Course Description

Combines intermediate and advanced statistical methods with practical research applications and computer software. Develops commonly used statistical models such as Two and Three-Way Analysis of Variance as well as the analysis of Fractional Factorial Designs for the solution of common business and industrial research problems. The statistical models are implemented and interpreted in the context of actual data sets using available statistical software.

Outline

1. Measures of Relationship - An overview of the major indices and tests associated with measures of relationship for Nominal, Ordinal, and Continuous variables; including
   1. Contingency Table Analysis; Phi, C, V
   2. Youden’s J-Index of Predictive Efficiency
   3. Cohen’s Kappa
   4. Kendall’s Coefficient of Concordance
   5. Spearman’s Rank Correlation Coefficient
   6. The Biserial and Point-Biserial Coefficients of Correlation
2. Simple regression and correlation
   1. Simple regression and correlation
   2. Using Simple Regression to Describe a Linear Relationship
   3. Testing Inferences About the Population Regression Line, the Intercept, and Slope, and the Underlying Assumptions of the Model
   4. Assessing the Fit of the Regression Line – Using the ANOVA Table
   5. The Coefficients of Correlation, Determination, and Alienation
   6. Prediction and Forecasting; Confidence and Prediction Limits
   7. Using SPSSPc for Correlation & Regression Analyses
   8. Generating Output & Testing Assumptions Using MVPStats and SPSSPc
3. Introduction to multiple regression analysis
   1. Underlying Theory & Assumptions – Linear Models
   2. Testing Inferences About the Regression Coefficients
   3. Assessing the Fit of the Regression Line; the ANOVA Table, the Coefficient of Determination and the
      Multiple Correlation Coefficient
   4. Full & Reduced Models; Forward and Backward Regression Approaches
   5. Prediction Using the Multiple Regression Model
   6. Lagging Variables in Time Series Analyses
   7. Generating Output with SPSSPc
4. Assessing the assumptions of the Multiple Regression Model
   1. Testing The 5 Basic Assumptions of the Model
      1. The Relationship is Linear
      2. Variance is Constant
      3. The Residuals are Normally Distributed
      4. The Residuals are Independent
      5. There is No Multicollinearity
   2. Residual Analysis and Corrections for Model Violations
   3. Utilizing Other Influence Statistics (Durbin-Watson, Cook’s D)
   4. Handling ‘Outliers’
   5. Generating Response Surfaces
   6. Multiple Regression and Model-Fitting Utilizing TableCurve3D.
5. Additional topics in multiple regression analysis
   1. Fitting Curvilinear Relationships
      1. Polynomial Regression & 2nd Order Models
      2. Reciprocal Transformation of the X Variable
      3. Log Transformation of the X Variable
      4. Log Transformations of Both the X and Y Variables
   2. Non-Linear Regression and Model-Fitting Using MVPStats and Table Curve2D.
6. Special topics in multiple regression analysis
   1. Using and Interpreting Indicator (Dummy) Variables
   2. Analyzing Interaction Effects
7. Introduction to Fractional Factorial Designs & Considerations Related To Conducting Screening Experiments
   1. The Fractional Design - Basic Theory
   2. Developing of Extreme Screening Designs: Plackett-Burnham Matrices
   3. Developing High Resolution Designs
   4. Extending the Latin Square : Orthogonal Arrays
8. Introduction to data screening
   1. CHAID and associated analytical procedures
   2. Combining CHAID analyses with factorial designs, ANOVA, and multiple regression analysis to identify critical, significant, and trivial factors and variable

Prerequisites

EMEN 5900 & 5610; or APPM 5580 & 7400a; or the equivalent.

Textbook(s)

None.

Hardware/Software

MVPstats, SPSS


Syllabus