## Stat 431: Statistical Analysis

### Fall 2013, University of Idaho, Section 01, CRN 36142

Instructor: Brian Dennis
Professor, Department of Statistical Science
316 Phinney, 208-885-7423, brian@uidaho.edu

Office hours: 1:30-3:00PM  Mon, Tues, Wed, Thur

Meeting times & places:
Lecture: MWF 9:30-10:20AM TLC 046 (not 45;  course was moved to a bigger room)

Textbook:
Ott, R.L. & Longnecker, M. 2010. An Introduction to Statistical Methods and Data Analysis, Sixth Edition. Brooks/Cole, Belmont, CA.

Other materials:
You should bring a scientific calculator (w/ memory, square root, logarithms) to tests.

Other resources:

Stat Assistance Center (SAC): first floor Library, 885-2929, for help with this course

Statistics Consulting Center (SCC):  Call 885-2929 for an appointment, for statistics advice concerning research or thesis

Based on 600 possible points, 90% A, 80% B, and so on

• Three tests (open book), 100 points each
• One partly comprehensive final (open book), 100 points
• Four take-home computer-aided assignments, 50 points each

Economy data

county demographics data

• Exercises assigned, but not graded (selected exercises may appear on tests)

Makeup tests: Makeup tests will not be given, except for legitimate scheduled university activities or grave, urgent reasons. Prior arrangements with instructor are required.

Final exam rescheduling: By Department of Statistical Science policy, no student may reschedule the final exam without permission of the Department Chair.

Outcomes:  Students successfully completing this course will be able to conduct and interpret standard statistical analyses used in research, including multiple regression, analysis of variance, analysis of covariance, analysis of categorical and other non-normal data, as well as to plan sample sizes and allocations for basic designs and surveys.

SAS documentation:

Introduction to SAS (dated, but informative for ordinary uses of SAS)

SAS online documentation

All the SAS examples you could possibly want

UCLA SAS starter kit

Introduction and documentation for PROC MIXED (pdf of now-defunct but informative website at Univ Kentucky)

The amazing, free R language for statistics, graphics, and computing

SAS examples for class:                                                                                R examples:

two-sample t-test example                                                                        two-sample t-test example

generalized linear models examples:

random and mixed effects examples:

MINITAB documentation:

Course evaluations

Aug 26

introduction

Aug 28          Ch 4
probability

distributions

Aug 30
probability

distributions

Sep 2

no class:

Labor Day

university closed

Sep 4
probability distributions, continued

Sep 6

Getting started with SAS

Sep 9

Getting started with R

Sep 11

sampling distributions

Sep 13          Ch 5

inferences for central values

Sep 16

inferences for central values, continued

Sep 18          Ch 6

inferences for comparing two central values

Sep 20

comparing two central values,

continued

Sep 23          Ch 7

variances

Sep 25          Ch 8

1-way AOV

Sep 27

1-way AOV

Sep 30

Test 1

Oct 2          Ch 9

multiple comparisons

Oct 4

multiple comparisons

Oct 7

inferences for non-normal models (& corresponding sections of Chapter 8)

Oct 9          Ch 10

Oct 11

Oct 14

categorical data

Oct 16

# Oct 18           Ch 11linear regression

Oct 21

linear regression

Oct 23

Oct 25

Test 2

Oct 28          Ch 12

multiple regression

Oct 30
multiple regression

Nov 1          Ch 13

model selection

Nov 4

nonlinear regression

Nov 6

Nov 8          Ch 14

Nov 11

Nov 13          Ch 15

block design

Nov 15

design

Nov 18

Test 3

Nov 20

AOV unbalanced & 3-way

Nov 22          Ch 16

Fall Break!

 Dec 2 AOV & regression Dec 4 Dec 6 generalized linear models (GLIM) Dec 9 mixed effects models Dec 11 mixed effects models Dec 13 other topics Final exam occurs in the scheduled final exam period, 10AM-12noon, Tuesday, Dec 17.  (No earlier reschedules will be allowed).

Items for further interest and enjoyment

Dallal’s essay on why significance level is 5%

Brian’s essay on Bayesian statistics

Aaron and Brian’s essay on statistics education for ecologists

Scientists on medical & health practices  (If it talks like a quack, and bills like a quack, it probably is a quack):

Paranormal/supernatural phenomena:

The Last Will and Testament of Philip J. Klass

To UFOlogists who publicly criticize me…or who even think unkind thoughts about me in private,

I do hereby leave and bequeath THE UFO CURSE: No matter how long you live, you will never know

any more about UFOs than you know today. You will never know any more about what UFOs really

are, or where they come from. You will never know any more about what the U.S. Government really

knows about UFOs than you know today. As you lie on your own death-bed you will be as mystified

about UFOs as you are today. And you will remember this curse.

— Philip J. Klass, prominent UFO skeptic, posted on a UFO discussion site shortly before his death

Scientists on evolution, intelligent design, biblical creation:

Dover trial transcripts  (read what scientists and “cdesign proponentists” say under oath)

The continuing futile attacks by evolution’s opponents reminds me of another legendary confrontation,

that between Arthur and the Black Knight in the movie Monte Python and the Holy Grail.  The Black

Knight, like evolution’s challengers, continues to fight even as each of his limbs is hacked off, one by

one.  The “no transitional fossils” argument and the “designed genes” model have been cut clean off,

the courts have debunked the “ID is science” claim, and the nonsense here about the edge of evolution

is quickly sliced to pieces by well-established biochemistry.  The knights of ID may profess these blows

are “but a scratch” or “just a flesh wound,” but the argument for design has no scientific leg to stand on.

Sean B. Carroll, in a review (2007 Science 316:1427-1428), of Michael J. Behe’s book

The Edge of Evolution:  The Search for the Limits of Darwinism.

Statisticians analyze USA elections:

Exit poll discrepancy analyses (2004 USA presidential election)

(How the 2004 election fraud was committed:  A.  B.)

Scientists on climate change research:

LaTeX (science/math/tech typesetting) resources:

Comparing public, private, and charter schools with NAEP data:

Postmodern comedy:  which of the following articles is a (deliberate) parody?

And of course, no statistics course website would be complete without: