your questions indicate a misunderstanding of the concept
studying skews the results, it's not supposed to be one of those things you study for <think urine test, just... different (and, yes, it is comparable, think of people who try all kinds of crazy **** to drive drug residues out of their kidneys)>
two points is within the margin of error, so that is a completly moot question, actually the farsical data that "on average habitual marijuana use causes a four i.q. point loss" god I would kill to remember which journal I saw that in, it's simply foam
the tests are not randomized, but nor are two given in sequence identical, however the variation is very nearly statistically meaningless, I say very nealy because, well, boom, margin of error, that is one of the contributing factors
different areas, well that is a fun one
there a few different "multiple intelligence" theories running around, this, yes, does just measure the classical idea of intelligence, hardly brushes by creativity and doesn't even acknowledge kinesthetic intelligence
however
if you look at the results of the gardner multiple intelligence tests the results do not form aggregate bell shaped curves, as they should, further, the questions are exclusionary
i.e. there are questions like
"do you prefer staying indoors painting or going outside to play ball" I would actually give you one of the questions verbatim as opposed to my poor paraphrase, they are just about that heinous, but my book is in my car and it's damn cold outside
now, think about that for a second
if you cannot score perfect, and you cannot score zero, how can the test provide accurate results?
now stop that knee jerk hippie "peace and flowers everyone is equal" reaction
the thing is
that test is broken, as are all of the other multiple intelligences tests because they use the same premise, if a person with absolutly no capacity could answer the questions they could still score just about average across the board on one of those because it's all touchie feely bull****
I.Q. variation over age, well, see, that's the fun thing about it, the most a persons i.q. is ever supposed to vary over their entire life span is fifteen points (if it varies more either something dramatic happened physiologically or the test was performed wrong)
I however am on the outs statistically even with that, most people don't change at all
a flat aquired\obtained knowledge test score is going to change as someone ages, as are virtually all other measures of personality, capability, and aptitude, granted the test does scale to age, it is supposed to, that is how it works
so many people detract from the value of the thing without having any beginning of an understanding what the hell the thing is to begin with