A case for the usefulness and plausibility of Thinking-Aloud verbal data