Course Files and Examples, Stat141 '07

Class 9/25
1. Course Introduction and Overview
  Handout 1

2. Opening Day Exemplars:
a. Was Darwin just messin' with us?        Topics: ordered categorical data, regression for explanation?
Public Acceptance of Evolution, Science 11 August 2006 Vol. 313. no. 5788, pp. 765 - 766   Jon D. Miller, Eugenie C. Scott, Shinji Okamoto    supplemental materials

b. Freshman 15. College weight gain: time 1, time 2 data
time1-time 2 measured data, comparing groups, t-tests
College: A weighty endeavor? 'Freshman 15' may be a myth, but many students pack on pounds. San Francisco Chronicle Tuesday, September 26, 2006
Rutgers study, Changes in Body Weight and Fat Mass of Men and Women in the First Year of College: A Study of the "Freshman 15.". By: Hoffman, Daniel J.; Policastro, Peggy; Quick, Virginia; Soo-Kyung Lee. Journal of American College Health, Jul/Aug2006, Vol. 55 Issue 1, p41-45,
accessed through the Lane Library portal, lane.stanford.edu (just go to the full journals listing)

c. Comparing proportions,
Parasitic Weed Uses Chemical Cues to Find Host Plant, Elizabeth Pennisi
    Science 29 September 2006 313: 1867 (in News of the Week)
  Volatile Chemical Cues Guide Host Location and Host Selection by Parasitic Plants Science 29 September 2006
  Supporting Online Material for Volatile Chemical Cues Guide Host Location and Host Selection by Parasitic Plants Justin B. Runyon, Mark C. Mescher, Consuelo M. De Moraes

d. Two-way factorial design, analysis of variance.
Kin recognition in an annual plant Susan A.Dudley and AmandaL.File, Biology Letters, June 2007

e. prediction, regression methods
Bigger is smarter: Overall, not relative, brain size predicts intelligence
Overall Brain Size, and Not Encephalization Quotient, Best Predicts Cognitive Ability across Non-Human Primates
Deaner, R.O. ; Isler, K. ; Burkart, J. ; van Schaik, C. Brain Behav Evol 2007;70:115-124 (DOI:10.1159/000102973)

3. Getting started. Describing Distributions: SW text Chap 2; esp secs 2.3, 2.4, 2.5, 2.6, 2.7
Class handouts:
Brain weight           data from SW text ex 2.12, pp.22-23
MAO                       data from SW text ex 1.4, pp.3-4

Class 9/27
Continue: Describing distributions (handouts and content from 9/25)
SW text Chap 2; esp secs 2.2, 2.3, 2.4, 2.5, 2.6, 2.7. Verzani 2.2, 2.3, 3.2 (comparing groups)
Transformations to change shape (SW 2.7);   
Count and Category Data SW ex1.5, sec 2.1, exs 2.1,2.2,2.5; Verzani 2.1

In the news:
1. Acupuncture May Be More Effective Than Conventional Therapy In Treating Lower Back Pain
  German Acupuncture Trials (GERAC) for Chronic Low Back Pain: Randomized, Multicenter, Blinded, Parallel-Group Trial With 3 Groups
Michael Haake; Hans-Helge Müller; Carmen Schade-Brittinger; Heinz D. Basler; Helmut Schäfer; Christoph Maier; Heinz G. Endres; Hans J. Trampisch; Albrecht Molsberger Arch Intern Med. 2007;167:1892-1898.
2. Spaceflight Can Change Bacteria Into More Infectious Pathogens    Germs Taken to Space Come Back Deadlier

Class handouts and resources:
Cricket Singing Times           data from SW text ex 2.19, pp.29
Female brain weights--illustration of percentiles, stem-and-leaf, mean sd
Count Data--
  Horseshoe Crab Satellites                     data from Agresti
Also in class Prussian Army Horse Kicks.
Extra calcs  cricket dichotomization, moments for grouped crab data

Biological aside: pigeons do logs, so can you.  Pigeon-brained birds can think in logarithms
Evidence that pigeons represent both time and number on a logarithmic scale
Behavioural Processes Volume 72, Issue 3 , 1 June 2006, Pages 207-214
Class 10/2
Samples and populations; Introduction to probability calculations,  SW text secs 2.8-2.9, 3.1-3.5; Verzani 5.1

