Spotting Bad Statistics and Misleading Charts
Welcome, champion! 🔥
You've mastered 0 of 45 lessons. Start your first lesson free — no signup needed.
About this course
A startup says its average salary is ₹15 lakh a year. That is true, and yet nine of its ten people earn ₹6 lakh, because the founder takes ₹96 lakh. Numbers like this are rarely invented; they mislead through what they choose to show. This online course on Abekus teaches you to spot the common tricks in headlines, pitch decks, opinion polls and WhatsApp forwards, in short reading-card lessons each followed by an MCQ quiz, with every example worked in real numbers.
Who this course is for
It is for curious students and working adults who read numbers every day and want to know when to trust them. School arithmetic is enough; there are no formulas for hypothesis tests, p-values or regression. Fictional recurring examples keep the arithmetic clear: Zipkart, a ten-person Bengaluru startup; a home-loan rate that rises from 8% to 10%; a snack that raises a disease risk from 2 in 1,000 to 3 in 1,000; and two monthly sales bars of ₹50 lakh and ₹55 lakh drawn on an axis starting at ₹48 lakh. Real cases include the 1973 Berkeley admissions data, the 1936 Literary Digest poll, wartime analysis of damage on returning aircraft, Anscombe's quartet and the correlation between chocolate eaten and Nobel prizes won.
Health examples are used only to show risk arithmetic, not as diet or medical advice, and the course stays away from current political and economic figures.
How the lessons work
Each lesson is a short set of reading cards on one trick or idea, followed by a quiz of multiple-choice questions. The course starts with the average salary nobody earns and the two questions to ask of any number: out of how many, and compared with what. It then covers averages and spread, percentage traps and risk numbers, correlation and causation including Simpson's paradox, biased samples and sample size, and misleading charts. Charts are described in words with their axis values, so you learn to redraw them in your head. You go at your own pace.
MCQ practice vs video courses for statistics
Video courses on Udemy, Coursera or Khan Academy are a good choice if you want a full statistics course with formulas and software. This course is built for a narrower job: reading a claim and deciding whether it holds up. Each short explanation is followed straight away by questions that put a real-looking claim in front of you, such as a rate that "rose 2%", a risk that is "50% higher" or a bar chart whose axis does not start at zero, and ask what it actually says. Practising on claims is what builds the habit of checking them. Many learners use both formats.
Common statistics mistakes
The course corrects these directly:
- A rate that rises from 8% to 10% rose 2%. It rose 2 percentage points, which is a 25% relative rise.
- A 50% higher risk means half of people are affected. In the snack example it means 3 cases in 1,000 instead of 2.
- The average salary is what a typical employee earns. When a few values are very large, the median is the typical value.
- A bigger sample always fixes bias. The Literary Digest received about 2.4 million replies in 1936 and still predicted the wrong winner.
- A fall of 50% followed by a rise of 50% gets you back to the start. ₹100 falls to ₹50 and then rises only to ₹75.
- The 1973 Berkeley figures proved discrimination against women. Department by department, there was no bias against them.
- The story of the returning aircraft is fully documented. The principle is; many popular details were added later.
Best way to spot bad statistics
Ask the same few questions every time: which average is this, out of how many, compared with what, who was counted and who was left out, and where does the chart's axis start. The last part of the course practises exactly this. It compares survivorship bias with sampling bias and confounding with reverse causation, shows which average to report and when a line or a bar chart suits a claim, and applies the checks to a startup growth claim, election opinion poll results, app star ratings and a health headline about a food.