Quality vs. quantity: comparing evaluation methods in a usability-focused software architecture modification task