Well, the right question to ask is whether you should attempt to crack the algorithm because, as of yet, there is no real proof that the algorithm can be cracked for your benefit.
Let me introduce the matter to those who have not thought or heard about it. As you all know, the Digital SAT is a CAT – or a computer adaptive test. Every test-taker has to navigate 2 verbal and 2 quantitative modules. Both the first verbal and the first quantitative modules contain base-level questions. Only if you answer a good number of questions correctly, in any of these sections will you get higher-level questions in the second section of the same type. Obviously, these tougher questions carry greater credit. So doing well in a later section with tougher questions gets you a much higher total score than doing well in a later section with base-level questions does. But how do the candidates proceed if they do well enough in the earlier modules to get the tougher questions in the later sections but then answer most of those questions incorrectly? Would you get a better total score if you do not answer enough questions correctly in the initial sections, get base-level questions in the later sections, and answer all or almost all of them correctly?
It’s an age-old dilemma faced in many walks of life: play aggressively or play defensively? If you up the stakes, you either win big or lose big. On the other hand, if you aim small, you miss small.
The matter is aggravated by the fact that the College Board does not disclose its scoring algorithm. It has never done so. The SAT has changed its avatar many times, but College Board never disclosed the exact scoring algorithm for any of those avatars. Not even for the discarded avatars.
The lack of authenticated information pokes everyone’s imagination into a state of hyperactivity in every instance, and a lot of students have inspired themselves or others to believe that doing exceptionally well in the later sections with base-level questions can get them a better score than doing sufficiently well in the initial sections and poorly in the later sections can.
To end the conundrum, one of my colleagues took all four bluebook tests separately. I tried to do my best in all the sections, and my colleague intentionally flunked the initial sections and did his best in the later sections. What we found was quite revealing: even if you answer all the questions correctly in two later sections with base-level questions – my colleague did answer every single question correctly in those two sections – you do not cross 1400. On the other hand, although I intentionally marked the wrong answer in about 40% of the questions in later sections with harder questions, I crossed 1400 on all the tests.
Of course, this was not an experiment that checked out all the possibilities. What would have happened if I had marked 60% of the harder questions incorrectly? I cannot tell because I did not. However, it is very unlikely that a student capable of answering about 80% of the base-level questions would go wrong in 60% of the tougher questions.
Therefore, for all practical purposes, it may be useless to even try cracking the system. In fact, it may actually hurt your prospects by preventing the full utilisation of your potential.
By Arijit Thakur, Senior Trainer
