The change in evaluation is considered equivalent to a win/loss of material (with pawns worth 1 point, bishops/knights 3 points, etc.)
There are times when losing a pawn, worth 1 point, is a pure blunder. There are times when it is a strategic sacrifice.
It is your job to determine which is which.
What engine analysis score change from a move would be classed as a blunder, bad or questionable move