What's New in SPSS 11.0 for Windows
All are planned features
and are subject to change.
New statistical procedures added in SPSS add-on modules
- Use Linear Mixed Models (MIXED, in SPSS Advanced Models) to get the
most accurate predictive models when you have a nested structure to your
data. Its flexibility means you can formulate a wide variety of models,
including: Fixed Effects ANOVA Model, Randomized Complete Blocks Design,
Split-Plot Design, Purely Random Effects Model, Random Coefficient
Model, Multilevel Analysis, Unconditional Linear Growth Model, Linear
Growth Model with a person-level covariate, Repeated Measures Analysis
and Repeated Measures Analysis with time-dependent covariate. In
addition, you can work with repeated measure designs, such as incomplete
repeated measurements in which the number of observations varies across
subjects.
- Multinomial Logistic Regression (MLR) enhancements made in SPSS
Regression Models. Use MLR to target customers, students or constituents
by saving predicted probabilities, predicted response categories,
probability of the predicted response categories and probability of the
actual response categories.
- Improved scalability and performance of MLR (added in SPSS 10.1)
means MLR runs approximately ten times faster than previous versions
did.
- Improved performance of General Linear Models (GLM) (added in SPSS
10.1) ensures GLM runs over ten times faster than in previous versions
of SPSS, especially when using it with large designs.
- Build predictive models to target customers, students or citizens --
even when the outcomes of interest are missing -- using MLR, PLUM or
GLM. These procedures now enable you to save the predicted value even if
the outcome is missing, as long as cases are complete on the predictors.
- Enhancements to Categorical Regression (in SPSS Categories) make the
procedure more powerful and easier to use:
- Easily work with continuous-level data using a method that
categorizes variables
- Guarantee smoother transformations using new transformations
(monotonic and nonmonotonic spline transformations of the variables,
with a pre-chosen degree and number of knots for each variable,
separately)
- Get more flexibility when data are missing using new options for
missing data.
- Fit predicted values for supplementary objects using regression
and optimal scaling results for "active" objects
- To visualize how well the data are fit with new plots displaying
the optimal, you can use nonlinear transformations with respect to the
residuals in the regression model
- To more easily examine how well a model fits the data, there is a
new design of tables for fit and model parameters: a new ANOVA table
with degrees of freedom according to optimal scaling level; a new
model summary table with adjusted RSQ for optimal scaling; t-values
for replacing F-statistics; significance levels; and a separate table
with zero-order, part and partial correlation, the importance and
tolerance before and after transformation
- Get more information in the correlation tables by using
eigenvalues
- Easily inspect output in the newly designed output tables
- Use long strings as input for greater flexibility
- Specify the treatment of missing values for more flexibility and
control
- Have greater flexibility when specifying models by no longer
having to specify the number of categories
- Save predicted scores and residuals to better target customers,
students or citizens
- Have more options for dating file saving by saving discretized
data
- Enhancements to Categorical Principal Components Analysis (in SPSS
Categories) make results easier to understand:
- Understand results more clearly with full output for categories
that occur only for supplementary objects, especially centroids that
classify supplementary objects into groups
- Have maximum flexibility in creating bi-plots using variable lists
for the LOADING keyword in the PLOT subcommand (one for loadings and
one for centroids)
- Understand one-dimensional solutions more easily using Plots for
one-dimensional solutions
New features added in SPSS Base 11.0 for Windows
Data
management
- Restructure your data easily to get it ready for analysis in
procedures, such as Linear Mixed Models, MANOVA, GLM and Cox Regression,
using the Restructure Data procedure. Using this feature, it is possible
to:
- Take a data file with multiple cases per subject and restructure
the data to put all data for each subject into a single record. This
is sometimes referred to as restructuring a data file from a
univariate form to multivariate.
- Restructure in the other direction. This means you take a data
file with a single case per subject and spread data across multiple
cases -- referred to as going from multivariate to univariate.
- Prevent data from accidental destruction by saving your data as
read-only
- Easily manage the contents of your SPSS data files with the ability
to drop and keep variables when saving using the Graphic Users
Interface
Data access
- Improvements to the Database Wizard make querying from databases
easier and more useful:
- Get database queries that immediately produce useful variables in
SPSS. The Database Wizard now allows you to recode categorical string
variables -- the result will be integers with value labels.
- Get more useful results when joining tables using the Database
Wizard. The logic used in auto-joining tables uses primary/foreign key
relationships.
- Run subsequent database queries using the Production Facility. The
syntax pasted from the Database Wizard can be used seamlessly in the
Production Facility.
- Know the status of your query using the progress indicator
- Easily inspect your data after performing a database query with
improved appearance in the Data Editor. The appearance of the Data
Editor will be restricted to 50 columns after a database query.
Additionally, a minimum value of seven columns is required for the
Data Editor. (Note these requirements have no affect on the actual
string width.)
- Analyze databases more efficiently by sampling the database when
doing an ODBC pull
- Easily read CSV-format text data (such as, "1," "5," ...) by
specifying text qualifiers from delimiters in the Text Wizard
- Get easier connectivity to Microsoft Data Access pack (especially
Microsoft Access)
- Get connectivity to Sybase 11 and 12; Infomix 7.3+, 9.14; Infomix
2000 (9.20); UDB (DB2 6.1 and 7.1); SQL Server 2000; Oracle 8.06;
Oraclei Releases 2 and 3 (8.1.6, 8.1.7)
- Read current versions of SAS data files and SAS portable files
Statistics
- Quickly size up the change in mean or sum between any two related
variables with percent change. For example, you can easily see how sales
increase from quarter to quarter.
- Get a better handle on your data using descriptive ratio stats:
- Coefficient of dispersion
- Coefficient of variation
- Price-related differential
- Average absolute deviance
- When you have outliers that might skew your data, summarize
aggregated variables using medians for a less biased measure of central
tendency
- Get more flexibility when creating customized models with density
functions for continuous and discrete distributions
- Get more accurate results with ONEWAY when the number of people (or
variances) in each group varies using Brown-Forsythe and Welch tests
- Get improved scalability and performance for Proximities and
Hierarchical Cluster Analysis (added in SPSS 10.1). The speed increase
is five to 50 times faster, depending on the type of computations made.
- Enhancement made to the Shapiro-Wilk's test of normality in EXAMINE
(added in SPSS 10.1) ensures EXAMINE can now handle up to 5000 cases
when weights are not specified.
- Export the Factor Score Coefficients using the MATRIX IN/OUT
subcommand in FACTOR (added in SPSS 10.1). This is can be done using
syntax only.
Graphical and output enhancements
- Save time by getting better-looking output using defaults; this
reduces the need to manually edit your output ad hoc
- Get output in a familiar, easy-to understand format -- rather than
deciphering scientific notation
- Improved the appearance of ASCII output in the Draft Viewer (e.g.,
wide output is no longer truncated)
General usability
- Get easier and more reliable scripting with support for a more
recent version of SAX Basic
Compare SPSS 11.0 for Windows to other versions