Meaning from Data: Statistics Made Clear

Who was the greatest baseball hitter of all time? How likely is it that a poll is correct? Is it smart to buy last year's highest-performing stock? Which hospital has the best outcome for a given procedure? When is it a good idea to buy a product's extended warranty? These questions all involve the interpretation of statistics, as do a surprising number of other mysteries, including: Is the "hot hand" among sports players real? How can you tell if Shakespeare is the probable author of a newly discovered poem? What is a guilt-free way to get someone to admit to cheating? And, how does a tax assessor calculate the market value of a house?

Meaning from Data: Statistics Made Clear is your introduction to a vitally important subject in today's data-driven society. In 24 half-hour lectures, you will explore the principles and methods that underlie the study of statistics. You have probably heard such terms as mean, median, percentile, quartile, statistically significant, and bell curve, and you may have a rough idea of what they mean. This course sharpens your understanding of these and scores of other statistical concepts and shows how, properly used, they can extract meaning from data.

Become Statistically Savvy

These challenging yet accessible lectures assume no background in mathematics beyond basic algebra. While most introductory college statistics courses stress technical problem solving and plugging data into formulae, this course focuses on the logical foundations and underlying strategies of statistical reasoning, illustrated with plenty of examples. Professor Michael Starbird walks you through the most important equations, but his emphasis is on the role of statistics in daily life, giving you a broad overview of how statistical tools are employed in risk assessment, college admissions, drug testing, fraud investigation, and a host of other applications.

Statistical Adventures

Professor Starbird is a master at conveying concepts through examples. Some of these include:

When is a Lottery not a Lottery? When it is not truly random. The 1969 Vietnam War draft lottery assigned young draft-age men a ranking for induction based on their birthdays, which were placed in capsules and drawn from a container, supposedly at random. But by computing the statistical correlation for the order-of-draw, it's clear that a nonrandom variable was at play. The most likely explanation is that the capsules with the dates were not thoroughly mixed.
The Birthday Challenge: What is the probability that out of 50 random people, two of them share the same birthday? The chances are much higher than most people think.
The Chicken Soup Method: How can 1,000 randomly chosen people serve as a predictor for the behavior of hundreds of millions of voters? This is the essence of a political poll, and its effectiveness should be no more surprising than the fact that that a single taste of chicken soup is enough to predict the overall saltiness of the batch, whether the batch is in a cup or a giant vat.
Beware of Fallacious Reasoning: At the O. J. Simpson murder trial, Simpson's lawyer Johnnie Cochran countered evidence that Simpson had beat his wife with a statistic that only 1 in 1,000 wife beaters go on to kill their wives. Therefore, Cochran argued, there was only a 1 in 1,000 chance that Simpson went on to commit the murder. Professor Starbird discusses the fallacies in this argument, including the fact that a wife was actually murdered in this case, so the relevant question should be: What is the probability that she had previously been beaten?
Who Really Won the 1860 Presidential Election? Establishing the will of the people in an election can be tricky, especially when three or more candidates are involved. Professor Starbird shows how the results of the four-way presidential race of 1860 can be interpreted as giving victory to each of three candidates, depending on the voting scheme employed. Abraham Lincoln won according to the rules in place, but given other equally valid rules, the victor—and history—would have been very different.

Statistics Is Everywhere

Statistical information is truly everywhere. You can't look at a newspaper without seeing statistics on virtually every page. You can't talk about the weather forecast without invoking statistics. Statistics obviously arises in business and social science but even enters the arts in analyzing manuscripts. And you'd better not go to a casino without understanding statistics. "It's really harder to find somewhere where statistics isn't important than to find the places where it is," says Professor Starbird.

Lectures:

1. Describing Data and Inferring Meaning
2. Data and Distributions—Getting the Picture
3. Inference—How Close? How Confident?
4. Describing Dispersion or Measuring Spread
5. Models of Distributions—Shapely Families
6. The Bell Curve
7. Correlation and Regression—Moving Together
8. Probability—Workhorse for Inference
9. Samples—The Few, The Chosen
10. Hypothesis Testing—Innocent Until
11. Confidence Intervals—How Close? How Sure?
12. Design of Experiments—Thinking Ahead
13. Law—You’re the Jury
14. Democracy and Arrow’s Impossibility Theorem
15. Election Problems and Engine Failure
16. Sports—Who’s Best of All Time?
17. Risk—War and Insurance
18. Real Estate—Accounting for Value
19. Misleading, Distorting, and Lying
20. Social Science—Parsing Personalities
21. Quack Medicine, Good Hospitals, and Dieting
22. Economics—“One” Way to Find Fraud
23. Science—Mendel’s Too-Good Peas
24. Statistics Everywhere