MANUALS FOR UNIVARIATE AND MULTIVARIATE STATISTICS

by Craig A. Wendorf

For the purposes of these manuals, univariate techniques are those which utilize only a single dependent variable. Therefore, designs which utilize multiple independent variables can be and are included in the univariate section of statistical techniques.

Multivariate methods provide an optimal linear combination of dependent variables that satisfy specific statistical criteria. These procedures include: (a) multivariate analogues of correlation and prediction techniques, (b) multivariate analogues of ANOVA significance tests, and (c) the classification of individuals into categories on the basis of predictor information.

Part I: Probability and Hypothesis Testing

• Introduction to Probability
• Intermediate Probability
• Distributions and Hypotheses

Part II: Statistical Tests for Nominal Data

• Binomial and Chi-Square Distributions
• Testing Chi-Square Hypotheses

Part III: 1 & 2 Sample Parametric Tests

• One Sample t and z Tests
• Decision Errors and Power
• Two Sample t and z Tests
• Further Issues in Hypothesis Testing

Part IV: Correlation and Regression

• Overview of Regression
• Types of Correlation Coefficients
• Hypothesis Testing with Correlations

Part V: Analysis of Variance (ANOVA)

• Terms and Concepts
• Computational Formulae for ANOVA
• Planned Comparisons
• Trend Tests
• Post Hoc Comparisons
• Repeated Measures Designs
• Randomized Blocks Designs

Part VI: Nonparametric Analogues

• Overview of Non-Parametric Tests
• Tests for Two Independent Tests
• Tests for Matched Groups

Part VII: Multiple Regression (MR)

• Overview of Multiple Regression
• Prediction with Continuous Variables
• Using Categorical Variables in MR
• Overview of Non-Linear Regression

Part VIII: Analysis of Covariance

• Overview of ANCOVA

Part IX: Multivariate Analysis of Variance

• Overview of MANOVA
• Analyses following a Significant MANOVA

Part X: Other General Linear Models

• Cannonical Variate Analysis
• Linear Discriminate Function Analysis