[dghlsakjg]

That’s likely apocryphal.

Morse requires you to know when the tone both starts and stops in order to differentiate a ‘dah’ from a break indicating a new letter or word or even just a ‘dit’.

Tapping a pencil only gives you the start of a tone since the pencil lift is silent. There is no real way to distinguish between a short tone and a pause (letter e) and a long tone (letter t) if you don’t know when the tone ends and a pause begins.

The same trope is shown in movies. You cannot tap Morse code if the recipient cannot hear when the tap ends.

In other words, the receiver has no way of knowing whether you sent an e or t without there being a signal that the pencil has lifted. Note that e or t can be substituted for any other number of paired letters.

[cortesoft]

Couldn't you just use like the tip of the pencil for a dot and the eraser end for a dash? Use the tonal differences instead of duration?

[dghlsakjg]

Yes. You could just invent a different code.

The other issue is that it can take weeks of study to even reach a minimal level of fluency in Morse. 5 WPM is the basic metric. You could tap faster with much more practice, but the proctor would almost certainly notice the student furiously tapping patterns onto their desk with their pencil.

To review you need to devote weeks of study to learn an encoding that is poorly suited for the task, and that has a very slow transmission rate so that you can transmit a message that everyone in the room might be able to hear?

Morse just isn’t a practical way to cheat.

[cortesoft]

Depends on the test. A multiple choice or true/false test would be easy to convey.