STAT411-3 - Methods of Data Analysis I - Fall 2017

When you can measure what you are speaking about and express it in numbers you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of an unsatisfactory kind - Kelvin (1824-1907)

  • Syllabus, Al's email, Office Phone: 994-5145, Office: Barnard (EPS) 304, Office Hours and Schedule.
  • R resources:
  • Writing a statistical report for Data Analysis problems; Summary of Statistical Findings examples; Scope of Inference Writing examples; report on iris analysis example and the markdown file that created it.
  • Other useful links:
  • Exams:
  • Homeworks:
    • DUE 11/20 HW9 on Chapters 9 and 10.
    • DUE 11/13 HW8 on Chapter 9.  Other problems on p. 261 (answers on p. 269): #1, 2, 4-11. 
    • DUE 11/6 HW 7 on Chapters 7 and 8.  Other problems on p. 227 (answers on p. 235): #1-14.
    • DUE 10/30 HW6 on Chapters 6 and 7.  Other problems on p. 170 (answers on p. 175): #1, 4, 6-11; on p. 198 (answers on p. 206): #3-11. 
    • DUE 10/9 HW5 on Chapter 5: non-parametric ANOVA and random effects ANOVA.  Biofilm dataSOLUTIONS.
    • DUE 10/2 HW4 on Chapter 5: ANOVA and PERMANOVA.  SOLUTIONS: pdf, Rmd. Other problems on p. 141 (answers on p. 147): #1-4, 6-10, 12.
    • DUE 9/22 HW3 on Chapters 2 and 3. SOLUTIONS. Other problems on p. 77 (answers on p. 84): #1-10, 14, 16, 17, 19.
    • DUE: 9/13 HW2: By hand (i.e., turn in a handwritten write-up), perform a 1-sample z-test and construct a 1-sample CI for some data set of interest (e.g., for the intrinsic writing group Sleuth3 dataset case0101 in R (see Chapter 1 notes)).   You choose a value for Ho: mu = mu0 that is different than 0.  You choose the confidence level.    Perform all 6 steps of the hypothesis testing framework, including a conclusion in terms of the problem.  SOLUTIONSOther problems on p. 51 (answers on p.56): #1-10, 11b.
    • DUE: 9/8  HW1 on Chapter 1, article: Your Brain on MethSOLUTIONSOther problems on p.22 (answers on p.26): #1-11, 15.
  • Labs:
    • 11/15 Working on R code for HW9.
    • 11/8 Working on R code for HW8.
    • 11/1 Working on R code for HW7.
    • 10/25 Working on R code for HW6.
    • 10/18 Lab 4 on regression, claims data.  Partial SOLUTIONS: pdf, rmd
    • 10/11 Review Chapters 1-3, 5, 6
    • 10/4 Working on R code for HW5.
    • 9/27 Lab 3, using R Markdown for reporting; working on R code for HW4.
    • 9/20 Working on R code for HW3.
    • 9/13 Lab 2, working with objects in R.
    • 9/6 Working on R code for HW1.
    • 8/30 Lab 1, getting started with R. Dairy data as: text, csv
  • Course Schedule
    • 11/22 - 11/24 THANKSGIVING BREAK!
    • 11/20 In class group work for extra credit
    • 11/17
    • 11/15 
    • 11/13 §9.4-9.6 Fitting any MLR, graphical tools for assessment
    • 11/10 VETERANS DAY observed, go thank a veteran
    • 11/8 §9.1-9.3 Fitting the basic MLR model with one predictor and one factor with an interaction, including factors with more than 2 levels, notes
    • 11/6 §9.1-9.3 Fitting the basic MLR model with one predictor and one factor with no interaction, notes
    • 11/3 §8.5 Extra sum of squares as model selection
    • 11/1 §8.5 Lack-of-fit test comparing SLR to ANOVA, notes
    • 10/30 §9.2-9.3 Basic MLR models: one predictor and a factor, or two predictors 
    • 10/27 §8.4 transforms, example from Lab 4: pdf, rmd
    • 10/25 §7.5, 8.2, 8.5 Correlation, robustness
    • 10/23 §7.4 Tests and CIs of regression parameters and mean response, PIs for a future response
    • 10/20 §6.4, 8.3, 8.6 Examples of planned vs unplanned comparisons, checking regression assumptions, notes
    • 10/18 §6.3 Planned vs unplanned comparisons, Bonferronis
    • 10/16 §7.1-7.3 Simple linear regression, notes: pdf, rmd
    • 10/13 Midterm Exam 1
    • 10/11 §6.4 Tukeys multiple comparisons after ANOVA
    • 10/9 §6.2 Testing linear combinations after ANOVA
    • 10/6 §6.1 Case study, Chapter 6 notes
    • 10/4 §5 When to use which ANOVA
    • 10/2 §5.6 Non-parametric ANOVA, random effects ANOVA
    • 9/29 §5 Permutation and randomization ANOVA
    • 9/27 §5.2-5.3 ANOVA as a comparison of reduced and full models, follow-up t-test for a planned comparison
    • 9/25 §5.3, 5.5 ANOVA model, checking ANOVA model assumptions with residuals
    • 9/22 §5.2-5.3 How ANOVA fits in our toolbox along with t-tests, randomization and permutation tests, non-parametric tests, Chapter 5 notes, skull data, diagANOVA.r
    • 9/20 §3.5 Transforming to normality: Box-Cox
    • 9/18 §3.4-3.5 Identifying and vetting outliers; identifying non-normality

    • 9/15 §3.2-3.3 Robustness to non-random sampling, resistance to outliers, Chapter 3 notes, timber data
    • 9/13 §2.3,  §3.2, §4.3.2 Two-sample z-test and t-test, robustness to non-normality
    • 9/11 §2 z and t confidence intervals

    • 9/8 §2.2  Central Limit Theorem, one-sample z-test and t-test, Chapter 2 notes: original, updated with 1-sample t-CI and example R-code
    • 9/6 §1.4 permutation test for analysis of observational data
    • 9/4 NO CLASS: LABOR DAY!

    • 9/1 §1.3: randomization test for analysis of experimental data
    • 8/30 §1.2: sampling, parameters, estimation, uncertainty quantification, Chapter 1 notes
    • 8/28 Welcome!   Remember the Scientific Method?