Because the computer "knows" it occasionally makes mistakes, for each word it offers as its best guess, it generates a list of alternative words. In some voice recognition programs, this list appears in a suggestion window on the screen and the words in it change with each dictation. The user can correct a mistake by choosing the desired word from this list if it appears there.
If the correct word is not in this list of alternatives, the user can spell it aloud, letter by letter, or begin typing the letters on the keyboard. The computer will use this information to predict the right word.
If you haven't seen the screen-by-screen demonstration
of how voice recognition works, now would be a good time to do so by selecting
the demo button below. Keep in mind that every system works a little bit
differently, but this generic demonstration will help you visualize the
essential procedures that all systems rely on to some degree. After the
demonstration you will have the opportunity to ask me questions.
Link to demo