Stat 425
Introduction to Nonparametric Statistics


Instructor:
  • Fritz Scholz
  • Office: Padelford C-310
  • Office Hours: Tu 1:00pm-2:00pm, Th 3:00pm-4:00pm or by appointment.
  • Office Phone: 206-543-3866
  • fscholz at u dot washington dot edu (best for messages)


    Initial Course Announcement

    Lectures: Tu/Th 9:00-10:20 Padelford C-301
    Text (required):
    Erich L. Lehmann, NONPARAMETRICS: Statistical Methods Based on Ranks (Springer Verlag 2006, Paperback).
    Text (optional):
    W. John Braun and Duncan J. Murdoch, A First Course in STATISTICAL PROGRAMMING WITH R.

    See also the many free introductory guides at the bottom of this page.


    Class Notes and Other Background Material (under construction):

    Technical Report: The Bootstrap Small Sample Properties.
    Fritz Scholz, March 1993.



    Lecture Slides (under construction):

    Introduction/Review of R (not formally covered in class), last updated 3/22/09.

    Rank Tests for Comparing Two Treatments, last updated 5/16/09.

    Two Sample Population Model, changed slide 49, inserted slides 73 and 74 on 4/27/09.
    Replaced slides 115-116, 120-123, 128 by new slides 115-117, 121-122 on 5/4/09.

    Blocked Comparisons for Two Treatments, updated 5/6/09, corrected slides 56, 65, 66, 69 on 5/12/09. Updated 5/19/09, corrected slide 87. Last updated 5/20/09, slides 64, 66, 78, 79, 80, 81, 82, 86

    Paired Comparison in a Population Model and the One-Sample Problem, updated 5/25/09, slides 41, 50-54, 60 and 66. Updated 5/26/09, slides 30, 40, and 48. Last updated 6/2/09, slide 91.

    The Comparison of More Than Two Treatments, last updated 5/23/09.

    Grades: Homework 100%.
    Assignments follow below (under construction).

    Homework 1: Chapter 1, problems 8, 15, 22, and 25. Due Thursday, April 9, 2009, by class time.

    Homework 1 Solutions

    Homework 2 Due Thursday, 3pm, April 16, 2009.

    Homework 2 Solutions

    Homework 3 Due Thursday, 3pm, April 23, 2009.

    Homework 3 Solutions

    Homework 4 Due Thursday, 3pm, April 30, 2009.

    Homework 4 Solutions

    Homework 5 Due Thursday, 3pm, May 7, 2009.

    Homework 5 Solutions

    Homework 6 Due Thursday, 3pm, May 15, 2009.

    Homework 6 Solutions

    Homework 7 Due Thursday, 3pm, May 22, 2009.

    Homework 7 Solutions

    Homework 8 Due Friday, 3pm, May 29, 2009.

    Homework 8 Solutions

    Homework 9 Due Friday, 3pm, June 5, 2009.

    Homework 9 Solutions



    Submit them electronically or on paper, written legibly!
    Approximate Schedule:

    Week   1 : Rank Tests for Comparing Two Treatments
                       
    Week   2 : Comparing Two Treatments or Attributes in a Population Model
                       
    Week   3 : Blocked Comparisons for Two Treatments
                       
    Week   4 : Paired Comparisons in a Population Model and the One-Sample Problem
                       
    Week   5 : The Comparison of More Than Two Treatments
                       
    Week   6 : Randomized Complete Blocks
                       
    Week   7 : Tests of Randomness and Independence
                       
    Week   8 : Nonparametric Maximum Likelihood Estimation
                       
    Week   9 : Buffer
                       
    Week 10 : The Bootstrap Method
    R Code for Class and Slide Examples:

    Chapter 1:

  • R function Wilcoxonsig.levels , code given also on slide 20.
  • R function Wilcoxon.normal , code that produced slide 32.
  • R function Wilcoxon.normalCDF , code that produced slide 33.
  • R function WilcoxonTiedRanks , code that produced
         slides 48-51, 53 and 54     .
  • R function psycholcounsel , code on slide 58.
  • R function psycholcounsel.sim , code on slide 62.
  • R function SiegelTukey , code used on slide 88
         (uses ST.rank, see next)    .
  • R function ST.rank , function required by SiegelTukey.
  • R function EDF.plots , code used to produce slide 95.
  • R function EDFpainEx6.plots , code used to produce slide 101.
  • R function KS.Kfun , code for calculating upper tail probabilities
         for the Kolmogorov-Smirnov test, slide 102    .
  • R function KSapprox , code producing slides 106-117,
         it uses ks.statistic.int (below) and KS.Kfun (above)     .
  • R function ks.statistic.int , function used by KSapprox (above).

    Chapter 2:

  • R function PsychEx1 , code used to produce results
         on slide 24    .
  • R function StochOrder , code used to produce the
         stochastic ordering plots on slide 29    .
  • R function shiftmodel , code used to produce the
         shift model plots on slide 31    .
  • R function PowerSim , code used to produce the
         simulation and approximation results on slide 53    .
  • R function Table2.5 , code used to produce the
         the table on slide 106    .
  • R function HLsim , code on slide 107.

    Chapter 3:

  • R function VitaminBsim , code on slide 38.
  • R function SignedRankExact , code discussed on slide 41.
  • R function BlockedWilcoxon , code introduced on slide 64, updated 5/20,09.
  • R function AlignedBlockedWilcoxon , code introduced on slide 78, updated 5/20/09.

    Chapter 4:

  • R function PowerSignRank , referred to on slides 30,31,37, 43, 45.
  • R function MuscleGain , referred to on slides 92 and 93.
  • R function MuscleGain.median , referred to on slides 96 and 97.

    R Installation Info and Introductory Guides:

    R Primer by Chris Green

    An Introduction to R by W.N. Venables and D.M. Smith and the R Development Core Team

    Verzani-SimpleR by John Verzani

    Mathematical Annotations in R by Paul Murrell and Ross Ihaka
    An Approach to Providing Mathematical Annotations in Plots by Paul Murrell and Ross Ihaka,
    Journal of Computational and Graphical Statistics, Vol. 9, No. 3, Systems and Languages (Sep., 2000), pp. 582-599. Can be downloaded via JSTOR (after logging in to get off-campus access).


    Recipes for fancy graphics in R

    For many more guides see R Contributed Documentation

    The R Project for Statistical Computing

    The Comprehensive R Archive Network

    For more documentation on R see:
    Manuals, FAQs, Newsletter, Wiki, Books,

    Home page of Fritz Scholz