TC3 → Stan Brown → Statistics → Spr12 ME50

# MATH200/ME50, Statistics

Spring 2012, T 6:30–9:20, room 224 (Stan Brown)

## News

Latest email to the class: 22 May. Here are all broadcast emails.

Grades have been posted and you may access them through myTC3/myInfo.

If you would like the details of how your grade was computed, you can request and receive that information via TC3 email.

Here is all news for the semester so far.

## Handouts

The handouts are marked with helpful symbols. A full explanation of the symbols is available, and most browsers will also show a “tool tip” if you just mouse over the symbol.

### Chapter 1: Data Collection

#### Optional extras

• Excel and your calculator generate pseudo-random numbers, but Random.org has lots of true random number generators, including numbers, coin flips, and dice rolls.
• In class we mentioned the 1936 fiasco of a presidential election poll. Read about it at Classic Polling Surprises (accessed 2012-03-25) and Introduction to Polling (accessed 2012-03-25).

The original Literary Digest article can be found at Landon in a Landslide: The Poll that Changed Polling (accessed 2012-03-25).

• In class we mentioned that Fox News viewers know less about the world than people who watch no news at all. Can you conclude that this is because they watch Fox News? Why or why not?

The relevant studies are Misinformation and the 2010 Election (accessed 2012-03-25) at the University of Maryland, and Some News Leaves People Knowing Less at Fairleigh Dickinson University (accessed 2012-03-25).

• Alternating Treatments (accessed 2012-03-25) gives lots of examples why you need to randomize your samples.
• What’s wrong with surveys where respondents select themselves? See Web Polls (accessed 2012-03-25).

### Chapter 2: Graphical Summaries

#### Optional extras

• The Joy of Stats video: Hans Rosling shows lots of great ways to present data (accessed 2012-03-25). Don’t miss the segment “200 Countries, 200 Years, 4 Minutes”, also available separately (accessed 2012-03-25).
• just for fun: self-referential graphs
• The Best Stats You’ve Ever Seen video: Hans Rosling uses descriptive stats to explode the myth of First World and Third World (accessed 2012-03-25).
• MATH200A Program part 1 can be used to make histograms, and MATH200B Program part 2 graphs time series.
• ticalc.org has five programs to make pie charts. Though I haven't evaluated them in depth, PIEGRAPH.ZIP looks most interesting based on the screen shots.
• STEMLEAF.ZIP (1 KB) is a TI-83/84 program that makes stemplots
• Charles Minard’s famous graph of Napoleon’s Moscow campaign (accessed 2012-03-25)
This is one of the many examples in Edward Tufte’s The Visual Display of Quantitative Information.

### Chapter 4: Correlation & Regression

#### Optional extras

• Cartoons about correlation and causation are here (accessed 2009-12-28) and here (accessed 2012-03-25).
• If correlation is not Causation (accessed 2012-03-25), how do we know that smoking causes cancer?
• What does R-squared mean? (accessed 2012-03-25)
• Least Squares — the Gory Details — Where do those formulas come from?
• MATH200B Program part 6 for inferences about linear correlation
• Under Chapter 12 are several more inferences you can make about correlation and regression.

### Chapter 8: Sampling Distributions and Sample Variability

#### Optional extras

• The Behavior of the Sample Mean (accessed 2012-03-25) — pictures show how sample means are distributed and why that distribution is different from the shape of the population.

### Chapters 9–12: Inference

• These two handouts apply to Chapters 9 through 12, and they help you choose the right case for analyzing a given situation:
• Hypothesis testing a/k/a significance testing has a language and philosophy of its own, but once you get it you’ll find it’s the same for Chapters 10 through 12. The following handouts replace sections 10.1 and 10.2 of your textbook, and they're included in Hypothesis Tests listed under Chapter 10 below:
• Look at this one after you’ve spent a week or two on hypothesis tests: Top 10 Mistakes of Hypothesis Tests

### Chapter 10: Hypothesis Tests for One Pop.

#### Optional extras

• The logic of hypothesis testing can seem very strange at first. One way to demystify it is to read many different writeups, not just your textbook. Here are two easy and excellent ones: Is Statistics Hard? by Gerald Dallal (accessed 2012-03-25) and Logic of Hypothesis Testing, part of HyperStat (accessed 2012-03-25).
• When α is the standard 0.05, one in 20 of your significant results are actually Type I errors. How do you interpret published results, then? See Sifting the evidence—what‘s wrong with significance tests? by Sterne and Smith (accessed 2012-03-25).
• Even professional researchers can make mistakes and publish false results through misusing hypothesis tests. “Data mining” (first gathering data, then looking for relationships) is one problem but not the only one. See Why Most Published Research Findings Are False by John Ioannidis (accessed 2012-03-25). If you find the article heavy going, just scroll down to read the example in Box 1 and then the corollaries that follow.
• You can also do hypothesis tests about σ, the population s.d.: see MATH200B Program part 5 or Inferences about One Population Standard Deviation.

### Chapter 11: Inferences on Two Samples

#### Extra cases

These are not part of the course syllablus, but are included for those who would like to learn more.