|
nQuery
Advisor eTutor is a modular-based online training course,
with each of the 6 main modules focusing on a different
aspect of working with nQuery Advisor.
Introduction to Statistical Sample Size
Concepts >>
Case Study: Calculating Sample Size >>
Randomization >>
Application Examples >>
Plots >>
Determining a value for the standard
deviation >>
Introduction
to Statistical Sample Size Concepts
Determining the appropriate sample size for a study is
very important to the design of an investigation. The
sample size must be correctly chosen to allow a study
to arrive at valid conclusions. A study which is too small
may produce inconclusive results, while a study which
is too large will waste scarce resources.
This
section of nQuery Advisor eTutor presents a brief introduction
to sample size determination concepts for significance
testing of a null hypothesis (two-group t-test), confidence
intervals and non-inferiority or equivalence testing.
It includes sample output from nQuery Advisor to illustrate
the concepts under discussion and tests the users understanding
of these concepts with a series of multiple choice questions.
Case
Study: Calculating Sample Size
|
|
In
the case study section, nQuery Advisor eTutor takes an
up-close and very detailed look at how nQuery Advisor
is used to calculate sample size. It uses the example
study question;
"Does
a new drug significantly reduce anemia in elderly women
after hip fracture?"
And
answers the question
"What
sample size do I need for the planned significance tests?"
using seven easy to follow steps
nQuery
Advisor eTutor goes through each of these steps in detail,
demonstrating important features and allowing participants
to use an audio-guided simulation of nQuery Advisor as
they proceed through the course. User comprehension is
again tested using a series of multiple choice questions
interspersed through-out the presentation.
Randomization
functionality was introduced to nQuery Advisor in Version
6.0. This section of nQuery Advisor eTutor focuses on
using the randomization feature to generate both basic
and advanced randomization lists. It again guides the
user through each step in the creation of the randomization
list, explaining each selection in detail. Multiple choice
questions again ensure that the user comprehends the concepts
and remains engaged with the course.
This
section features 9 detailed examples of how nQuery Advisor
can be used for sample size and power calculation in different
research scenarios. The examples include:
DNA Microarray: Testing fold-change with fold-change threshold
Oncology: Survival analysis
Psychology: Two-way ANOVA
Genetics: One-way ANOVA with unequal n's
Epidemiology: Confidence interval for odds ratio
Cardiology: Confidence interval for interclass kappa
Pharmacology: Contrast in ANOVA
Dental Research: Non-inferiority tests for proportions
Bioequivalence: TOST for ratio of means in crossover study
Each
application example looks at typical sample size calculation
issues that might arise in that particular field of study
and goes on to show how nQuery Advisor can be used to
resolve these issues.
The
plots module shows how to get the most from nQuery Advisor's
plotting facilities. It demonstrates how various elements
of a plot can be edited and of course how the plots can
be incorporated into other programs for reporting purposes.
Determining
a value for the standard deviation
|
|
One
of the great features of the nQuery Advisor program is
its ability to assist a user in determining a value for
the standard deviation. This section of nQuery Advisor
eTutor shows the full extent of this very useful element
of the program. It shows how to use the Data Entry Standard
Deviation Calculator, how to estimate the standard deviation
from any one of 10 different options, and how to specify
the covariance matrix for a repeated measures design.
|
