New Predictive Solutions features
New types of visualization of results
-
MR Bar Chart – the bar chart MR is a bar chart designed to present the results of multiple-choice questions, which often appear in surveys. Questions of this type can be difficult to show in a visualization because the percentages do not necessarily add up to 100%—each respondent can select more than one answer. This type of data also allows for the use of two bases for calculating the percentage of the total number of respondents or answers. One of the advantages of the new form of visualization is the ability to simultaneously show the percentage of answers (what share of all answers a given category represents) and the percentage of respondents (what percentage of people selected a given option). This allows the results to be presented in a more complete and easier to interpret way.
MR Bar Chart (Assets owned by respondents)
(click to zoom-in)
-
Table MR Bar Chart – the table MR bar chart is a tool designed to present the results of multiple choice questions in the form of a table supplemented with bar charts. This combination provides the user with both numerical values and their graphical representation, which facilitates the comparison of answers and interpretation of results. As with a standard MR bar chart, it is possible to show both the percentage of responses (the share of all responses given to a given category) and the percentage of respondents (the percentage of people who selected a given option). The percentages of respondents do not necessarily add up to 100%, as each respondent can select more than one answer. The table chart can also be enhanced with the number of responses in each category and cell coloring, which allows you to quickly spot significant differences in the data.
Table MR Bar Chart (Assets owned by respondents)
(click to zoom-in)
New procedures for data preparation
-
Replace Names – this procedure allows you to easily organize and customize variable names in a dataset. This is particularly useful when working with files from different sources, where names can be long, inconsistent, or difficult to read. With the Replace names procedure, you can add prefixes or suffixes, replace parts of variable names, and anonymize them. It also allows you to convert characters and change case, which lets you quickly standardize and organize the file structure. Additional options, such as test mode, duplicate handling, and automatic repair of incorrect names, guarantee security and full control over the process.
The window of the Replace Names procedure
(click to zoom-in)
- Moving Measures – the new procedure allows you to analyze changes in data over time using a moving window. Various moving statistics are available to the user, such as: sum, mean, median, minimum, maximum, standard deviation, variance, and coefficient of variation. For selected variables, you can specify the frame spans (the number of periods taken into account in the calculation of statistics). The value calculated on the basis of several consecutive or previous values allows you to smooth the data, remove random spikes, and see trends more easily. This approach is particularly useful in time series analysis, e.g., when tracking changes in consumer sentiment, product ratings, or sales results. The result is new, more stable variables that help to better understand the direction and dynamics of change.
New procedure for exporting the results
-
Export Results – this procedure allows you to quickly export tables and charts to both raster and vector graphic files. With this functionality, you can easily prepare the resulting objects for presentation in applications such as PowerPoint or reports created in Word. The export preserves the appearance of the PS IMAGO PRO results window without losing formatting and ensuring high quality of the output objects. The user can select the file format and assign names with an automatic or custom prefix. Exporting results directly from the program significantly reduces working time and eliminates the risk of errors during manual copying.
The window for the Export Results procedure
(click to zoom-in)
New IBM SPSS Statistics 31.0 features
New Analytical procedures
- Bland-Altman Analysis – a graphical method used to compare two ways of measuring the same value, e.g., two devices or research methods. It helps to check how consistent the results are by showing both the difference between them (systematic error) and how much they differ from each other (range of variability of results).
- Proximity Mapping – a visualization technique that allows for the reduction of data dimensionality and the representation of relationships between objects in a spatial configuration. It allows you to see the relationships between objects – e.g., research subjects or products – and how close they are to each other in the analyzed structure. It is particularly useful in exploratory analyses, where it is important to capture similarities and differences between units.
- Distance Correlation – a new measure that allows you to examine the relationship between two variables independently of their shape. It detects both classic linear relationships and more complex patterns, such as curvilinear or irregular ones. Thanks to this measure, you can detect relationships that would remain hidden in classic correlation coefficients. This measure provides a more complete picture of the data and better responds to the challenges of practical analyses, where real-world processes are rarely perfectly linear.
Improvements for individual procedures
- Chi-Square Independence Test – now available as a separate procedure (ChiSquare) in the Descriptive Statistics menu, as well as in the new CHISQUARE INDEPENDENCE syntax.
- Levene's Test of Homogenity of Variance – added as an option in the Independent Samples t-test procedure, allows you to assess the equality of variance in the compared groups.
- Coefficient of variation (CV) – ratio of standard deviation to mean expressed as a percentage, available in the Frequency and Descriptive Statistics procedures, as well as in the Complex Samples module.
- Fill Image – new option in the Chart Properties window, allowing you to add background images to charts.
General improvements
- Dark Mode – a new interface appearance option that allows you to customize the analytical environment to your preferences and working conditions.
- Ergonomic Improvements – including the ability to change the properties of multiple variables simultaneously in the Variables view and new output composition configuration options that allow you to save your favorite table, chart, and results view styles as ready-made sets.