New features in XLSTAT 2017
New features added to XLSTAT in 2017 include:
- ANOVA: A chart summarizing multiple comparison grouping results is now displayed when the ANOVA is launched on several dependent variables. Available in all XLSTAT solutions.
- Data sampling: It is now possible to split your data into a training set and a test set. Available in all XLSTAT solutions.
- Descriptive statistics: The level at which values should be split into stem and leaf can now be controlled. Also available in all XLSTAT solutions.
- Reliability analysis: An innovative feature that allows you to compute the reliability of a group of quantitative variables. Available in the Describing Data Menu of all XLSTAT solutions, it includes the internal consistency analysis based on Cronbach’s alpha. Many more new options will be added to our upcoming releases, which will be made available to you shortly.
- Agglomerative Hierarchical Clustering: The cophenetic correlation matrix as well as the associated cophenetic coefficient enable you to evaluate the classification quality. It can be computed by activating the corresponding option in the outputs tab of the dialog box. Available in all XLSTAT solutions.
- PLSPM: Many Goodness of Fit indexes have been added, including SRMR, GD, dULS, Chi-Square and Likelihood. To activate them, click on the corresponding option in the outputs tab of the Run dialog box. Available in XLSTAT-Psy, XLSTAT-Marketing and XLSTAT Premium.
- Sensory wheel: This new feature lets you build an attractive visual representation of the different sensations that can be triggered by a product. Available in XLSTAT-Sensory, XLSTAT-Marketing and XLSTAT-Premium.
- Correlation tests: The polychoric correlation measures the correlation between ordinal variables. Computation speed has been significantly increased. You can now filter results to display only the variables with the most important correlations (check the Options tab in the dialog box). Available in all XLSTAT solutions.
- Reliability analysis: The Guttman index is now available in all our solutions.
- A new guide that will let you choose an appropriate descriptive statistics feature is now publicly available online.
- Features history Save time by directly accessing recently used XLSTAT features through the Recent menu. Now available in all XLSTAT solutions.
- Propensity score matching This new feature helps pairing individuals in order to reduce confounding effects in observational studies. Available in XLSTAT-Biomed, XLSTAT-Ecology and XLSTAT-Premium under the Survival Analysis menu.
- Generalized Procrustes Analysis It is now possible to filter out PCA elements according to the sum of squared cosines. Feature available in XLSTAT-Sensory, XLSTAT-Marketing, XLSTAT-Ecology, XLSTAT-Psy et XLSTAT-Premium.
- Linear Regression Heteroscedasticity and autocorrelation among residuals can now be taken into account while building a model. Both options are available in all XLSTAT solutions. They can be found in the Linear Regression dialog box. Simply select the Options tab and then covariances sub-tab.
- Reliability Analysis It is now possible to compute the Intraclass Correlation Coefficient (ICC) as well as the Guttman L4 and Spearman-Brown reliability indices. The above options can be accessed in all XLSTAT solutions within the Reliability Analysis dialog box (Describing Data menu).
- Time-Intensity With this new feature you can now model the evolution of perception intensity related to an attribute across time and according to products and assessors. It is available in XLSTAT-Sensory, XLSTAT-Marketing and XLSTAT-Premium under the Sensory Data Analysis menu.
- Principal Component Analysis The dialog box has been redesigned in order to make it even more flexible to use, especially when launching analyses on several groups. It is now possible to direct variable tags following vector directions in the correlation circle chart. Available in all solutions, the Principal Component Analysis dialog box can be found under the Analyzing Data menu.
- One sample Wilcoxon Signed ranks Test: This new feature is a non-parametric equivalent of Student’s one sample t-test.
- Principal Component Analysis (PCA): Bootstrap confidence ellipses or convex hulls can now be plotted around observations. These options are available under the Charts / Bootstrap charts subtab (see below).
- Support Vector Machines (SVM): Multiclass responses can now be analyzed either through one-versus-one or one-versus-all analysis.
- Cox Proportional Hazards: Several new options including interaction effects, proportional tests & predictions.
- MS PowerPoint and MS Word export buttons (PC only): All XLSTAT outputs can now be exported in just a few clicks.
- XLSTAT Designer (PC only): This new interface allows you to configure customized statistical applications composed of XLSTAT features and options.
- Theme for tables: It is now possible to customize the theme of table outputs.
The changes from version 2015 to 2016, and those added in 2016, can be found here.