Teaching
Courses I design and teach, with the slides and class notes open to everyone.
Through Stat.BD, which I founded, I teach data analysis and research methodology to hundreds of learners, from university students to clinicians and other professionals. Many of them are doing research of their own, and I guide them through it, from study design to analysis.
Two courses have open materials here: Applied Biostatistics & Research Methodology for Clinical Research, for clinicians, and Statistics & Data Analysis Fundamentals, for university students.
Stat.BD · 1:1 course · 5 of 30 sessions released
Applied Biostatistics & Research Methodology for Clinical Research
A 1:1 course I designed for a practising clinician with ten years of clinical experience who is moving into research. I teach it through Stat.BD, which I founded. The plan is 30 sessions over 15 weeks, with 14 SPSS labs, and sessions are published here as they are taught.
Handed a medical paper you have never seen, in a field that is not yours, you can say out loud and in front of people what its question was, whether its design could answer it, what its numbers mean, where it is vulnerable, and whether you believe it.
How I teach it
-
Real papers from day one
Two published South Asian studies run through the whole course, read in full rather than as simplified extracts.
-
Build every number yourself
Each summary a paper prints is rebuilt by hand in Excel before any software is used. The SPSS lab always follows the concept and never leads it.
-
Start from what the learner knows
Clinical experience is treated as the one thing in the room that cannot be taught. Analogies come from the ward, not the textbook.
Syllabus
Four phases, each building the language the next one needs. Every released session has its projected slides and the student class notes; some add Excel practice workbooks.
Phase I · Statistical foundations
Sessions 1-8
The language, summaries, distributions, sampling variability, confidence intervals, p values.
-
Session 1
The Story of Research
Where a research question comes from, and what statistics is actually for
-
Session 2
Population, Samples, Variables & Data
What one row is, what one column is, and what job each column does
-
Session 3
Describing Categorical Data
Counts, and the four things people do to them
-
Session 3.5
Working with Data in Excel
Every number you have been shown, made by you instead
Excel workbooks: Practice register · Khanam tables · COBRA-BPS · Age bands · Worked solution
-
Session 4
Describing Numerical Data
Where the middle is, how spread out it is, and which pair of numbers to print
Excel workbooks: Practice data
-
Session 5
Understanding Distributions & Outliers
What a column looks like, and the values out in its tails
- Sessions 6-8: released as they are taught.
Phase II · Research methodology
Sessions 9-18
Reading papers, questions, study designs, measurement, bias, sampling, sample size, protocol and ethics.
- Sessions 9-18: released as they are taught.
Phase III · Applied biostatistics
Sessions 19-26
Choosing a method, group comparisons, effect measures, regression, multivariable thinking.
- Sessions 19-26: released as they are taught.
Phase IV · Research literacy & communication
Sessions 27-30
Diagnostic and prediction studies, survival, systematic reviews, writing the manuscript.
- Sessions 27-30: released as they are taught.
Anchor papers, read in full on the publishers' sites: Jafar et al., NEJM 2020 (COBRA-BPS) · Khanam et al., BMJ Open 2019
Slides, class notes and workbooks © Muhtasim Munif Fahim / Stat.BD, shared under CC BY-NC 4.0. You are welcome to use them for non-commercial teaching with attribution.
Stat.BD · classroom course · 6 theory sessions released
Statistics & Data Analysis Fundamentals
A classroom course for university students from any department and any year, with no statistics or programming assumed. The plan is 14 two-hour sessions over two weeks: six theory sessions, six tool sessions in SPSS, R and Python, then a project lab and a capstone. I teach the six theory sessions, and their slides and class notes are below.
Run and report a real analysis, start to finish, on data you helped create: import it, clean it, describe it, test it, and say what the result does and does not show. Then choose between SPSS, R and Python knowing what each is for.
-
The class is the dataset
The room fills in an anonymous survey in Session 1 (sleep, screen time, study hours, CGPA band, a stress scale), and every later example is about people in that room.
-
One real paper, read critically
A published survey of Bangladeshi university students runs through every session, including its misprints and the places where its tables disagree with each other.
-
Statistics once, then three tools
The ideas are taught without software first. The same analysis is then run in SPSS, R and Python, so the choice of tool comes after the understanding.
Theory
Sessions 1-6
Research questions and data, description, sampling and intervals, tests, relationships, regression and survey scales.
-
Session 1
Data and Research Basics
From a question about your own life to a dataset you can analyse
-
Session 2
Describing Data
Tables, averages, spread and charts, before any test
-
Session 3
From Sample to Population
Probability, the normal curve, and how far an estimate can be trusted
-
Session 4
Testing Hypotheses
Comparing groups, and when a difference is more than luck
-
Session 5
Relationships between Variables
Chi-square, correlation, and why association is not causation
-
Session 6
Regression and Survey Scales
Holding things fixed, yes or no outcomes, and whether a scale hangs together
Tools
Sessions 7-12
The same class dataset, from raw file to reported result, in SPSS, R and Python.
Practice
Sessions 13-14
A project lab on a question of the group’s own, and a capstone that ends with choosing a tool.
Running paper, read in full: Islam et al., PLoS One 2020 . The class dataset in these materials is simulated until the real class survey is collected, and every page that uses it says so.
Slides and class notes © Muhtasim Munif Fahim / Stat.BD, shared under CC BY-NC 4.0.
Mentoring
Outside the course, I informally advise students on ML and statistics projects. I can help with:
- Reading guidance: what to read next (ML/NLP/NAS/causal inference) and how to take notes effectively.
- Project feedback: clarifying problem statements, evaluation plans, and research communication.
- Reproducibility: setting up experiments, baselines, and reporting results.
Email me with a short description of your background, what you're trying to do, and a concrete question. If you're asking for feedback on a project/paper, include a link and what kind of feedback you want.