1. OMR Scanner VS Image Only Scanners with OCR Software
1. Image Only Scannerswith OCR Software cannot do OMR Mark
Discrimination since they don't use ScanTools Plus to process the
exams.
2. The image only scanners with the OCR Software solution would
use a pixel count approach for OMR that will create errors when
erasures are on the sheet. This approach sets a level of 'fill' for the
bubble, say 10% which would be interpreted as a 'hit' for the
answer. If there is an erasure on the sheet, and the pixel count for
that bubble is 10%, they have a double 'hit', so this results in an
error. To overcome this inaccuracy, their solution is to send all
multiples to the editing operators causing a lot more editing labor,
not to mention to the inaccurate information.
3. Because of using image scanners, so they do everything postscanning which means OMR as well as Constructed Response is
processed from an Image File (this is NOT true OMR), but we do
true OMR in real-time for accuracy and speed.
4. They don't plan to use drop-out colors on the forms to save cost,
but with no drop out color for the answer bubbles, there is less
accuracy because the system may conclude the non-drop out
printing (black) is a mark, causing additional errors. In high stakes
situations, this erroneous information would be cause for high
visibility (as in TV and Newspaper) exposure.
5. If, during scan, the document has a fold-over or double sheet
feed, they will not know this during the scanning process, unlike
our system which shuts down should these faults happen. This
means that they will have to locate the mis-scanned document,
which could be scanned days before, and re-scan causing lost
time, extra labor and decreased accuracy.

2. 6. In addition, they risk skewing of the sheets as they are fed through
the image only scanners. This will cause significant errors. The
feed mechanism of the image only scanners is not the same as the
feed mechanism of the OMR scanners.
We have spent 2 years changing the feeding mechanism design to
eliminate skewing of the sheets, resulting in very high accuracy.
No image only scanners have our proprietary feed design and the
mechanism used in image only scanners can cause skewing errors,
up to 7% in some cases, resulting in corrupt data and enormous
editing time.
7. The feed rollers used in image only scanners will transfer pencil
graphite from the sheet being scanned to subsequent sheets
which can easily be read as a mark. Scantron Engineering changed
feed roller material and density to eliminate this graphite transfer
problem. As you can imagine, when pencil graphite is transferred
to the following sheets, and read as a mark when there really is no
mark intended, high stakes testing results are at serious risk.
8. The image only scanner doesn't know if there is an error until 3
sheets after the sheet is fed - much too late to do anything about
the error due to scanner buffering design.
9. In image only scanners there can be no information about the
specific sheet being processed because the mechanical distance
between the read station (or read head) and the printer is less
than 2 inches which does not allow time for real time processing
and then to print on the exam form as it goes through the
scanner. We validate the scanning process for each sheet by
printing on them during scan eliminating double scanning, and this
also documents that the sheet has been scanned for legal
requirements in high stakes tracking or validation situations.

3. Accuracy
Speed
White Space around
bubble
Timing Marks
mark discrimination
Easy to use
Capture Data
sheets Skewing
OMR Scanners
Consistently 99.9+ %
Up to 250
pages/minute
form can have the
bubbles very close
together, making it
possible to collect
more data in the same
space.
Need Timing marks
Image Only Scanners
Up to 99.9% with
editing
Up to 120
pages/minute
form needs to have
more “white space”
around the bubbles.
Don’t need Timing
mark
can perform mark
cannot perform mark
discrimination,
discrimination.
thereby being able to Multiple
determine erasures
marks in an area,
from darker marks
when not allowed, will
be flagged as potential
problems.
More easy and can be Less easy and should
left unattended
have operator to take
decision about
potential problems
Can capture much
Capture less data in
more data on an OMR
one form coz you
form reducing paper
should leave more
costs.
“white space” around
the bubbles.
very high accuracy by
Less accuracy and
eliminate skewing
cause significant errors