results. Then answer, in a structured report:

  • What was the need for this survey?
  • How many people were surveyed, and what was their composition?
  • What were the key findings?

Format: about 3 pages (600 words), clear language, citations where needed, answers under clear headings.

Deadline: same as Assignment 1.

Assignment 3 · classwork

The reaction-time experiment — a live t-test

We will run a two-sample t-test on data you generate in class.

  1. The class splits into two groups at random (odd vs even roll numbers).
  2. Each student measures one reaction time (in ms) at humanbenchmark.com/tests/reactiontime and reports it to her group.
  3. Each group computes its mean and standard deviation on the board.
  4. Together we test H₀: the two groups have equal mean reaction time using the pooled two-sample t-test from the lecture: pooled variance, standard error, t, degrees of freedom, critical value at α = 0.05, conclusion.
  5. We then repeat the test with a different split (front vs back benches, spectacles vs no spectacles, …).

Because the first split was random, H₀ is true by construction — so if any split comes out “significant”, you have just witnessed a Type I error live. Think about it: if we tried 20 different splits, how many “significant” results should we expect by pure chance?

Submit: one page in your binder — the data table for both groups, the full calculation, and your conclusion in one sentence.

Assignment 4 · homework

The error hunter — Type I and Type II in the wild

Pick three real systems that make accept/reject decisions. The first one must be the IGDTUW biometric attendance machine; choose the other two from: an email spam filter, a bank’s OTP fraud check, a smoke detector, a metro baggage scanner — or any system from your own branch.

For each system, write exactly four lines:

  1. H₀, stated explicitly. (Everything depends on this line — most wrongly-labelled errors trace back to a sloppy H₀.)
  2. The Type I error in that system, in plain words.
  3. The Type II error, in plain words.
  4. Which error is costlier for that application, and therefore which way the decision threshold should lean.

For the biometric machine, your answer should connect Type I/Type II to FRR and FAR as done in the lecture. One page total. Deadline: next class.

Lecture slides

Worth reading

All course material