Wednesday, March 31, 2010

Actual Scantron Exploits

Chapstick don't work. Face it, fiddling with the Scantron form is going to get noticed, especially if it makes it sticky. And guess what? It won't spoof the reader, either.

Light lines across the bubbles, or shading? OK, now you're just rolling dice on whether the scoring machine is going to pick up a right answer off one of those--and not a wrong one, or a double-mark (insta-wrong!) You're better off with a blind guess!

Cutting the form and other such nonsense will just get you into academic probation.

What to do to beat the Scantron?

Before The Test: The Ultimate Exploit

Pre-test prep is crucial when you're graded on the curve. Spread crap about the Chapstick and form-spoiling exploits among your classmates. Send them to websites about it, hand out chapsticks and exactos and colored pens before the test, show them a (dummy) hacked form of your own and tell them how cool it is. Make crap up about sucking a Mentos and licking the form, or spilling green tea then drying it in the oven. Whatever.

Then, when their Scantrons get bounced and flagged, you'll be golden!

Now, let's get real...

Play the Test, NOT the Tech

First, don't try phreaking the Scantron form reader unless you're the one working on it and running the tests through. Different readers work differently, and most spoofs are going to either get the form spit out, get the form marked wrong, get ignored, or hose the machine.

None of these gets you what you want, and two of them put your spoofed Scantron sheet in a human's hands where trouble starts. Not good.

So, first, don't eff up the form. When you know an answer, don't screw up marking it. Use the right pencil and mark it properly. Exploits are worth nada if they take away what you can do right (whatever that may be.)

Working the tech is asking for trouble. It leaves a trail marked "this way to loser." So...

A scantron is just a multiple choice test. MC tests have been getting worked since they've been given. Scantron neither detects nor corrects these techniques. They leave no tell-tale marks on paper. And they can turn failure into hope for survival (which you don't have otherwise, or you wouldn't be looking up hairbrain schemes for putting lip wax on your test form on the internet, right?)

First, how's the test scored? If you don't know this, learn it or drop the class. Are blanks ignored or marked wrong? Does a blank count against you more than a wrong marked answer? How many points do you get/lose for each right/wrong/blank?

Let's say blanks are worse than wrong marked answers. You are now a wild guessing machine--no answers goes blank. See below for cutting down on the wrong guesses.

Wrongly marked answers are worse than blanks? Mark nothing on the Scantron that you either don't know or haven't got a strong guess on--keep reading to fill in as many bubbles as you can...

Exploit the test as follows:
Go through the whole test fast, answer only what you're sure of. Often the answers for some questions are built into others.

Watch for tricky words "are not often mainly" "less regularly than at first" and garbage like that can wait. Skip it and do fine parsing later (if at all.)

Go back, picking up anything that a later question answered. Get the marks in the right freaking bubble! Do not get half the test off by one then discover it too late!

It might seem risky to bounce around the test, rather than going top to bottom one by one, but really your chances of picking up an off-by-one bubble row shift is better this way, and you'll affect less of the test if you do it. Few things such as bad as having to move 300 test bubbles up one when you're moments away from "pencils down".

How many possible answers are there on the questions you have left? If it's two it's worth guessing, especially if you've got even the faintest clue of which seems better.

If it's more, throw out the howlers. If you can get it to two, you're in guessing territory even for a wrong mark is worse than a blank. Once you've done this as far as you can, sweat your way through as many blanks as you can in the remaining time.

If a blank is worse than a mark, get it to three possibles and guess. Then finish off with your totally random shots in the dark. Flag the total guesses, and use any extra time to go back and see if you can eliminate any of the answers you've marked as wrong. When you do, pick another of what's left.