Class handouts and materials:
p's and q's: recap of percentiles quantiles, with brain data
overheads for populations and samples
Random samples and Probability
false-positive, false negative Medical Diagnosis    see SW text section ex3.17
Class 10/4
Continue basic probability calculations (medical diagnosis)
Discrete Random Variables: Bernoulli Trials and Binomial Distribution; Poisson Distribution and Count Data; SW text 3.7-3.9;  Verzani 5.1.1, 5.2.2

Class handouts and materials:
1. Binomial Distribution Examples and Binomial-Poisson approximation
2. Poisson Distribution for Count Data.   link1  link2   link3
Class Handout is an annotated "link 2" plus Poisson fit to horsekick data
3. Discrete Distribution examples
Class 10/9
Continuous Distributions for measured data (Normal, Uniform, Exponential). SW text 3.6, 4.1-4.3;  Verzani 5.1.2, 5.2.1-5.2.2
Probability calculations with Normal distribution. SW text 4.3.
Binomial approximation, SW text 4.5 and 5.5; Verzani 6.1

Class handouts and materials:
1. Continuous Distributions (including keepsake)
2. normal approximation to discrete distrib, continuity correction
3. sums and differences of random variables, incl normal distribution
Class 10/11
Continue: probability calculations, continuous distributions
Assessing normality, q-q plots. SW text 4.4, Verzani 3.2.4;
Begin: Sampling Distributions of sample means: dichotomous (proportions) and quantitative observations. Central Limit Theorem
SW text 5.1-5.4; Verzani book, 5.1.3, 5.1.4, 5.3, 6.1

Class handouts and materials:
1. Exponential distribution materials: Wolfram Mathworld    Wikipedia
2. Calculation of Moments for Discrete and Continuous Distributions:    Mathematica notebook      pdf of notebook
3. Normal Distribution calculations, Assessing Normality including qq-plots (prior year)
data for qqplot example        fruit moisture, SW ex 4.8
4. Distribution of Sample Means: How do sample means behave? (prior year)
Class 10/16
Continue Central Limit Thm examples and calculations; (handouts from 10/11)
Begin: Inference for a single mean (including proportions): Confidence Intervals and Hypothesis tests.
SW text 6.1-6.3, 6.5,6.7; paired samples (including nonparametric tests) SW Ch.9;
Verzani book, 7.1-7.3, 7.5.3, 7.6 ; 8.1-8.4, 8.6.2

Class materials:
Inference for a single mean (including paired data)         squirrel data, SW ex 9.6
Inference for a single mean: paired data, virus example       mengovirus SW ex 9.10 data
Class 10/18
Continue: Inference for a single mean (including proportions): Confidence Intervals and Hypothesis tests.
inference for single mean: SW text 6.1-6.3, 6.5, 6.7, Verzani, 7.1, 7.3 8.2,8.3; paired samples SW Ch.9, Verzani, 7.5.3; 8.6.2.
proportions: SW 6.6 (ignore "p-tilda"), Verzani 7.2,8.1

Class materials:
Lecture        Additional computing

Class 10/23
Inference for proportions (Verzani 7.2, rather than S.W, 6.6); Binomial test (Verzani 7.2); Wilcoxon-Mann-Whitney test (S.W. 7.11);
Sign test (S.W. 9.4); Wilcoxon Signed Rank test (S.W. 9.5)

Class materials:
Lecture        Additional computing

Class 10/25
Further examples: two sample (treatment/control) comparisons, inference procedures, SW Ch 7, Verzani 7.5.2, 8.6, 9.3.1 (Kolmogorov-Smirnov)
Research Design: Overview, value of experiments, Causal inference, Association vs Causation, SW Ch 8. (esp 8.1-8.3, easy reading)
Research Design: Sample Size Calculations. SW 6.4, 7.8, App7.1.

