Textbook Notes
(363,141)

Canada
(158,218)

University of Ottawa
(6,458)

BPS1101
(3)

William Ogilvie
(3)

Chapter 2

# BPS1101 Chapter 2: Use of Natural Frequencies

Unlock Document

University of Ottawa

Biopharmaceutical sciences

BPS1101

William Ogilvie

Winter

Description

!"#$%&$'()*+(,$-+#.*#/01#"$)%$#2(,*()#$
")()1")10(,$1/&%+3()1%/4$
$
1. The following data was presented in the assigned reading:
a) The probability that a woman has breast cancer (prevalence) is 1
percent.
b) If a woman has breast cancer, the probability that she tests positive
(sensitivity) is 90 percent.
c) If a woman does not have breast cancer, the probability that she
nonetheless tests positive (false-positive rate) is 9 percent.
(Note that b and c indicate a test accuracy of about 90 %).
How to convert to Natural Frequencies:
- Pick a big number of patients (100,000) then do the math. To get numbers
using percentages, multiply the total by the percent and divide by 100. Using a
flow-chart is helpful
100,000 women
1% have breast cancer
Cancer Healthy
1000 99000
(100000 X 0.01) (100,000-1000)
90 % test positive False positive rate 9%
900 cancers 100 cances 90, 090 don't 8910 test positive
detected missed have cancer for cancer but
(1000 X 0.90) (1000 X 0.1) (99000 X 0.91) don't have it
(99000 X 0.09)
for every 9,810 positive tests (8910 + 900) only 900 actually have cancer (1 in 10)
If a woman tests positive for breast cancer, hopefully her doctor does follow-up
tests to make sure she actually has the disease.
2. A company with 10,000 employees is performing lie detector tests to find 100
people who have embezzled money. If the test is 99 % accurate, how many innocent people will be falsely accused? If a person tests as “guilty” what is the
chance that they actually did it?
Answer: There are 100 guilty people and (10000-100) 9900 innocent people
10,000
100 embezzlers
Guilty Innocent
100 9900

More
Less
Related notes for BPS1101