SAS Analytics

SAS Analytics

Courses Info

Prerequisite: SAS Basic + advance

Knowledge about Statistics is additional advantage

The below list is not exhaustive. There are many more concepts and industry applications covered in

the training. Below course is designed to cover both the certification topics as well as industry

applications requirement.


Basic Statistical Concepts

• Descriptive and Inferential Statistics

• Populations and Samples

• Parameters and Statistics

• Use of variables dependent and Independent

• Types of Variables Quantitative and Categorical

• Scales of measurement Nominal ordinal interval ratio

• Statistical Methods

• Exploring your data


Descriptive Statistics

• Describing your data

• Measures of Location

• Percentiles

• Measures of Variability

• Using descriptive statistics to answer data questions

• Means procedure – Using Proc means to generate statistics


Picturing your data

• Histogram

• Normal Distribution

• Assessing Normality

• Measures of Shape :: Skewness

• Measures of Shape : : Kurtosis Normal

• Probability plots

• Box Plots

• Assessing Normality with examples

• Univariate Procedure

• Statistical Graphics procedure in SAS

• Using SAS to picture your data


Confidence Intervals

• Point Estimators

• Variability and Standard Error

• Distribution of Sample Means

• Interval Estimators

• Confidence Intervals

• Normality and Central Limit Theorem

• Calculating confidence interval for the Mean

• Using Proc Means to generate confidence Interval


Hypothesis Testing

• Decision Making Process

• Steps in hypothesis testing

• Types of Errors and Power

• The P value Effect Size Sample Size

• Statistical Hypothesis Test

• The T Statistic and the T Distribution

• Comparing Sample and Hypothesized Means

• Using Proc univariate to generate statistics

• Using Proc Univariate to perform a hypothesis test


The T-test

• Assumptions for Two sample T Test

• F test for Equality of Variance

• Comparing Group Means

• Identifying your data for the T test

• Running Proc T test in SAS Examining the equal variance t- test and p- value

• Examining the unequal variance t- test and

• Interpreting two sample t test result

• One Sided Test Test for difference on one

• The T test procedure and Side option Performing a one sided test.


One Way Anova-Introduction

• Anova Overview

• The ANOVA hypothesis

• The ANOVA Model Sum of Squares

• Assumptions for Anova

• Predicted and Residual Values

• Comparing group means with one way Anova

• Examining the descriptive statistics across means

• The GLM procedure


Anova with data from a Randomized block design

• Observational Studies Vs Controlled Experiments

• Nuisance Factors Including a blocking variable in the model

• More Anova Assumptions

• Creating a Randomized block design

• Performing Anova with Blocking


Anova -Post Hoc Test

• Multiple Comparison Methods

• Tuckey’s Multiple comparison method

• Dunnet’s multiple comparison method

• Determining which mean is different

• Diffogram and Control Plots

• Proc GLM with LS means

• Performing post hoc pairwise comparison


Two Way Anova with Interactions

• N Way Anova

• Interactions

• Two way Anova Model

• Using Two way Anova

• Identifying your data

• Applying the two way Anova Model

• Performing two way Anova with Interactions

• Performing Post Hoc Pairwise Comparison


Exploratory data analysis

• Using scatter plots to describe relationship between continuous variables

• Using Correlation to measure relationship between two continuous variables Hypothesis testing for a correlation

• Avoiding common errors in interpreting correlations

• Avoiding common errors- Outliers Causal and Effect

• Types of relationships

• Producing correlation statistics and scatter plots using Proc Corr.

• Using PROC CORR to produce correlation matrix and scatter plots

• Examining correlations between predictor variables.


Simple Linear Regression

• How SAS performs simple linear regression

• Measuring how well the model fits the data

• Comparing regression model to a baseline model

• Hypothesis testing for linear regression

• Assumption of simple linear regression

• The REG procedure: Performing Simple linear regression

• Scoring predicted values using parameter estimates

• Storing parameter estimates using Proc Reg

• Score using Proc score.

. Multiple Linear Regression

• Advantages and Disadvantages of Multiple Regression

• Common Applications

• Picturing the model for Multiple Regression

• Hypothesis testing for multiple regression

• Assumptions for multiple regression

• The Reg Procedure performing multiple linear regression


Model Building and interpretation

• Approaches to selecting model

• SAS and automated approaches to modeling

• All possible regressions approach to model building

• The REG procedure using all techniques

• The REG procedure using automatic selection

• The REG procedure Estimating and testing coefficients for selected models

• The Stepwise selection approach to model building

• Specifying Stepwise selection in SAS

• The REG procedure performing stepwise regression

• Using alternate significance criteria for stepwise models


Examining Residuals

• Assumption for regression

• The importance of plotting data and checking

• Verify Assumptions using residual plots

• Detecting outlies using residual plots

• The REG procedure producing default diagnostics

• The REG procedure specific diagnostics


Detecting Collinearity

• Understanding Collinearity

• The REG procedure detecting Collinearity

• The REG procedure Calculating diagnostics for Collinearity

• The REG Procedure Dealing with Collinearity

• Using an effective modeling cycle

• Categorical data analysis


Logistic Regression-Introduction

• Modeling a binary response

• The Logistic Procedure

• Dummy variables

• Specifying a parameterized method in class statement

• Effect Coding Reference Cell coding

• Fitting a binary logistic regression model

• Interpreting odds ratio for a categorical predictor

• Interpreting odds ratio for a continuous predictor

• Comparing pairs to assess the fit of a model.


Multiple Logistic Regression

• Multiple logistic regression

• The backward elimination method for variable selection

• Adjusted Odds ratio

• Specifying the variable selection method in model statement

• Fitting a multiple logistic regression model

• Comparing the binary and the logistic regression model

• Backward elimination method with interactions

• Specifying interaction in the model statement

• Fitting a multiple logistic regression with interaction

• The Odds ratio statement

• Fitting a multiple logistic regression with all Odds ratio

• Comparing multiple logistic regression models

• Interaction plots

• Lift Charts

• ROC curve

• Sensitivity and Specificity

• Variable transformation using bivariate analysis