A person can be born on a Monday with probability 1/3 and on any other
day with equal probability. What is the probability that 4 randomly
chosen persons were born on different days of the week.
6!/4! * (1/9)^4 + 6!/3! * (1/9)^3 * 1/3
denominator = 7^4
*took two cases where one 4 days are possible but not monday
other case is when it might be on a monday and 3 other days
the answer seemed not correct
Best Answer
You have the correct idea, but the selection of days is incorrect. In the first case, where you have 4 different days, none of which is a Monday, you have
$$6\times5\times4\times3 = {6! \over 2!}$$
possibilities to choose the (ordered) Weekdays ($\neq$ Monday).
In the 2nd case (one Monday in the set), you have
$$6\times5\times4 = {6! \over 3!}$$
ways to select the order of the non-Monday days, and $4$ ways where you put the Monday into that order.
That means your overall probability is
$${6! \over 2!}\left({1 \over 9}\right)^4 + 4\times{6! \over 3!}\left({1 \over 9}\right)^3{1 \over 3} = {360 + 4\times120\times3 \over 9^4}={1800 \over 9^4}={200 \over 9^3} \approx 0.274\ldots$$
I'd like to stress that it is imperative that you use ordered weekdays in the counting, even though the actual condition does not depend on any order. The reason is that those $\left({1 \over 9}\right)^4$ and similar probabilities depend on an order.
To make one specific example, the probability that the first person picked has birthday on Tuesday, the second picked on Wednesday, the third picked on Thursday and the 4th on Friday is $\left({1 \over 9}\right)^4$. The probability that the birthdays are (in order from 1st to 4th picked) (Friday, Thursday, Wednesday and Tuesday) is also $\left({1 \over 9}\right)^4$, and the same is true for the other 22 permutations of those 4 days. Each of those possibilities is a different event, each has probability $\left({1 \over 9}\right)^4$.
If ones makes the error of not considering order, one is lumping all those 24 cases into 1 case. If the used probability is still $\left({1 \over 9}\right)^4$, this is too small by a factor of 24.
EDIT: I see Henry got the same result in a comment, I think that result was added after I looked at the comment (I remember it was there, but shorter).