Like a piece of machinery, a brain can be given a tune up. Various performance metrics (SAT and IQ tests) of intelligence are affected by my mood, sleep habits, and nervousness. And various activities (such as listening to musical pieces by Mozart or doing math exercises) can also affect the results of such tests.
I first realized this when I was in preschool. Several graduate students in the psychology department at UCLA got me into a program for studying child development and related intelligence issues. For about a year I'd visit them pretty much on a weekly basis so that they could study me. They constantly measured my intelligence quotient and monitored the development of various motor skills. And I was in a group of about a dozen or so children, all of whom were being compared against me. These researchers were always taking my mood, my sleep habits, my health, and what I had been doing lately into account along with my intelligence measurements.
As it turned out, a nervous little 4 year old kid suffering from allergies and chronic migraine headaches such as myself could be dropped into a lab full of strangers and only score in the 125-135 range on an IQ test on one day. And then a few months later with soothed nerves, some familiarity with the other people in the lab, and when not suffering from headaches and sneezing fits, the same kid could end up scoring in the 165-175 range on average. The agony associated with having a cold or other illness could put a spike on the graph near 120 on a bad day. And an exceptionally good night's rest and delicious breakfast could result in a spike near 200 the week after that. That's exactly how I fared on those tests, and I wasn't unique within the program.
More statistical noise is introduced due to the raw difficulty in quantifying intelligence. And people in what is supposedly the genius category are apparently prone to some of the widest variances, due in no small part to the ways in which the metrics themselves tend to be tuned to focus most of their precision on the task of measuring people within only one standard deviation from average intelligence. (Or, depending on the intended purpose of the tests, they might focus more precision on different ranges.)
And how many of us have encountered questions on an IQ test that had two correct answers, yet only one answer deemed "the best answer" actually scores in your favor? Worse yet, the determination of the best answer can be culturally biased. I've even taken some IQ tests where marking multiple answers and diverging from the instructions of the test was actually the right way to maximize my score.
But I'm going off on a tangent. Anyways, there are two basic things involved in tuning my brain. The first is my overall emotional state, and the second is to exercise specific neural pathways.
For the emotional part: being happy, stress free, and having fun all help. And being optimistic and hopeful also helps. But really, it's all just about being relaxed. Stress and nervousness clutters up my mind and affect my ability to focus on the problems I'm solving or the questions I'm answering. A relaxed, confident attitude is best. It's sort of like when you take the SAT multiple times. The first time you are generally more nervous and concerned about how well you'll do, and consequently students end up performing better their second or third times when this stress is reduced.
And to exercise specific parts of your brain, well that should be obvious. It's like studying in school. Or solving puzzles. It was proven in one study I'm aware of that playing Tetris increases one's score on an IQ test. Likewise, paying attention to a piece of music written by Mozart will also exercise certain parts of your brain and tune it up. Basically you've got to just practice solving certain types of problems or thinking about certain kinds of things, and over time your brain will improve in those areas.
Anyways, if you have a fragile ego and care about such things like SAT scores of 1400 to try and validate yourself to the world, then you might want to apply my observation to your own brain to try helping yourself out. It's all any of us can really do, I suppose, until we can start overclocking our brains. (By the way, my score is above 1400. Does that mean I get a cookie, or what?)