In compound jinās, one expression is a single word while the other is composed of separate words or a compound expression that sounds like the first. The continuous acoustic stream conceals a difference in segmentation.
Correct analysis must identify word boundaries: a single lexical item on one side, multiple components on the other. Merely making two sentence endings rhyme does not establish compound jinās unless the expected likeness between the actual expressions is present.
Subtypes depend in part on whether the written form agrees. Pronunciation, spelling and composition therefore need to be analyzed together.