Frazz by Jef Mallett for January 27, 2010
January 26, 2010
January 28, 2010
Transcript:
Caulfield says, "Assuming a typical distribution of answers on a multiple-guess test, it works out that your best odds are with (b) or (c) 60% of the time." Caulfield says, "Which drops to 48% if that was the previous answer, and another five percentage points per consecutive occurrence. Plus or minus 1.5." Caulfield says, "Beyond that it gets complicated." Frazz says, "I thought you hated math." Caulfield says, "It's different when you can apply it to something."
…Just one problem: this depends on how many choices there are, and whether the test designer took into account the “guess method” and randomized the pattern of correct answers to foil this method.