STAT 705 FINAL EXAM REVIEW SHEET

(The final exam will include new material and questions from Exam 1 
and Exam 2 material.  Also study the review sheets for Exams 1 and 2.)

MORE REGRESSION TOPICS
------------------

I. Regression with Time Series data
  A. Meaning of autocorrelation
    1. Problems resulting from Autocorrelation 
    2. Errors composed of previous error term and independent noise
  B. First-order Autoregressive Error SLR Model
    1. Expected values of error terms
    2. Variances, covariances, correlations of error terms
    3. Test for significant covariate effect
  C. Durbin-Watson Test
  D. Remedies for Autocorrelation
    1. Estimating rho
    2. Transformations and back-transforming model
    3. Cochrane-Orcutt and other methods to estimate rho
    4. Forecasting and Forecasted Prediction Intervals

II. Regression and ANOVA with Missing Data
  A. Types of Missingness (MCAR and MAR)
  B. Listwise deletion, Multiple Imputation, Maximum likelihood
    1. Advantages/disadvantages of each
    2. Linear Regression method of imputation
  C. Repeated Measures with Dropouts
  D. Regression with Predictors having Missing Values


CATEGORICAL DATA ANALYSES
-------------------------

I. One-Way Tables
  A. Summarizing Categorical Data with Contingency Table
  B. Modeling Success Counts with Binomial
  C. Inference about Binomial Probability pi
    1. Sampling distribution of pi-hat
    2. CI for pi (Wald & score)
    3. z-test about pi
    4. Large-sample checks
    5. Small-sample alternative
  D. Chi-square goodness-of-fit Test (1 by c table)
    1. Calculating Observed and expected counts
    2. Test statistic and its null distribution
    3. Large-sample checks
II. Two-Way Tables
  A. Comparing Two Proportions (Independent Samples)
    1. Representation as 2 by 2 table
    2. Sampling distribution of (pi_1-hat - pi_2-hat)
    3. CI for difference of two proportions
    4. z-test to compare two proportions
    5. Large-sample checks
  B. Small-sample Alternative:  Fisher's Exact Test
    1. Assumptions of test
    2. Rationale behind P-value
    3. Odds ratio
  C. Comparing Two Proportions (Paired Samples)
    1. Difference between paired and independent samples
    2. Test statistic for McNemar's Test
    3. McNemar-type CI for (pi_1 - pi_2)
    4. Large-sample checks
    5. Small-sample alternative
  D. Chi-square Test for Independence (r by c table)
    1. Model for Cell Counts
    2. Structure of table of cell counts and of cell probabilities
    3. Calculating Observed and expected counts
    4. Test statistic and its null distribution
    5. Large-sample checks
    6. Small-sample alternative
  E. Chi-square Test for Homogeneity (r by c table)
    1. Difference in sampling process compared to test for independence
    2. Difference in hypotheses compared to test for independence
    3. Large-sample checks
    4. Small-sample alternative
III. Other Topics
  A. Loglinear Models
    1. Method for analyzing three-way tables
    2. Types of dependence represented by interaction terms in model
    3. Interpretation via odds ratios
    4. Comparing potential models using chi-square tests
  B. Multicategory (Polytomous) Logistic Regression
    1. Baseline-Category Logit Models for Nominal Responses
    2. Cumulative Logit Model for Ordinal Responses