The fact that people are debating the meaning of a question, probably means it's not as clear as it could be, or that the answer is somewhat subjective. Creating a multiple choice test question is a lot tougher than it would seem. Writing an entire test is a huge endeavor. Poorly written test questions often have one or more of the following characteristics: 1. The are ambiguous 2. They attempt to measure somthing for which they are not well-suited 3. They contain clues as to the correct answer The triple mat question could be labeled as at least one of those. Also, while few of us would use NG glass with a triple mat, it's at least somewhat subjective, and it also doesn't account for other factors, such as the the amount of contrast in the item being framed. Not only is writing a test question difficult, but writing the 3 wrong answers (assuming 4 choices) is really tough. The wrong answers can't be so obviously wrong that you can easily eliminate them. Good wrong answers make guessing more difficult. In theory, you could blindly guess at every answer and have a mathematical chance of getting 25% correct. A good test is one that evolves from feedback like what we've read in this thread. Every multiple choice test in all but a few time-tested fields needs to go through a feedback-review-modfication process.