|
|
|
PASS
|
|
Analysis of Variance (ANOVA) Introduction
A fundamental tool in hypothesis testing is the F-test for testing the
differences among several group means. It may also be used for analyzing
more complex experimental designs. You specify an effect size, standard
deviation, alpha, beta (or power), number of factor categories, and
sample size. The program calculates the power using the non-central F
distribution. What if analyses let you solve for any of the
parameters.
PASS includes power analysis and sample size calculations for
several experimental designs, including:
One-Way Design
Used when there is only one factor. For example, the factor might be
temperature or dose level. Individual planned comparisons may also be
entered and analyzed.
Factorial Design
Used when you have up to three factors in a factorial experimental
design. This routine may also be used to analyze special designs such as
the Latin Square design.
MANOVA
Used when want to analyze a design with up three factors and multiple
responses per subject.
Randomized Block Design
Used when you have a randomized block design with up to two factors.
Multiple Comparisons
Used when you want to use multiple comparisons. Several type of multiple
comparisons can be analyzed including pair-wise, treatments versus
control, and multiple contrast.
Repeated Measures Design
Used when each subject is measured at least twice. The subjects may have
been separated into groups. The analysis allows for the
Geisser-Greenhouse correction to be taken into consideration. It allows
you to specify various types of covariance patterns. This version allows
for up to three grouping factors and up to three within-subject factors.
It includes analysis for the F-test, Wilks Lambda, Hotelling-Lawley
trace, and Pillai-Bartlett trace.
Simulation
Using Monte Carlo simulation, the power and precision of a test
procedure can be evaluated when the data do not followed the normal
distribution. This lets you investigate the impact of outliers,
skewness, and other types of non-normality on your design.
[Back to PASS]
[Back to
PASS
Procedure List]
|
|
Download a 7-day Trial Copy of PASS 11
Use Evaluation
Serial Number "TRYPASS11" |
Place an Order Using Our
Secure Online Order Form
 |
|
|
|
|
PASS 11
Now Available
We are pleased to
announce the availability of PASS 11. This edition
adds over 17 new procedures and features to PASS.
Click here for pricing.
Read about what’s new in
PASS 11.
|
"Researchers seeking software to determine sample size have a wide
range of choices... Researchers seeking a comprehensive,
commercial program for analysis of statistical power and precision should
consider PASS 2008."
- Patrick Dattalo
"A Review of Software for Sample Size Determination"
Eval Health Prof, 2009.
|
Click here
for the complete list of procedures.
Download a
PASS Update
Most Recent
Version:
PASS 11
Version 11.0.8
Released December 2, 2011
What problems have been fixed by
recent PASS 11 updates?
PASS Training Course
A 4-week, online course on sample size and power, featuring PASS, is now being offered through
statistics.com. The course is
taught by Dr. Thomas P. Ryan.
Click
here for more information.
|
|
|
|
|
Minimum
System Requirements
Click on one of the links below to view the minimum system
requirements for NCSS and PASS products:
PASS 11
PASS 2008
NCSS 2007
Previous Versions |
Accuracy
We at NCSS have put a
great deal of effort into finding the most
accurate algorithms possible. The programs have
been tested and verified over and over, both by
us and by our customers. Each routine has been
verified against textbooks, journal articles,
and, where possible, other software. NCSS and PASS are some
of the most accurate statistical analysis
programs available. NCSS and PASS calculate with
seventeen-digit, double-precision accuracy. |
Guarantee
If you are not completely
satisfied with an NCSS product during the first
30 days for any reason, return the program for a
full, prompt refund (excluding shipping). |
|