 Regression
Modeling SoftwareWhat
is StatModeler?
StatModeler is a new,
superior successor of Applied Analytic Systems' Wingress.
StatModeler is a
complete end-user application that allows anyone with a
basic knowledge of statistics and regression to 1) do
summary analysis of any database field, 2) perform two
mean and paired hypothesis testing and 3) to model and
conduct either a simple (single variable) or multivariate
regression analyses from any supported database.
The focus of StatModeler
is multiple curvilinear regression. Modeling and solving
a regression model with StatModeler is as simple as
opening a data file and selecting selecting the dependent
variable, selecting the independent variable(s), then
solving and viewing the regression plots and ANOVA
outputs. Predictions can be conducted based upon the
solution of the model.
Like WinGress,
StatModeler 2K3 was developed in Delphi (5) and it uses the powerful multiple regression analysis Delphi/C++ Builder components
from Applied Analytic Systems, and incorporates advanced
modeling capabilities including curvilinear, polynomial,
and response surface regression, and advanced model
refinement processes like stepwise & exhaustive
selection optimal regression. Note: StatModeler is not a programmer's tool. You do not need Delphi or any programming language to use it.
Multiple Regression
Modeling
Regression (simple or
multiple) is an inferential statistical analysis
methodology that uses general or least-squares
differential equation solutions to identify the
relationships between a dependent (response) variable and
one or more independent (control) variable(s). Analysts
primarily use regression for forecasting and prediction
modeling, using the equation coefficents that result from
the solution of the least squares system. It is not
necessary to understand the calculus of least squares to
use regression analysis effectively.
Some example
applications of multiple regression modeling
- CAMA (Computer
Assisted Mass Apraisal) Systems [real estate]
- Trend Analysis
[revenue and sales, stock market/investment]
- Revenue/Yield
(Perishable Inventory) Management [Airline,
Hospitality, Retail]
- Forecasting
[demand, sales, financial, time-series]
- Prediction [e.g.
Performance vs. Test Scores and Grades]
- Compliance [e.g.
equal employment opportunity or environmental
studies]
StatModeler
Application, Input
Datasets, and Output Reports
Analysts can
use Excel datasets and SPSS datasets with StatModeler.
StatModeler is easier to use than Excel.
StatModeler is a 32-bit
Windows executable application requiring Windows 95/98/ME
or NT/2000/XP and 64 Meg memory. It can analyze data
stored in StatModeler's own database file format as well
as dBase/Foxpro, Paradox, and Excel CVS file formats.
Outputs include full
ANOVA (analysis of variance) reports including coefficient values, r-squared, adjusted r-squared,
s-squared, F, t, and p-values, and charting / plotting
utilities which facilitates full residual analysis,
allowing analysts to validitate the model by testing the
basic assumptions of a regression model and its
appropriateness for the given dataset.
This program is ideal
for researchers and analysts with quick projects
that need to draw inferential conclusions and predictive
functions through regression.
StatModeler 2K3 Screen
Shots
Dataset
"Bank.xsm"
|
The model associated with this sample
dataset is one which allows the analyst to
predict, for example, an employee's current
salary based upon other factors like beginning
salary, educational level, age, job
description...
Of course you could use the same dataset to
create predictive models for other employee
variables. This dataset was initially used as
part of an EEOC study of fairness in hiring. |
System Requirements
- Windows 98, ME, 2000, XP
- 64 Meg Ram
- Data in dBase (.dbf), Excel
(.cvs), or Paradox (.db) format
Questions About StatModeler?
Purchase
Orders are processed by our SOHO
Software Affiliate.
Contact The Statistical Software
Group for multiple or
site license pricing not covered on this page. No credit card? For a Free one day Trial download, contact Mike Laing at laing@aasdt.com during regular business hours.
StatModeler Developers...
Applied Analytic
Systems, Inc
Carnegie Office Park
600 North Bell Ave
Bldg. 2, Suite 2700
Carnegie, PA 15106
Phone:
412.253.4160
FAX: 412.253.4161
AAS Web: www.aasdt.com
Supported database
formats are dBase, Access, FoxPro, Paradox, InterBase.
Paradox is a trademark of Corel Corporation. dBase and
Delphi are trademarks of Borland Software Corporation.
Foxpro, NTand XP are a trademarks of Microsoft
Corporation.
|
