sfwon
Not sure if someone has an elegant solution to it - i'd be interested in seeing one. But for test taking purposes, here's how I would tackle the problem.
Looking at the first data:
Knowing that a=b does not provide sufficient information. When b is negative, then 1/a > 1/(b+1) whereas when b is positive then 1/a < 1/(b+1)
Looking at the second data:
Knowing b>0 alone does not provide sufficient information. a could be any number so 1/a could be greater or less than 1/(b+1).
Looking at the two data combined:
If a is equal to b and b is a positive number (thus a is a positive number) then we have sufficient information to answer to question. if we sub in a for b, the question becomes is 1/a < 1/(a+1) where a is positive? The answer is intuitively no.
Thus, 1 and 2 combined is enough to answer the question but neither alone is sufficient.
For data 1), when a=b and b is positive, lets just take a=b=4, then 1/4 > 1/5, correct? So when a=b>0, I think 1/a > 1/(b+1) (not 1/a < 1/(b+1) as you pointed out).
So my conclusion is with a=b, we can always conclude that 1/a > 1/(b+1) - thus the answer should be A as data 1 is sufficient to answer the question.
Where did I go wrong here?