
![]() |
This lecture discusses some common ways that people use statistics to draw invalid or misleading conclusions. Image courtesy of spweber on Flickr. |
Lecture 24: Avoiding Statistical Fallacies
About this Video
Topics covered: Statistics, plotting, correlation, causation, bias, logical fallacies, data enhancement, Texas sharpshooter fallacy.
Resources
Recitation 10: Introduction to Dynamic Programming
About this Video
Topics covered: Dynamic programming, memoization, overlapping subproblems, optional substructure, Fibonacci memoization example.
What does GIGO stand for?
› View/hide answer
Garbage In Garbage Out: if your data is flawed, any conclusions you draw from it will be, too.
What does 'cum hoc ergo propter hoc' mean?
› View/hide answer
It is Latin for "with this, therefore because of this," a logical fallacy which involves incorrectly inferring a causation from a correlation. You should always remember—correlation does not imply causation.
What is a lurking variable?
› View/hide answer
One which doesn't appear in the plot of two other variables, but could have an effect on both. It is also sometimes referred to as a confounding variable.
In this problem set you will write a solution to an optimization problem on how to find the shortest route from one building to another on the MIT campus given that you wish to constrain the amount of time you will spend walking outdoors (because generally speaking, the nocturnal beaver… err, um, the nocturnal MIT engineer… hates the sun).
These optional resources are provided for students that wish to explore this topic more fully.