What's new in versions 29, 29.0.1 and 29.0.2.

IBM SPSS Statistics 29.0.2

Python Extensions

Three new Python Extensions are now made available in the Extension Hub:

SPSSINC_TRANS
SPSSINC_MODIFY_TABLES
STATS_IMBALANCED

General Enhancements: Link to the IBM watsonx.ai page on welcome dialog available.; Upgrade to Apache POI from 5.2.1 from 4.1.2 to address the security and vulnerability issues.

Deprecated function

User Feedback: From version 29.0.2 onwards, Medallia will not be available to provide feedback.

IBM SPSS Statistics 29.0.1

Data Editor: A new Overview tab provides information on characteristics of the data in a dataset or file, with summaries of variable types, measurement levels, missing data, and allows drill down into individual variables with appropriate charts and summary statistics based on measurement level definitions.

Analyze procedures

Parametric Survival Regression models for recurrent events data

Click Analyze > Survival > Parametric Shared Frailty Models to estimate parametric survival models for recurrent events data by incorporating a shared frailty term. This term is treated as a random component to account for an unobserved effect due to individual or group-level variability.

PRESS (Predicted Residual Sum of Squares) statistic in Linear Regression

Click Analyze > Regression > Linear and the Statistics button to obtain the predicted residual sum of squares (PRESS) statistic, a popular cross-validation-type statistic for assessing linear regression models.

Youden's Index for ROC (Receiver Operator Characteristic) Curves

Click Analyze > Classify > ROC Analysis and the Display button to request Youden's index at each point along an ROC curve. Youden's index is a popular single-number summary of sensitivity and specificity for each cut point.

Export random effects predictions (EBLUPs) to datasets or files when estimating mixed models

Click Analyze > Mixed Models > Linear and when one or more Random specifications request the display of random effects predictions, select the Export button to create new datasets or files with the EBLUPs.

Click Analyze > Mixed Models > Generalized Linear and when one or more Random specifications request the display of random effects predictions, on the Model Options tab, select the Export item to export create new datasets or files with the EBLUPs.

Enhancements to Dialog Box Functions

Percentiles

Click Analyze > Descriptive Statistics > Percentiles to access a new dialog box featuring full access to all percentiles functions in the EXAMINE procedure, including specification of quartiles or custom percentiles, by using any of the five available estimation methods, and bootstrapped confidence intervals.

Click Analyze > Descriptive Statistics > Explore, then the Statistics button to have full access to the same function as the stand-alone Percentiles dialog box.

Linear Regression

Click Analyze > Regression > Linear, then the Statistics button to access a Selection criteria checkbox to access SELECTION keyword statistics.

Click Analyze > Regression > Linear, then the Options button to be able to specify the TOLERANCE keyword level for the CRITERIA subcommand for handling variables exhibiting near collinearity.

Cox regression

Click Analyze > Survival > Cox w/ Time-Dep Cov to access a redesigned dialog box that now allows specification/computations of multiple time-dependent covariates for Cox regression models.

Command syntax

SURVREG RECURRENT: Estimates parametric survival models for recurrent events data via incorporation of a shared frailty term. This term is treated as a random component to account for an unobserved effect due to individual or group-level variability.
ROC ANALYSIS: PRINT subcommand's COORDINATES=ROC keyword offers YOUDEN option for Youden's Index, combining sensitivity and specificity into a single measure of discrimination at each possible cut point on the ROC curve.
MIXED: Adds an OUTFILE subcommand with an EBLUPS keyword to export EBLUPs or random effects parameter predictions to datasets or .sav files. If multiple sets of EBLUPs are requested on RANDOM subcommands via SOLUTION keywords, the FILE_SEPARATE keyword can be used with TRUE or FALSE to save predictions in one or multiple data sets or files.
GENLINMIXED: Adds an EBLUPS keyword to the OUTFILE subcommand to export EBLUPs or random effects parameter predictions to datasets or .sav files. If multiple sets of EBLUPs are requested on RANDOM subcommands via SOLUTION keywords, the FILE_SEPARATE keyword can be used with TRUE or FALSE to save predictions in one or multiple data sets or files.

General Enhancements

Output Modify: Output Modify enhancements provide Right-click Modify Output shortcuts in the output pivot tables to quickly access several common features like transpose, sort column, hide column, and highlight cells in column.
Search dialog: Enhancements to the Search dialog include searching for variables, searching the IBM SPSS Community, searching the IBM SPSS YouTube channel, and more.

Deprecated Functionality

Import from and Export to dBase files: The options to import from or export to a dBase file format are removed from the graphic user interface. The underlying GET TRANSLATE and SAVE TRANSLATE command syntax still supports this feature.
ROC Curve: The menus and dialogs for ROC command are removed and this command is now available as syntax only.

