Buy Now
What's New in Amos 6.0
Experienced and new users alike will appreciate these new features and benefits to Amos 6.0, which enable you to:
- Obtain Bayesian estimates of model parameters and other quantities. Bayesian analysis enables you to apply your subject-area expertise or business insight to improve estimates by specifying an informative prior distribution. Markov chain Monte Carlo (MCMC), the underlying estimation method, specifies the Bayesian technique that Amos will use. Bayesian analysis enables you to:
- Reliably fit structural equation and related models to smaller samples
- Estimate any function of model parameters. For example, compute the difference between direct and indirect effects. Observed data can be complete or incomplete.
- Investigate the assumptions of maximum likelihood estimation by plotting the marginal likelihood of any parameter
- Avoid inadmissible model parameter estimates, such as negative variance estimates, through the choice of prior distribution or a global option
- Prevent unstable solutions in systems of linear equations in non-recursive models (for example, models with bi-directional causality) through a global setting
- Perform tests of custom hypotheses that are not easily obtained using maximum likelihood or other estimation methods
- Obtain optimal asymmetric credible intervals for indirect effects
- Impute missing values or latent factor scores. Choose from three data imputation methods: regression, stochastic regression, or Bayesian. Use regression imputation to create a single completed dataset. Use stochastic regression imputation or Bayesian imputation to create multiple imputed datasets. You can impute missing values or factor scores.
- Use regression imputation to create a single completed dataset. Predicted values form the regression equations that result and are substituted for the missing values.
- Use stochastic regression imputation or Bayesian imputation to create multiple imputed datasets. Stochastic regression imputation uses maximum likelihood-based parameter estimates from structural equation modeling. Missing values are imputed by drawing at random from their conditional distribution given the observed data and assuming that the parameter values are equal to their maximum likelihood estimates. Multiple Bayesian imputation is similar to stochastic regression imputation but takes into account the fact that the parameter values are estimated rather than known.
- Expand the capabilities of Amos using popular Microsoft programming languages, including Visual Studio® and C#.
- Enjoy more control over the Amos interface. Improved user interface features enable you to:
- Use print previews to see how path diagrams will look when printed
- Zoom, scroll, and magnify objects more effectively
- Create variable path diagrams with one click
Buy Now
