Quiz 7 a
1. Give an example of a random variable.
Remember that a random variable is a random, *numerical* outcome of an experiment.
So your answers should have an experiment mentioned, and the outcome
should be (a) random and (b) numerical. The most common mistake was
to describe non-numerical variables. The second most common mistake
was to give a vague, sometimes impossibly vague, description of the experiment.
Here are some sample answers from class:
• Whether or not you get in a car accident
First -- there's no experiment. We
could create one by saying "when you drive home tonihgt" or "in your lifetime"
or something that gives us a period of observation. Second, the outcome,
while random, is not numerical. We could improve it by saying, "a 1
if you get in an accident, and a 0 if not."
So an improved version: X is
a random variable that is a 1 if you are in a car accident anytime before
midnight tonight, and a 0 otherwise.
Let's hope for 0's.
• The height of a person in the US that you will choose at random.
Got the experiment--- choose a person. Got the randomness
-- while one's height is not random, choosing a person at random means that
the height we choose is random. And got the number. A near-perfect
answer. Could be made more precise by saying "heihgt in inches" for
example.
• A card chosen out of a deck of normal playing cards.
We've got an experiment with a random outcome, but the
outcome is not numerical. Maybe if we said "the numerical value of
a card chosen at random, where Jack, Queen, King = 10, Ace = 1."
• conducting an experiment on how many teenage girls get pregnant before
the age of 16. Asking random 16 year olds or younger girls.
The experiment is a little fuzzy. We could fix this up by saying "Take
a random sample of 16 year olds and ask if they've been pregnant." The
numerical outcome, then, is the number of girls in the sample who answer
"yes".
• If parents are trying to have kids but want a girl, then the expected number
of tries before having a girl is 6 and the random variable would be girl
or boy.
Here's its not clear what the experiment is. Is
it having a single child? Having 6 children? And its not clear
where the 6 comes from. It is clear that the outcome "girl or boy"
is not numerical. This could be fixed by saying "the number of girls
in a randomly selected family of 6 children."
• Drawing the top card from a newly shuffled deck of cards. Having
to guess the color outcomes (black or red) provides the variable. Again,
no numerical value.