Amos 4.0

With AMOS 4.0, you can quickly specify, view, and modify your model graphically using simple drawing tools. Then assess your model's fit, make any modifications, and print out a publication quality graphic of your final model. And AMOS 4's online Help system answers the most basic to the most complex SEM questions. AMOS 4.0 quickly combines components of regression analysis, simultaneous equation models, factor analysis, and ANOVA into one comprehensive model. Any variable--observed or latent--can predict any other variable. And AMOS takes measurement error into account, giving you better estimates. It can also be used to create latent variables from observed data. AMOS 4.0 features Full Information Maximum Likelihood treatment of missing data to ensure cases are not left out when data are incomplete. The program automatically calculates correct standard errors and fit statistics, assuming normality and a Missing at Random missing data mechanism. Your estimates become more efficient and less biased at the same time. AMOS's nonnative bootstrapping options evaluate model estimates with any empirical data--not just normally distributed data. The bootstrapping and Monte Carlo options make it easy to obtain bias and standard error estimates for any parameter, including standardized coefficients and effect estimates, and provides derived statistics. AMOS 4.0 also tests multivariate normality and multivariate outlier diagnostics. Analysis of missing data by AMOS's state-of-the art maximum likelihood analysis now comes with a complete set of fit indices. AMOS 4.0 attempts to estimate saturated and independence models automatically with missing data. If successful, it computes fit chi-square statistics and all derived fit indices. The AMOS 4.0 User's Guide contains an overview of the most common AMOS functions, three initial tutorials and a walk through 21 worked examples. This guide explores many facets of SEM and shows how to apply AMOS to give more complete results when multivariate models are fit to all the data.