Class handouts and materials:
Recap: Inference for two-sample comparisons  Two-sample(treatment/control)inference: t-test, nonparametric alternatives (Mann-Whiney, Duckworth)
                ancy data SW ex7.7
Additional data sets:      rats sniffing glue, tuolene data SW ex 7.9-7.11        soil respiration SW ex 7.17

Class handout new materials Pre-post comparative experiment; power and sample size calculation
Research paper for urea synthesis data: Comparative Analyses of Pretest-Posttest Research Designs, Donna R. Brogan; Michael H. Kutner, The American Statistician, Vol. 34, No. 4. (Nov., 1980), pp. 229-232.   JSTOR link
     urea synthesis, BK data     

Short description of Duckworth and history (Dick Deveaux text pp.29-30)
Slides recapping nonparametric methods Courtesy of Department of Environmental and Plant Biology, Ohio University, Athens Ohio
Class Materials:
another nonparametric test: ks.test(): Kolmogorov-Sminorv test (stats)      A general document that includes K-S examples: Fitting distributions with R


Background reading on research design.
1. Classic paper on Medical experimentation. Statistics and Ethics in Surgery and Anesthesia. John P. Gilbert; Bucknam McPeek; Frederick Mosteller Science, New Series, Vol. 198, No. 4318. (Nov. 18, 1977), pp. 684-689.     JTSOR link
2. From Stat209         Cause and Effect Examples:
         a. Prayer, more harm than good?
From Harvard  Study calls prayer for sick people ineffective Rate of problems after surgery unchanged by act   
Study of the Therapeutic Effects of Intercessory Prayer (STEP) in cardiac bypass patients: A multicenter randomized trial of uncertainty and certainty of receiving intercessory prayer American Heart Journal, Volume 151, Issue 4, Pages 934-942 (April 2006). [compliance note]
Prior Duke   flip   flop
         b. Carbonated soft drinks and Cancer    Yale press release      News report: Soda-Cancer Link Revealed as Myth      
Journal of the National Cancer Institute     JNCI paper  Carbonated Soft Drink Consumption and Risk of Esophageal Adenocarcinoma Mayne et al. J Natl Cancer Inst.2006; 98: 72-75
        c. Is TV bad or is it bad parenting? Attention Deficit Disorder and TV
2004 version : Pediatrics. 2004;113:708-713. Christakis DA, Zimmerman FJ, DiGiuseppe DL, McCarty CA. Early television exposure and subsequent attentional problems in children.   Publication   summary    press release    news report      audio NPR interview    interview transcript and publication
2006 reversal? Pediatrics. March 2006. Stevens T and Mulsow M. There is no meaningful relationship between television exposure and symptoms of attention-deficit hyperactivity disorder. Pediatrics. 2006; 117(3):665-672.   News Reports: TV may not cause kids' attention disorders   Researchers say TV is not to blame for ADHD   TV may not cause kids' attention disorders: study.  Good general commentary in Slate Feb '06
      d. Causes? of autism (a student asked about this).
Vaccines and Autism. The CDC statement   a pending court case
Does Television Cause Autism?
A cautionary comment, including my Nobel-laureate Jim Heckman
Citizen and blooger comments: Autism Bulletin     Ariana Huffington       Economists' Full paper: Does Television Cause Autism?
3. Another textbook treatment (Dick DeVeaux)

Power and Sample Size Calculations: Some resources:
JAMA review article: Statistical Power, Sample Size, and Their Reporting in Randomized Controlled Trials (JAMA. 1994;272:122-124)
Resources on power calculations:     Graphpad            R-helpfile: Power calculations for one and two sample t tests          R Tutorial: Calculating The Power Of A Test

Class 10/30
Comparing Many Means: Introduction to Analysis of Variance, single classification. SW Ch. 11.1-11.5, 11.8 . Verzani Ch 11.
One way anova and post-hoc multiple pairwise comparisons, SW 11.4,11.8, Verzani

Research publication examples from lecture:
1. fat drunk mice. Big doses of red wine extract help obese mice stay fat and happy 11/1/2006 USA TODAY
Resveratrol improves health and survival of mice on a high-calorie diet Nature, 16 November 2006.
2. Cannabinoids promote embryonic and adult hippocampus neurogenesis and produce anxiolytic- and antidepressant-like effects   J. Clin. Invest., Nov 2005; 115: 3104 - 3116.

