For large values of n, the distributions of the count X and the sample proportion "\\hat{p}" are approximately normal. In practice, the approximation is adequate provided that both "np\\geq5" and "n(1-p)\\geq10" , since there is then enough symmetry in the underlying binomial distribution.
Given the proportion "p" is expected to be 0.48 and from past records "\\hat{p}" was found to be 0.4 and "n=10." Then
"np=10(0.48)=4.8<5,"
"n(1-p)=10(1-0.48)=5.2<10"
Normal distribution cannot be used.
Binomial distribution is appropriate for this event. Then "10(0.4)=4"
Probability of breaking the record
"P(X>4)=1-P(X=0)-P(X=1)-" "P(X=2)-P(X=3)-P(X=4)="
"=1-\\dbinom{10}{0}(0.48)^0(1-0.48)^{10-0}-\\dbinom{10}{1}(0.48)^1(1-0.48)^{10-1}-"
"-\\dbinom{10}{2}(0.48)^2(1-0.48)^{10-2}-\\dbinom{10}{3}(0.48)^3(1-0.48)^{10-3}-"
"-\\dbinom{10}{4}(0.48)^4(1-0.48)^{10-4}\\approx"
"\\approx1-0.0015-0.0133-0.0554-0.1364-0.2204\\approx"
"\\approx0.573"
Comments
Leave a comment