Blimp Software

NEW!  Blimp 3.0 Beta Version Update

Blimp 3 for Mac OS and Windows is now available as a beta version while we finish the user guide. Blimp 3 includes new computational engine for single- and multilevel path and latent variable models with missing data. The factored/sequential estimation approach allows for easy specification of complex latent variable models that are difficult or impossible to estimate in a joint modeling framework (e.g., latent factors interacting with manifest variables, latent variable by latent variable interactions, random effects as predictors or outcomes). Many new features for missing data handling, including simple specification of selection and pattern mixture models for missing not at random processes.

Blimp 2.2 Application Download

Blimp 2.2 for Mac OS and Windows is now available with many exciting new features. Blimp was originally designed as a multiple imputation program, but the application now offers general-purpose Bayesian estimation for a wide range of single-level and multilevel regression models with two or three levels, with or without missing data. In particular, the application offers estimation routines for interactive and polynomial effects that are not yet available in other software packages, including support for binary, ordinal, and nominal variables. Blimp was developed with funding from Institute of Educational Sciences awards R305D150056 (Craig Enders, PI; Roy Levy, Co-PI) and R305D190002 (Craig Enders, PI; Brian Keller and Han Du, Co-PIs).  Algorithmic development by Craig Enders, Brian Keller, and Han Du. C++ programming by Brian Keller. Qt graphical user interface development by Brian Keller and Behrouz NematiPour.

The newest version of Blimp will automatically download new updates as they become available, so your software will always be current.

  • Blimp 2 for Mac OS (64 bit; requires Mojave OS 10.14 or higher)
  • Blimp 2 for Windows (64 bit)
  • Blimp 2 with statically linked libraries for Ubuntu (64bit) and Fedora (64bit) distributions of Linux
  • Blimp 2 with dynamic linking to user-supplied libraries for other Linux distributions (see README file for details)

Blimp 2.2 Features

  • Bayesian estimation of single-level, multilevel (up to three levels), and multiple group regression models with complete or incomplete data
  • Posterior summaries of all model parameters from Bayesian estimation (posterior mean, median, standard deviation, and credible interval)
  • Fully conditional specification multiple imputation for single-level, multilevel (up to three levels), and multiple group regression models
  • Missing data handling for normal, binary, ordinal, or nominal variables
  • Automatic dummy coding for nominal variables
  • New simplified scripting language and redesigned output
  • New graphical interface with automatic updates when new features become available
  • New graphical engine that creates trace plots for all model parameters
  • Rights and Sterba variance explained effect sizes for multilevel models
  • Bayesian estimation for interactive and polynomial effects with complete or missing data
  • Bayesian estimation with grand mean centering (all models) and group mean centering (two- and three-level models)
  • Post-hoc probing of interaction effects with continuous or categorical moderators
  • Bayesian estimation of conditional effects (simple intercepts and slopes) in regression models with interaction effects
  • Discrete and latent imputations for binary, ordinal, and nominal variables
  • Fully conditional specification or Bayesian estimation with level-2 and level-3 cluster means modeled as latent variables
  • Contextual effects models with latent group means or manifest group means
  • Interaction effects with latent group means
  • Bayesian estimates of random intercepts and slopes for multilevel modeling diagnostics
  • Automatic updates, so the software is always up to date
  • Various algorithmic and interface enhancements (e.g., random starting values, options for saving various estimates and output)

User Guide Examples

The Blimp User Guide illustrates imputation for several common single-level and multilevel analyses. Each example comes with a zip archive that includes slides explaining the analysis, the raw data files, Blimp scripts, and analysis scripts for Mplus, R, SAS, SPSS, and Stata. The examples are also included with the Blimp installation in the Applications folder.

Blimp Examples from Missing Data Workshop at UCLA

  • Slides, software, data, and analysis scripts from a three-day missing data workshop at UCLA. Includes several Blimp scripts for Bayesian analyses and multiple imputation that use new data sets from the forthcoming Applied Missing Data Analysis v 2.0.
  • Click here to download the slides, data files, and analysis scripts.

Blimp Papers

The Bayesian estimation and FCS imputation routines in Blimp are described in the following papers.

Enders, C.K., Du, H. Keller, B.T. (2019). A model-based imputation procedure for multilevel regression models with random coefficients, interaction effects, and other nonlinear terms. Psychological Methods. Advance online publication.

  • Click here to download the paper
  • Click here to download the online supplement and technical appendix

Enders, C.K., Keller, B.T., & Levy, R. (2018). A fully conditional specification approach to multilevel imputation of categorical and continuous variables. Psychological Methods, 23, 298-317.

  • Click here to download the paper
  • Click here to download the technical appendix
  • Click here to download the data file and analysis scripts

Blimp Support

Questions or suggestions? Email Craig Enders