Strip the theory jargon away and constructive alignment is one question asked three times: what should someone be able to do? The outcome states it, the activities rehearse it, the assessment verifies it. When all three point at the same capability, the course works almost regardless of style. When they don't, learners optimise for the assessment — because learners always optimise for the assessment — and the stated outcome quietly loses.
The misalignment tells
You can audit any course in ten minutes. Find an outcome verb ("apply," "evaluate," "perform") and then find the assessment. If the outcome says apply and the assessment says select the correct definition, that's the gap. Common tells: outcomes written after the content was built, activities that demonstrate rather than rehearse, and assessments inherited from whatever the tool makes easy.
Work backward, always
Alignment is cheap if you start with it and expensive to retrofit. The working order: define the outcomes with whoever owns the problem, design the assessment that would genuinely evidence each one, and only then build activities that prepare learners for that assessment. Wiggins and McTighe called it backward design; in practice it's just refusing to write content before you know what it must produce.
One honest example
A recent financial-literacy build had three outcomes: know the journey, speak the language, decide well early. Every activity and every assessment item maps to exactly one of the three — and anything that mapped to none, however interesting, was cut. That single discipline did more for the course than any interaction pattern in it.
Alignment isn't a framework you cite. It's a subtraction you enforce.