Data Quality for Analytics Using SAS

Chapter 8: Analytics for Data Quality

8.1 Introduction

8.2 Correlation: Problem and Benefit

Problem description

Imputing missing values based on other variables

Substituting the effect of unavailable or unusable variables

Multicollinearity or the need for independent variables

Sign inversion

Derived variables from transactional data

Derived variables for customer behavior

8.3 Variability

General

Statistical variability and the significance of p-values

Introducing variability

Instability of the business background and definitions

Undescribed variability

8.4 Distribution and Sparseness

General

Missing values

Distribution of interval variables

Outliers

Categorical variables and rare events

Grouping sparse categories

Sparse values in time series ...