|
The Advanced Statistics Module of StatPac gives you more
power and control than any other statistical analysis software on
the market today. It provides state-of-the-art sophistication and
capabilities. Whether you're an experienced statistician
or a beginner, StatPac is the solution.
The Advanced
Statistics Module contains:
Analysis of Variance
Analysis of variance is used to compare variances from
more than two groups. The ANOVA program in StatPac offers
impressive versatility. There are eleven ANOVA models to
handle practically any kind of experiment you design
(e.g., repeated-measures, split-plot, randomized block,
complete block, nested and Latin square models). The
output selection includes an analysis of means, classical
ANOVA table and post-hoc least significant difference
t-tests. The Kruskal-Wallis test is available for
non-parametric analysis.
Linear and Non-Linear Regression
Regression analysis in StatPac is sophisticated and
remarkably easy to use. An automatic curve-fitting option
makes non-linear regression an effortless procedure.
Another special robust technique is available to reduce
distortion caused by extreme data points. StatPac can
produce a rich variety of tables for regression,
autocorrelation and residual analysis.
Stepwise Multiple Regression
The multiple regression program in StatPac has been
given top ratings by reviewers for its speed, accuracy
and completeness. The stepwise method is forward
inclusion with backward elimination. Output includes all
the regression statistics and matrices. You can even
switch to interactive prediction to try the regression
equation on new data, or save the model for future use.
Probit and Logistic Regression
Probit and logistic regression are similar to multiple
regression except they are used when the dependent
variable is dichotomous (can take on only two values). A
banker might use these methods to determine the
probability that a person will pay back a loan, or a
medical researcher might use them to determine the
probability that an experimental drug would be
successful. Both techniques use accurate non-linear
algorithms.
Canonical Correlation
Canonical correlation is a powerful multivariate
technique to study the intercorrelational structure
between two sets of variables. One set is usually
regarded as dependent and the other as independent. For
example, a set of "buying behavior" variables
might be considered dependent, while a set of
"personality characteristics" variables could
be thought of as independent. Canonical correlation
provides a convenient way to understand the complex
relationships that might exist between the variables.
Principal Components Analysis
Principal components analysis is often used in
conjunction with multiple regression in an attempt to
reduce the number of predictor variables. This is
important because it helps to reduce future data
collection costs. Usually, most of the variation in a
large group of variables can be captured with only a few
principal components. StatPac also contains a complete
selection of collinearity diagnostics that measure
relationships between predictor variables and how they
affect the stability and variance of the regression
coefficients.
Factor Analysis
Factor analysis is used to identify and group
variables by their common dimensions. It is often used
with newly designed questionnaires to examine the
cohesiveness of variables. The factor analysis program in
StatPac picks up where others leave off. There are two
methods of extraction, three types of rotation and
several different ways to control the exit criteria.
Every parameter is adjustable to give you complete
control of the analysis.
Cluster Analysis
Cluster analysis is used to identify and group
respondents that are similar. It is frequently used in
marketing research to identify and target segments of the
population for an advertising campaign. StatPac contains
six outstanding clustering techniques. A hierarchical
tree diagram provides a visual summary that makes it easy
to identify the clusters. The cluster membership can be
saved for inclusion in additional analyses.
Stepwise Discriminant Function Analysis
Discriminant function analysis is used to predict a
categorical variable. Marketing researchers often use
this procedure to understand the factors that determine
why consumers choose one brand over another. StatPac
offers a complete output selection including canonical
variable analysis.
Perceptual Mapping
Multiple correspondence analysis (perceptual mapping)
is a very powerful and easy to use technique for studying
the relationships between two or more categorical
variables. It is frequently used in marketing research to
understand consumer perceptions of a product and to
determine the effectiveness of an advertising campaign
designed to modify their perceptions.
|
