DURATION | PRICE | inc-VAT | |
---|---|---|---|

4 | £1,800 | £2,160 | |

This course provides an introduction to and understanding of the application of statistical techniques such as Analysis of Variance, Linear Regression and Logistic Regression. This course is ideally suited to someone with prior statistical knowledge who undertakes statistical analysis in the language of SAS. |

### WHAT YOU'LL BE TAUGHT...

#### An introduction to Statistical Procedures

- Background statistical concepts
- Generating random samples using PROC SURVEYSELECT
- Generating descriptive statistics using PROC MEANS and PROC UNIVARIATE
- Examining the distribution of a dataset
- Producing confidence intervals
- Performing hypothesis tests
- Compare the mean statistic of two populations using PROC TTEST

#### Analysis of Variance (ANOVA)

- Use PROC GLM to compare the mean statistic of two populations using a One-Way ANOVA
- Use PROC GLM to compare the mean statistics of multiple populations using a One-Way ANOVA
- Use PROC GLM to compare the mean statistics of multiple populations using a Multiple-Factor ANOVA
- Learn how to deal with interactions in a Multiple-Factor ANOVA.
- Understand the impact of a blocking factor in ANOVA.

#### An Introduction to Linear Regression

- Examine the relationship between continuous variables using scatter plots
- Use PROC CORR to identify the strength of linearity between continuous variables
- Use PROC REG to perform Simple Linear Regression (SLR)
- Use PROC REG to perform Multiple Linear Regression (MLR)
- Understand the impact of Collinearity in Multiple Linear Regression and learn how to manage it
- Learn methods of building Multiple Linear Regression models
- Learn how to interpret different models
- Use Scatter plots to examine the residual statistics in a given model
- Learn methods of detecting outliers and influential observations

#### An Introduction to Logistic Regression

- Learn how to detect associations between variables using PROC FREQ
- Use PROC CORR to obtain correlation statistics for strength of association in categorical variables
- Use PROC LOGISTIC to build a simple Logistic Regression model with a continuous predictor
- Use PROC LOGISTIC to build a simple Logistic Regression model with a categorical predictor
- Use PROC LOGISTIC to build a Multiple Logistic Regression model
- Compare the explanatory and predictive ability of Multiple Logistic Regression models

#### A practical Business Application of Predictive Models using Logistic Regression

This session is aimed at allowing the Learner to run through a full model build process with the Instructor so that they gain a comprehensive understanding of the full model build process.

Learn 9 clear steps to build regression models, in a real-world environment, through an interactive model build process, including:

- The use of training and validation data
- Dealing with missing value
- How to reduce levels of categorical predictors using PROC CLUSTER
- How to remove redundant predictors using PROC VARCLUS
- Final model assessment and validation
- This session is aimed at allowing the Learner to run through a full model build process with the Instructor so that they gain a comprehensive understanding of the full model build process.

### WHAT YOU SHOULD ALREADY KNOW...

- Knowledge of descriptive and inferential statistics (including p-values and hypothesis testing) is essential
- Knowledge of Analysis of Variance and Regression techniques is preferable, not essential.
- Experience of running a program in the Language of SAS, creating a dataset, applying formats and running basic Statistical Procedures. This knowledge can be gained on our Foundations - The Language of SAS Course