Class handouts and materials:
Slides (scanned overheads) from lecture:     anova intro    post-hoc pairwise comparisons
One-way anova examples
Class handout, One-way anova examples     
Data for Oct 30    Oberlin dance data     Full Moon mental health                ancy data SW ex7.7

squid respiration case study    Squid respiration data
Class 11/1
Continue: Comparing Many Means: Introduction to Analysis of Variance, single classification. and post-hoc multiple pairwise comparisons; SW Ch. 11; Verzani Ch 11.
Power and sample size calculations
Nonparametric alternatives, Verzani 11.1.4
Begin: Randomized Blocks and Factorial Designs. SW 8.4, 11.6, Verzani 11.4

Class handouts and materials:
      power calculations
     nonparametric alternative, Kruskal-Wallis     Clergy mental health data
factorial design analysis of fullmoon data
randomized blocks, factorial designs
Randomized blocks ex: alfalfa data       factorial design ex soybean data      interaction plot for soybean


Power and sample size resources:
JAMA review article: Statistical Power, Sample Size, and Their Reporting in Randomized Controlled Trials (JAMA. 1994;272:122-124)
Resources on power calculations:    
Graphpad            R-helpfile: Power calculations for balanced one-way analysis of variance tests         Lesson 21: Power and Sample Size in ANOVA courtesy of West Point

Class 11/6
Factorial Designs. SW 8.4, 11.6, Verzani 11.4
Begin Inference for categorical data: Introduction to analysis of categorical data, 1 X c tables (univariate) SW text Ch. 10, Verzani 9.1


Class handouts and materials:
randomized blocks, factorial designs
      factorial design ex soybean data      interaction plot for soybean
Introduction to analysis of categorical data    Slides (scanned overheads) from lecture:     polychotomous data, 1xC table example   

Readings on Phillips and death dates:
Can the Famous Really Postpone Death?           JAMA, 1990 Postponement of death until symbolically meaningful occasions

Class 11/8
Analysis of categorical data, 1 X c tables (univariate) ; r X c tables (bivariate); chi-square tests, Fisher's exact test.
SW text Ch. 10, sections 10.1-10.7, 10.10; Verzani 9.1,9.2

Class materials:
Lecture        Additional computing

        Comparing proportions (2x3 table) ex: villi data

Class 11/13
Analysis of categorical data continued.  SW text 10.8, 10.9.
Odds ratios and relative risks;
Paired data (McNemar);
Bayes Thm calculations;

Class materials:
Lecture        Additional computing
The vcd Package Visualizing Categorical Data. October 16,2007 Version 1.0-6
Class 11/15
Categorical Data continued.  SW text 10.8, 10.9.
Recap Odds ratios and relative risks;
Simpson's paradox and higher dimensional tables.
Begin: Association among measured variables
Introduction to correlation (for measuring associations) and regression (for fitting, and prediction).     SW text 12.1-12.6, Verzani Ch.3, Ch.10

Class materials:
Odds ratios and relative risks; Simpson's paradox and higher dimensional tables
Introduction to correlation and regression     figures for correlation examples
mammals brain weight and body weight example
data sets:  platelet.dat SW p.550
Class 11/27
Continue, Association among measured variables
correlation (for measuring associations) and straight-line regression for fitting and prediction
SW text 12.1-12.6, Verzani Ch.3, Ch.10

Class materials:
Introduction to regression models, SW rat example         plots for ratreg examples
data sets:        ratspeed.dat  amphetamine data SW ex 12.1
rank correlation snippet
regressions for mammals brain weight and body weight example        4795 known species of land mammals.
Class 11/29
Fitting curves: Transformations vs polynomial regression, SW text sec 12.6, Verzani 10.3
Dichotomous outcomes, logistic regression.  SW text, sec 12.7, Verzani 12.1

Class materials:
Data transformations and fitting polynomials.    plots for tumor data ex
   tumor.dat Tumor growth data

Inference for regression models, Logistic regression, donner party and esophogeal cancer exs
   plot for donner logistic model fit    plots for Esophageal cancer data ex
data sets:  donner.dat Donner Party Survival by Gender and Age        
   lymph.dat Esophageal cancer data , SW ex 12.44, table 12.9 p.583.
Class 12/4
Dead week: Regression recap   lecture notes 12/4
return TH2 papers
Class 12/6
Dead week: Course overview, Discussion of grading

  12/6 Lecture notes
Review and summary materials, Tables of tests and R commands: (updated from 11/13 handout)
Discussion of Exam3 (Dec12)