When you teach the pronunciation of an English word, you probably know the traditional approach: the teacher writes the word on the board or puts it on a slide, explains the meaning and the stress, reads it aloud, and the class repeats. Is that actually the best way?
Under the Cambridge approach, teaching a word's pronunciation has three stages.
1. Model. The teacher says the word, or plays a recording of it being said correctly.
2. Process. Analyse the pronunciation. The teacher talks through the stress, any diphthongs or triphthongs, and the sounds that are commonly misread so learners can avoid them.
3. Drill. The teacher says it and learners repeat a few times to get used to it, with corrections where needed.
So the written form only earns its place in the first two stages, where you are analysing. By stage three, when learners are imitating what they hear, the thing to attend to is the sound, not the spelling.
On top of that, English words frequently are not pronounced the way they are written. Looking at the letters at this point can pull learners off course, because they will fall back on habit and read it the way they assume it should sound.