The Arduino selects an audio track to play, plays it, waits for the user to type what he/she heard onto the PC keyboard , then compares the results, expected versus actual, and switches a led accordingly. Is that it?
I suggest starting with Serial Input Basics - updated then work out how you are going to associate the written form of the sound with the sound track so you can check the spelling. Progress from there. Have you already got an SDcard or similar with recorded test words for the player?