Developing better models for calculating missing data

 

Statistics Missing Values is used for survey research, social science, data mining and market research to validate data. This module allows you to examine data to identify missing data, estimate summary statistics and impute missing values using statistical algorithms.

With Statistics Missing Values, you can impute missing data, arrive at better conclusions and eliminate hidden influences.

 

• Quickly diagnosing missing data imputation problems using diagnostic reports.

• Replacing missing data values with estimates using a multiple imputation model.

• Visualising and analysing models to obtain detailed information and improve data management.

 

 

Quickly diagnosing missing data imputation problems

 

• Examining data from different points of view using six diagnostic reports.

• Diagnosing missing data using the data model report, which provides a specific overview of the data.

• Determining the extent of missed data and extreme values for each case.

 

Replacing missing data values with estimates

 

• Understanding the missing patterns in the dataset and replacing missing values with plausible estimates.

• Benefiting from an automatic imputation model that selects the best method according to the characteristics of the data or customising the imputation model.

• Modeling the individual datasets created, using techniques such as linear regression or expectation maximisation algorithms to produce specific parameter estimates.

• Obtaining final parameter estimates by pooling estimates and inferential computation statistics that consider variations in and between imputations.

 

Visualising and analysing models

 

• Displaying missing case data and all variables using the data model table.

• Determining the differences between missing and non-missing groups for a variable associated with the separate t-test table.

• Evaluating the percentage association of missing data for one variable with missing data for another variable, using the percentage mismatch of the model table.