In: Statistics and Probability
Probability trees 1. It is estimated that 100 out of 9,900 women who participated in a breast cancer screening had breast cancer. Of the women who participated in the screening and have breast cancer 80 out of 100 will have a positive mammogram. Of the women who participated in the screening and do not have breast cancer 950 out of 9800 will have a positive mammogram.
a) Create a probability tree
b) Calculate the probability that a woman has a positive mammogram given that she does not have breast cancer.
c) Calculate the probability that a woman has a negative mammogram.
d) Given that a woman has a negative mammogram calculate the probability that she has breast cancer.
e) Construct a contingency table comparing cancer diagnosis vs mammogram result.
f) Based on the contingency table what proportion of women have a negative mammogram result given that they have breast cancer.
g) Based on the contingency table what proportion of woman have breast cancer and a positive mammogram test.
h) Calculate the probability that a woman has breast cancer or a positive mammogram result.
i) Can the event “ Cancer Diagnosis or Positive test ” be considered independent events?
Solution-
Let the events that women having breast cancer diagnosis, breast cancer not diagnosis, positive mammogram test and negative memmogram test are denoted by C, C' , + ,and – respectively.
Now,
According to question, 100 out of 9,900 women who participated in a breast cancer screening had breast cancer.
So, P(C) = 100/9900 = 0.01 .....(1)
And number of women not having breast cancer is 9900-100 = 9800
So, P(C') = 9800/9900 = 0.99 ....(2)
The women who participated in the screening and have breast cancer 80 out of 100 will have a positive mammogram.
So, P(+/C)= 80/100 = 0.8 .....(3)
And women having breast cancer bit negative test are 100-80 = 20
So, P(-/C) = 20/100 = 0.2 ....4)
The women who participated in the screening and do not have breast cancer 950 out of 9800 will have a positive mammogram.
So, P(+/C') = 950/9800= 0.0969 ...(5)
Women not having breast cancer and negative mammogram test are = 9800-950 = 8850
So, P(-/C') = 8850/9800 = 0.9031 ....(6)
(a)
Uding the above information ,the probability tree of the above events is shown below-
(b)
The probability that a woman has a positive mammogram given that she does not have breast cancer.
P(+/C') = 0.0968 (from equation 5)
(c)
Total women = 9900
Women having negative test = 20 +8850=8870
So, the probability that a woman has a negative mammogram is
P(-)= 8870/9900 = 0.89
(d)
Number of women having negative memmogram test =20+8850 =8870
Number of women having breast cancer but negative test = 20
So, of it is given that a woman has a negative mammogram then the probability that she has breast cancer.
P(C/-) = 20/8870 = 0.00225479
(e)
The contingency table comparing cancer diagnosis vs mammogram result is shown below-
Number of women | having positive mammogram test | having negative memmogram test | Total |
having breast Cancer diagnosis | 80 | 20 | 100 |
having breast Cancer not diagnosis | 950 | 8850 | 9800 |
Total | 1030 | 8870 | 9900 |
(f)
Based on the contingency table the proportion of women have a negative mammogram result given that they have breast cancer is
= 20/100 = 0.2
(g)
Based on the contingency table the proportion of woman have breast cancer and a positive mammogram test is
=20/9900 = 0.00202
(h)
Number of women having breast cancer or positive test result = 9900 -8850 = 1050
The probability that a woman has breast cancer or a positive mammogram result.
P(C+) = 1050/9900 = 0.10606
(i)
No, the events “ Cancer Diagnosis or Positive test ” can be considered independent events because if a person has cancer diagnosis then he has more chances of positive test with respect to women who has not cancer diagnosis.So, two events are depend on each other.