|
|
|
|

Webcast Audio Seminar Series
“Educational Technology Toolkit: A Consumer’s Guide”
|
| |
|
Moving from Paper-Pencil to
Electronic Exams: What it Takes to Get it Done -
April 25, 2006, 12:00 pm:
Presented by Scott
Elliott
University of Iowa
|
|
Description
In this
presentation, Scott Elliott will discuss the benefits
and drawbacks of computer-based testing (CBT), and what
it takes to make CBT successful. Computer-based testing
has potentially wide applications in undergraduate,
graduate and continuing medical education. An early
step in evaluating CBT is to be sure that the exam
format is measuring the examinees’ knowledge and not
their comfort level with the technology. It is,
therefore, important that the CBT reproduce or
accommodate traditional test-taking behavior. CBT also
provides for a number of enhancements not easily
achieved with traditional paper and pencil exams. These
include easier control and editing of exam items using
and itembank, better incorporation of testing into the
learning environment using specific feedback, and
enhancing exam questions by incorporating multimedia.
In addition to traditional multiple choice and
true-false questions, CBT can also offer different item
styles including simulations, intelligent continuation,
and long-menu, non-cueing question types. Storing exam
items (itembanking) in intelligent databases is becoming
more popular. Itembanks exist to help get exam items
organized, hierarchically and statistically, and to
eliminate “out-dated” and “haphazard” storage methods of
the past. Trends in itembanking now include features
that track item and exam performance and provide a
statistical prediction as to how an exam might perform.
Statistical analysis of exams is very important to
ensure acceptable item performance and exam reliability
and validity. As for resources, knowledgeable people
who are committed to the success of the CBT program are
critical. Many different skills are necessary for
successful CBT deployment, including database
technology, instructional design, statistical expertise,
and computer and networking expertise.
|
|
 |
|
|
 |