[nemo]

To expand on Mondo's post...

Since being "ENFP" and "Gifted" obviously aren't independent, you'd have to use Bayes' Theorem.

E.g.

P(Gifted | ENFP) = P(ENFP|Gifted)*P(Gifted) / P(ENFP)

Because

P(ENFP and Gifted) = P(ENFP|Gifted)*P(Gifted).

Assuming that the probability of being gifted is 0.05, the probability of a student being gifted given that they are an ENFP would be

P(Gifted | ENFP) = (.1545)(0.05) / (0.076)
P(Gifted | ENFP) = 0.1016 = 10.16%

And for all the types, this gives:

P(Gifted | XXXX) = P(XXXX|Gifted) / P(XXXX) * P(Gifted)

If you notice, the term P(XXXX|Gifted) / P(XXXX) is exactly what's in Mondo's table.

So if you just take Mondo's table and multiply by the probability that any old student is in the gifted program, you get the probabilities that a person of type XXXX is "gifted", at least using data from CC's table and a source I found saying 5% of a school is in the gifted program...

P(Gifted | INTP) = 16.99%
P(Gifted | INTJ) = 14.37%
P(Gifted | INFP) = 13.38%
P(Gifted | INFJ) = 13.35%
P(Gifted | ENTP) = 11.61%
P(Gifted | ENFP) = 10.16%
P(Gifted | ENTJ) = 7.43%
P(Gifted | ENFJ) = 6.3%
P(Gifted | ISTJ) = 4.93%
P(Gifted | ISTP) = 3.88%
P(Gifted | ESTP) = 2.46%
P(Gifted | ISFJ) = 2%
P(Gifted | ISFP) = 1.99%
P(Gifted | ESFP) = 1.4%
P(Gifted | ESTJ) = 1.3%
P(Gifted | ESFJ) = 1.18%

Anyway, that's the basic interpretation of Mondo's table. You can adjust for different percentages of a school enrolled in a gifted program, but it won't change the ranking at all.

Edit:

So as not to confuse anyone, what Mondo gave was the likelihood that someone was gifted given that you already knew what type they were. The data CC gave was the opposite: the probability that someone who you already knew was gifted was of a certain type.

Mondo's table showed that, of all the types, the likelihood of an INTP being gifted was highest; and CC's table showed that, if you had a class of gifted students, you'll have more ENFPs than any other type. The discrepancy comes from there being more ENFPs in the general population.