Things to Just Memorize

Question 1

What is wider, a 95% CI or a 90% CI?

Answer 1

95% CI is wider, the 90% CI is more narrow

Question 2

If a result for a CI of a group difference excludes 0, is it statistically significant?

Answer 2

It is statistically significant if it excludes 0

Question 3

If a result for a CI of a group ratio excludes 0, is it statistically significant?

Answer 3

Not necessarily, for group ratios, the CI must exclude 1 in order to be statistically significant

Question 4

“Does this patient have appendicitis?” is an example of deductive or inductive reasoning?

Answer 4

Deductive reasoning: comparing the current case against a pattern

Question 5

“What could this patient with right lower quadrant pain have?” is an example of deductive or inductive reasoning?

Answer 5

Inductive reasoning: generating a list of differential diagnoses and finding a pattern that fits the symptoms

Question 6

“I know my patient has the disease. What is the chance that the test will be positive?” is sensitivity or specificity? What is the equation?

Answer 6

Sensitivity, = TP/(TP+FN)

Question 7

“I know that my patient does not have the disease. What is the chance that the test will be negative?” is sensitivity or specificity? What is the equation?

Answer 7

Specificity, = TN/(TN+FP)

Question 8

“I just got back a positive test result. What is the chance that my patient actually has the disease?” is positive or negative predictive value? What is the equation?

Answer 8

Positive predictive value, = TP/(TP+FP)

Question 9

“I just got back a negative test result. What is the chance that my patient does not have the disease?” is positive or negative predictive value? What is the equation?

Answer 9

Negative predictive value, = TN/(TN+FN)

Question 10

Which of the following are characteristics that are dependent on the disease prevalence:
Sensitivity
Specificity
Positive Predictive Value
Negative Predictive Value

Answer 10

Positive and negative predictive value will change with disease prevalence. Sensitivity and specificity are characteristics that are inherent of the test.

Question 11

What is the equation and interpretation of a positive likelihood ratio?

Answer 11

is a ratio of people w/ positive test result among disease over people w/ positive test result without disease)
PLR = Sensitivity/(1-specificity)

Question 12

What is the equation and interpretation of a negative likelihood ratio?

Answer 12

Ratio of people w/ negative test result among disease over people w/ negative test result without disease
NLR = (1-sensitivity)/specificity

Question 13

Which of the following depend on disease prevalence?
Positive and negative predictive values
Positive and negative likelihood ratio

Answer 13

Only positive and negative predictive values depend on disease prevalence. Positive/Negative likelihood ratios are derived from sensitivity and specificity which does not depend on the disease prevalence

Question 14

What are the attributes for that you want for a screening test?

Answer 14

High sensitivity (so as to not miss someone with the disease)

Question 15

What are the attributes desired for a confirmatory test?

Answer 15

High specificity (so not to confirm someone with the disease who doesn’t have it)

Question 16

What is Baye’s theorem?

Answer 16

Pre-test odds * likelihood ratio of test result = post-test odds

Question 17

What is the relationship between odds and probability?

Answer 17

Odds = Probability/(1-Probability)