On the Wechsler and Stanford-Binet scales, a score of 62 is 2.533 standard deviations away from the mean and lies below 99.4% of those who take the test. From the evidence posted by
@Vixen, it appears someone miscalculated the 99.4% as 99.9%. When compared to most of the other mistakes
@Vixen has posted within this thread, that's a minor error.
And reporting 99.4% as 99.9% is only a small exaggeration, when compared to most of the others.
As for the 158, I suspect someone got the leading digit wrong, or perhaps the second photo is reporting someone else's test results. I note that the superimposed signatures ("Vixen") are identical, which means they were added after the photograph was taken, which means the evidentiary value of those photographs is nil.
The highlighted score is
@JayUtah's mistake.
The Cattell Culture Fair Intelligence Test (CFIQ) maxes out at 152, so
it is not even possible to score 162 (or 158!) on the Cattell scale. But a score of 62 is not only possible, but entirely plausible.
(Edited to add a second citation for the fact that the CFIQ maxes out at 152. While I'm at it, I'll mention another possibility: The 62 might have been a raw score of 62%, which would then have to be converted to the scale of a normal distribution with μ=100 and σ=24 or σ=16. If that conversion was done by a member of Mensa, that conversion might have been done incorrectly.)
If the 62 was her score on the Cattell scale, then her score was only 1.583 or 2.375 standard deviations from the mean, depending on whether she took the original version of the Cattell test (with standard deviation of 24) or the more recent version (with standard deviation of 16). So her score lies outside only 94.43% or 99.1% of the population, depending on which test she took.
It seems, therefore, that she is only a single-niner or double-niner, depending on whether her score of 62 was for a standard deviation of σ=24 or σ=16.
That she is only a single-niner or double-niner is far more plausible. A true triple-niner, more than three standard deviations from the mean, would probably have a hard time dealing with the sophistication of this forum's software.
Do we really need more rhetorical questions?