## Statistics 302, Sections 505-506, H.J. Newton Studying for the Final Exam Tuesday May 5, 1997, 10:30-12:30, Blocker 165

### Structure of the Exam

• 5 multiple choice similar to those on first exam.
• 5 multiple choice similar to those on second exam.
• 10 multiple choice covering material since second exam.
• A rank-sum test problem.
• One multi-part regression problem based on Stataquest output.
• One multi-part 1-way ANOVA problem based on Stataquest output.
• A Tukey-Kramer multiple comparison problem.

• Any needed formula will be supplied except the sample mean and variance and interquartile range.
• Make sure you can use the tables 2, 3, 4, 6, and 7 in the back of the book.
• Make sure you can plug numbers into formulas and get the right answer.

### Outline of Material Since Second Exam

• Chapter 10
• Mann-Whitney Rank Sum Test
• Assumptions
• Null hypothesis
• Ranking data
• Using Table VI (pg 611)
• Chapter 11
• Simple Linear Regression
• Probabilistic and deterministic models
• Independent and dependent variables
• Understanding assumptions about populations of y's for fixed x.
• Linearity of means
• Normality of populations
• Constancy of variances
• Independence of populations
• Checking assumptions
• Interpreting graphs of residuals
• Interpreting slope and intercept
• Understanding residuals
• Interpreting R^2
• Testing hypotheses about a slope
• Prediction intervals and confidence intervals
• Inferences about correlation coefficient (rho)
• The bivariate normal distribution
• Performing a test that rho is zero
• Chapter 13
• A completely randomized design one factor (1-way) design
• A randomized block design
• A completely randomized two factor (2-way) design
• Sources of variation for 1-way, randomized block, and 2-way
• The assumptions for ANOVA
• All poulations are normally distributed with the same variance
• Between and within sample variation
• The null and alternative hypotheses
• The Tukey-Kramer multiple comparison procedure for 1-way ANOVA
• Properties of the F curves
• Always positive, average value 1, curve gets more symmetric as degrees of freedom increase