STATS PROPOR REGR

Linear models for dependent variables that are proportions

This procedure estimates a linear model with a link function for equations where the dependent variable is a proportion with a beta distribution. Mean and precision parameters of the distribution are estimated. Residuals and fitted values can be saved. The estimated model can be saved and used for predictions on new data with the STATS PROPOR REGRPRED extension command.

---

Requirements

• IBM SPSS Statistics 18 or later and the corresponding IBM SPSS Statistics-Integration Plug-in for R.

Note: For users with IBM SPSS Statistics version 23 or higher, the STATS PROPOR REGR extension is installed as part of IBM SPSS Statistics-Essentials for R.

---

Installation intructions

1. Open IBM SPSS Statistics
2. Navigate to Utilities -> Extension Bundles -> Download and Install Extension Bundles
3. Search for the name of the extension and click Ok. Your extension will be available.

---

License

• Apache 2.0

Contributors

• JKP, IBM SPSS