IBM SPSS Statistics 29

Analyze procedures

Linear OLS Alternatives

Elastic Net: Click Analyze > Regression > Linear OLS Alternatives > Elastic Net to obtain a Linear Elastic Net Regression analysis. The new Linear Elastic Net extension procedure uses the Python sklearn.linear_model.ElasticNet class to estimate regularized linear regression models for a dependent variable on one or more independent variables. Regularization combines L1 (Lasso) and L2 (Ridge) penalties. The extension includes optional modes to display trace plots for different values of alpha for a given L1 ratio, and to select the L1 ratio and alpha hyperparameter values based on cross validation. When a single model is fitted or cross validation is used to select the penalty ratio and/or alpha, a partition of holdout data can be used to estimate out-of-sample performance.
Lasso: Click Analyze > Regression > Linear OLS Alternatives > Lasso to obtain a Linear Lasso Regression analysis. The new Linear Lasso extension procedure uses the Python sklearn.linear_model.Lasso class to estimate L1 loss regularized linear regression models for a dependent variable on one or more independent variables, and includes optional modes to display trace plots and to select the alpha hyperparameter value based on cross validation. When a single model is fitted or cross validation is used to select alpha, a partition of holdout data can be used to estimate out-of-sample performance.
Ridge: Click Analyze > Regression > Linear OLS Alternatives > Ridge to obtain a Linear Ridge Regression analysis. The new Linear Ridge extension procedure uses the Python sklearn.linear_model.Ridge class to estimate L2 or squared loss regularized linear regression models for a dependent variable on one or more independent variables, and includes optional modes to display trace plots and to select the alpha hyperparameter value based on cross validation. When a single model is fitted or cross validation is used to select alpha, a partition of holdout data can be used to estimate out-of-sample performance.

Parametric Accelerated Failure Time (AFT) Models

Click Analyze > Survival > Parametric Accelerated Failure Time (AFT) Models to obtain a Parametric Accelerated Failure Time (AFT) Model analysis, which invokes the parametric survival models procedure with nonrecurrent life time data. Parametric survival models assume that survival time follows a known distribution, and this analysis fits accelerated failure time models with their model effects proportional with respect to survival time.

Pseudo-R² measures in Linear Mixed Models and Generalized Linear Mixed Models

Pseudo-R² measures and the intra-class correlation coefficient are now included in Linear Mixed Models and Generalized Linear Mixed Models output (when appropriate). The coefficient of determination R² is a commonly reported statistic because it represents the proportion of variance that is explained by a linear model. The intra-class correlation coefficient (ICC) is a related statistic that quantifies the proportion of variance that is explained by a grouping (random) factor in multilevel/ hierarchical data.

Command syntax

GENLINMIXED: The output now includes pseudo-R² measures and the intra-class correlation coefficient (when appropriate).
LINEAR_ELASTIC_NET: The new extension command uses the Python sklearn.linear_model.ElasticNet class to estimate regularized linear regression models for a dependent variable on one or more independent variables.
LINEAR_LASSO: The new extension command uses the Python sklearn.linear_model.Lasso class to estimate L1 loss regularized linear regression models for a dependent variable on one or more independent variables. The command includes optional modes to display trace plots and to select the alpha hyper-parameter value that is based on cross-validation.
LINEAR_RIDGE: The new extension command uses the Python sklearn.linear_model.Ridge class to estimate L2 or squared loss regularized linear regression models for a dependent variable on one or more independent variables. The command includes optional modes to display trace plots and to select the alpha hyper-parameter value that is based on cross-validation.
MIXED: The output now includes pseudo-R² measures and the intra-class correlation coefficient (when appropriate).
SURVREG AFT: The new command invokes the parametric survival models procedure with nonrecurrent life-time data.

Python and R upgrades

Python 3.10.4 and R 4.2.0 are a part of the IBM® SPSS® Statistics 29.

Select Cases - hidden cases

Unselected cases are no longer hidden in the Data Editor when a subset of cases is selected, and the unselected cases are not discarded. This represents a return to the behavior of Statistics 27.0.1 and earlier versions.

Violin plots

The Graph board Template Chooser includes a new violin plot, which is a hybrid of the box and kernel density plots. Violin plots show peaks in the data and are used to visualize the distribution of numerical data. Unlike a box plot that can show only summary statistics, violin plots depict summary statistics and the density of each variable.

Workbook mode enhancements

Two new workbook toolbar items: Show/Hide all syntax windows and Clear all output.
New button on the Status bar to switch between Classic (Output and Syntax) and Workbook modes.

Search enhancements

The Search feature now provides options for entering terms directly in a toolbar field and for viewing results in a drop-down pane.