Inference questions test your ability to read between the lines and figure out what the author is indirectly implying.
To infer means: to draw as a conclusion, using your own deduction, an information or an implication that has not been stated directly anywhere in the passage. So, very simply put, if you see a clear statement in black and white, you know it cannot be an inference. It can only be an inference if it states something visibly different from what has been stated explicitly. Now, the way RC questions do this is that they facilitate room for this implicit inference in the immediate vicinity of whatever the passage does make explicit. In other words, If explicit statements A and B have been strongly and very obviously mentioned in the passage, then the “inference” question type will ask you about inference C that flows so logically from A and B that any logical person will be able to reach it without a second thought. In fact, C has to be the most obvious inference and hence the correct answer. If it is possible for A & B to be true, but not for C to be true, then C is not the correct inference. By its very nature, and under these test conditions, the inference must be an absolutely and unarguably necessary consequence of what is stated explicitly (A & B here).
So now, if it is indeed this obvious, then why is the “inference” question type considered one of the toughest? Well, it’s actually because C is so obvious that errors happen. More often than not, because C is such an unavoidable consequence of A and B, the reader of the passage often automatically assumes the explicit presence of C! Indeed, assumptions can work in the trickiest of ways. Whenever we assume the presence of a fact automatically, we tend to imagine it’s explicit mention in the passage even if it is not actually there. Hence, you must bear caution when approaching these questions. The good inference will be that statement that you unavoidably assume, on the basis of what is explicitly stated, but if you look carefully, that statement itself has not explicitly been written in the text.
Always remember to stay hyper-faithful to the passage. If something seems extraordinarily obvious but is not mentioned anywhere, take notice of that, and only believe whatever the passage has explicitly offered you to read and analyze.
For example, if the passage reads, “Akshay has read every book by Enid Blyton at least once”, we cannot necessarily infer that “Akshay enjoys reading” — maybe he hates to read, but has had to read Enid Blyton specifically as part of some kind of assignment!
On the other hand, an undeniable implication is: “Akshay has read the Secret Seven series at least once in his lifetime.” If you read that sentence carefully and compare it with what the original sentence in the passage read, you would see that this second sentence has to be true without a question, simply because the first statement makes it that much obvious. That’s the level of logical undeniability that you should seek when you answer “inference” questions.