Home
Ngene
Forum
Contact
 
Features
Screenshots
Documentation
Download
Purchase
Ngene is designed to be the single source of stated choice (SC) experimental designs. As such, it has an extensive range of features and outputs.

With Ngene you can:

Specify designs with great flexibility:

  • Generate designs with any number of choice situations, alternatives, attributes and attribute levels.
  • Maintain attribute level balance, or specify that an attribute must occur an exact number of times or between a minimum and maximum number of times.
  • Dummy and effects code attributes.
  • Specify logical rules to limit what attribute levels can coexist in a choice situation.
  • Interrogate design level correlations as calculated using a range of correlation formulas (Pearson product moment, G index, J index, Spearman rank, Point biserial, CP coefficient, H index).

Generate full and fractional factorial designs.

Generate orthogonal designs:

  • Maintain orthogonality either across or within alternatives.
  • Obtain orthogonal designs for a very large range of design dimensions.
  • Add blocking and foldover columns.
  • Generate optimal orthogonal in the differences designs.
  • Find the most efficient orthogonal design.

Generate efficient designs:

  • Report and optimize on efficiency measures including d, a, s (sample size), b (utility balance), and wtp (willingness to pay).
  • Report and optimize on efficiency measures for multinomial logit (MNL) models, mixed multinomial logit (MMNL) models (panel and cross sectional) and error components (EC) models (panel and cross sectional).
  • Account for prior uncertainty with normally and uniformly distributed Bayesian priors.
  • Report and optimize on the Bayesian mean, median, minimum, maximum and standard deviation.
  • Draw Bayesian and random parameter distributions with random, Halton, Sobol and MLHS draws, as well as Gaussian quadrature.
  • Optimize on more than one model and error measure type using model averaging.
  • Search for efficient designs using the pair swapping, RSC and modified Federov algorithms.
  • Report utilities, probabilities, the Fisher matrix and the covariance matrix for each model type.

Generate formatted HTML mockups:

  • Format the scenarios by placing design levels, text and radio buttons wherever you like within a table.
  • Format or relabel design levels for presentation.
  • Apply cascading style sheets (CSS) to instantly modify the appearance of the formatted scenarios (CSS files included, or create you own).
  • View the mockups directly within Ngene.

Interact with a modern user interface that maximizes flexibility:

  • Open and evaluate existing data files and designs.
  • Open files independently in the workspace or maintain syntax, data and output files within a project.
  • Retain all syntax runs and associated outputs during a session.
  • Interrogate any design found during a search.
  • Report design properties as needed - no need to decide what to report before the syntax runs, and no calculation of unnecessary properties during the search.
  • Easily view any number of user selected design properties in a grid, and copy directly to other applications including Excel.

and do much more...