1. What do the default values (the variable
values already chosen by BRAINS) mean?
2. What does BRAINS actually predict?
3. Why can't I make comparisons across regions?
If you have other questions about BRAINS,
you might also want to read about BRAINS
or read the using BRAINS section.
1. What do the default values (the variable
values already chosen by BRAINS) mean?
Each default value represents either the most common occurrence
for variables that allow you to choose an option, (e.g., boat
type); or the average value for variables that allow you to insert
a number (e.g., wind speed, boat length) for that specific variable
independent of any other variable values.
For example, assume that in your accident model the default
boat type is Personal Watercraft (PWC), and the default boat
length is 30 feet. This means that PWC is the most common type
of boat involved in this type of accident and that the average
length of all boats involved in this type of accident is 30 feet.
Importantly, the default values here do not mean that the
average or most common boat involved in this accident type is
a 30-foot PWC.
2. What does BRAINS actually predict?
The BRAINS system uses the statistical relationships between accident report variables within the U.S. Coast Guard's Boating Accident Report Database (BARD). These relationships allow the BRAINS user to determine the probability that a specific accident could occur given the conditions that are entered by the user. BRAINS enables the user to isolate the effect of changes in a specific variable on the likelihood of a specific type of accident occurring.
In addition to allowing the user to predict the likelihood
of an accident occurring under common conditions, the flexibility
of BRAINS allows the user to predict the likelihood of an accident
under uncommon or hypothetical conditions.
For example, the average length of a Personal Watercraft (PWC)
in BARD is approximately nine feet. By inserting this combination
of values, BRAINS will provide the user with the likelihood of
a specific type of accident occurring under those typical conditions.
However, what if the PWC where longer, let's say 15 feet?
BRAINS gives you the flexibility to construct this hypothetical
scenario and predict the change in accident probabilities due
to an increase in the length of PWC. Although this flexibility
has the potential to increases the power and usefulness of the
system, the user needs to understand that the accuracy of any
prediction becomes less reliable as the variable values become
less realistic.
NOTE:
For a specific example about how BRAINS work, a simple case study is presented
in About BRAINS.
3. Why can't I make comparisons across
regions?
It is difficult to make direct comparisons across regions
when using any of the models because appropriate accident variables
may differ across regions.
