Testing

[spaceboots06]

I was having a discussion with my friend and it quickly changed subject from Michael Jackson to whether a monkey would do better on a test that is 10000 questions long over a test that was 1 question long. Although I am aware that as something is tested over and over the right/wrong ratio will even out, but with a monkey I think it would be entirely different. I think and will continue to think that a monkey will do better on a test that is one question long because there's less opportunity for the monkey to choose a wrong answer. (We were referring to four choice multiple choice tests.) I emailed my old professor and he was pretty neutral about this, he couldn't give us a definitive answer. I do not think monkeys are like flipping a coin or rolling dice.

I just wrote this on Facebook to a friend in defense of my argument.

Let's say it was a fill in the blank test. If it was a one question test the monkey has an infinite amount of answers to put in the blank, if it was a two question test, the monkey has an infinite amount of answers to put in both blanks. Given that the questions aren't the same, wouldn't it become double, triple, quadruple more times to put in a correct answer[as the test progresses]? And given that a monkey can enter numbers in a blank space.

What do you guys think? Sorry if my rambling isn't too clear.

 

[RITZFit]

are the results of each answer instant? (will it quickly find out it was right or wrong before going to the next question).