Status: Urban Legend
Here's an odd urban legend that I just stumbled across. Supposedly if you smear chapstick down the side of a scantron sheet (the kind used for standardized tests such as the SAT), the grading machine will mark all your answers correct. The theory is that the chapstick will interfere with the scanning light, confusing it into thinking that your answers are correct. Needless to say, this doesn't work.
Some guy named Richard Mangahas has written a short article
detailing all kinds of theories about ways to cheat on scantron tests, including: marking or deleting the black lines along the side of the page, filling in the bubbles with cross-hatches, or placing tape along the side of the page. I don't think any of these methods would work either. (Though Mangahas claims some of them work 25-30% of the time... which is about the same percentage you would expect from guessing at the right answer.)
Maybe it was kids armed with chapstick that caused all those SAT-test score errors recently.
If a dose of chapstick did anything at all, it'd probably just mark all the answers as 'incorrectly filled out', which would probably send it off into the bin to be hand-checked. Worse, it'd probably just mark everything as false positives, resulting in an abysmal score.
If someone really believed this, then how did they ever get into college?
All these test tricks remind me of all the breathalizer cheats. Maybe you can get the scantron to mark your answers all correct if you take the test with a battery in your mouth.
First, he says method #1 is the best and recommended method of cheating, but then he says don't use it on every answer. If this is the method which has the highest success, why not use it on all answers?
Second, this is purely conjecture. But wouldn't the machine be programmed to fail on error, rather than pass? If the scanner hits an error, it should either flag it for manual correction, or flag it failed.
I am sure that the people who design these machines have considered cheating, and put in appropriate checks. It's like an ATM - if you cause the ATM to fail, it's not going to spew out $100 bills.
Yeah, they hand-check tests that have problems. I knew a kid who purposefully Christmas-treed the whole test & his score was, "N/A". (To Christmas-tree is to make a pattern or image using your bubbles. Apparently his image/pattern was "YOU SUCK". He was a funny kid...I dunno if they ever let him re-take it.
You mean I don't have to give them back, then?
I did not recommend using the first method in every answer because I also received false negatives (correct answer marked wrong) and they appeared at a higher pace than false positive (incorrect answer marked correctly), which is the desired outcome.
The "failing" on error was the desired outcome because in my experimentation (which I would like to stress are machines I never had physical access to) a failure appeared to lead to non-marking of the particular answer. As for complete machine behavior, never having had my hands on a manual or capabilities to manipulate various settings on these machines, I cannot say much more than I have written.
Ultimately, the original text file was published because I wished to debunk chapstick as a method (which was popularly spread at that time).
And as a final note, I have received maybe about 50 emails over the course of the past 10 years. About half of them just say "I don't think they would work." And I would like to point out that I hear "don't THINK it would work" more than "didn't work for me."
The other half would either say it worked for them or not. I would say about 10 of the emails backed up my claims and just asked if I had anything with better chances. And one person even claimed to have mastered the first method to the point where it was 50%. So, that's about a 20% success rate on emails I've received.
I've heard enough people call it BS, and I just don't care anymore. I have a real job and have not touched a scantron in over 5 years. I just thought I could address some of the points.
The programs that score the sheets look for correct answers, not incorrect answers. If they looked for incorrect answers then students turning in a totally blank answer sheet would receive 100% and they don't.
on the bubble sheet, simply answer in this order:
C-A-B-B-A-C-C-A-B-B-A-C and repeat.
you'll be a genuine pretend genious.
(im not even jokin either im serious it works)you will get +50% easy
but just in case, u better study...
(PS. Dont break the chain of cabbaccabbac)
I ended up getting a 34% and failing the class by just 1%
THERE ARE NO PATTERN OR A WAY TO TRICK HUMAN AND ROBOTIC RESOURCES
SO ITS TIME TO CRACK OPEN THAT TEXTBOOK UNDER THE HALF-EATEN SANDWICH AND READ UR MATERIALS
THE ONLY FASTEST WAY TO A GOOD GRADE IN TESTS IS ONLY IF YOU STRATEGICALLY STUDY....
-results may varie...... for jocks- >=D
for every answer.
it's guaranteed to get you a 60% or higher.
i have done it before and i got a 64%
Reading this thread, it seems that students no longer understand that it is their job to learn the material in their classes. . .
Don't throw away your chapstick - it's great in the winter. Just don't waste it by putting it on your Scantron forms.
The best way to do well on a Scantron scored test? STUDY!