Chapter 5 - Probability Flashcards
Conditional Probability
Probability of a certain thing given another certain thing
Book example 1 –> P(9 year old | girl) –> probability of being a 9 year old in the subcategory of girls
Sensitivity and Specificity
Sensitivity, or true positive fraction = probability that a diseased person screen positive
P(screen positive | disease)
Specificity, or true negative fraction = probability that a disease free person screens negative
P(screen negative | disease free)
False positive fraction vs. False negative fraction
False positive fraction = 1 - specificity
P(screen positive | disease free)
False negative fraction = 1 - sensitivity
P(screen negative | disease)
Positive and negative predictive value
Positive predictive value - probability that if I have the disease, the test will come back positive
P(disease | screen positive)
Negative predictive value - probability that if I don’t have the disease, test will come back negative
Independent events
occurrence of one event is not affected by the occurrence or non-occurrence of another event
P(A | B) = P(A) OR P(B | A) = P(B)
Bayes’ Theorem
Probability rule that is used to compute conditional probability based on specific available information
See Page 74 for equation
Complementary Events
probabilities of complementary events sum to 1
Binomial distribution
probability distribution for experiments/processes have 2 outcomes
must clearly specify which outcome is the "success" and which outcome is the "failure," even if the negative outcome is the success See page 75 for equation
appropriate use must satisfy three assumptions:
1) only two outcomes
2) probability of success is the same for both outcomes
3) probabilities of success and failure are independent
Binomial mean number of successes and standard deviation
Mean = mu = np
standard deviation = little sigma = sqrt(n(p)(1 - p))
The Normal distro
appropriate when an experiment or process results in a continuous outcome
Gaussian distro
Assumptions
1) Mean = median = mode
2) ~68% of population values fall between one std dev above and below the mean
3) ~95% of population values fall between two std devs above and below the mean
4) ~99.9% of population values fall between three std devs above and below the mean
5) symmetric about the mean
see page 79 for equation - needs calculus o\_\_O
for any probability distibution, total area under the curve is 1
Probability of a Characteristic
(Number of people with a characteristic)/(Number of people in the population)