NAG Statistical Add-Ins for Excel - Technical Content Details

Changes from Release 1.0

Users of Release 1.0 of the Add-Ins should note that while the basic functionality hasnot been changed between Release 1.0 and Release 1.1 there has been a major rewrite of theinterfaces taking into account comments from users. The interfaces of Release 1.0 are not compatible with the interfaces for Release 1.1. We hope users find the new interfaces easier to use and with the improved Help system will quickly adapt to them. Any future releases of these Add-Ins should not involve such a major change to the interfaces.

Additional functionality is provided by two new functions, MM_GLM and MM_MLR, in whichyou can specify a linear model by means of a model formula. There is also improved functionality in NAGExtract.

Improvements in Release 1.1

The latest version of the NAG Statistical Add-Ins for Excel is now available. Release 1.1 maintains the functionality of the original Add-Ins but there has been a major effort to improve the ease-of-use. We have taken into account the helpful comments we have received from our users in the design and presentation of the Add-Ins. The main improvements are:

  • New, more user-friendly interfaces for many functions • Greater use of suitable default values
  • Better annotation of the output
  • Algebraic specification of linear models
  • Improved help system with additional introductory material
  • New set of examples
  • Comprehensive error messages
  • A printed version the User's Guide is available
  • RTF and HTML versions of the User's Guide are provided
  • 'How to use' demonstration movies

Release 1.1 Contents

  • Basic Statistics distributions
    • Normal
    • t
    • F
    • gamma
    • beta
  • summary statistics
    • means
    • standard deviations etc.
    • five-point summary
    • tests for two-way tables
  • Regression and correlation
    • correlation and partial correlation
    • multiple linear regression (general linear model)
    • dummy variable generation
    • standardised residuals and influence measures
    • Durbin-Watson test
  • Time series
    • acf and pacf
    • ARIMA fitting and forecasting
    • transfer function fitting and forecasting
    • spectral analysis.
  • Generalised linear models (using standard links) for:
    • Normal
    • binomial
    • Poisson
    • gamma distributions
  • Multivariate methods
    • principal component and factor analysis
    • discrimant analysis
    • cluster analysis
    • metric and non-metric scaling

For more technical details, contact the NAG Response Centre at