critical inquiry

Question 1

sensitivity

Answer 1

the proportion of people who HAVE the disorder who test POSITIVE

a/(a+c)

Question 2

specificity

Answer 2

the proportion of people who DO NOT HAVE the disorder who test NEGATIVE

d/(b+d)

Question 3

positive predictive value

Answer 3

the proportion of the people who test positive who have the disorder
( probability that someone who tests positive has the disorder

a/(a+b)

Question 4

negative predictive value

Answer 4

the proportion of people whoe test negative who do not have the disorder
(probability that someone who tests negative does not have the disorder)

d/(c+d)

Question 5

+ LR

Answer 5

• sensitivity/(1-specificity)
• the higher the number, the more certain you can be that a positive test indicates the person has the disorder
• > 10 is very good for ruling disorder IN
• tests with highest + LR provide the most info in event of + test

Question 6

• LR

Answer 6

• (1-sensitivity)/specificity
• lower the number, the more certain you can be that a negative test indicates the person does not have the disorder
• <.1 is very good at ruling disorder out
• tests with lowest - LR provide most info in event of - test

Question 7

pre test probability

Answer 7

• very high means negative test are usually false negatives
• very low means positive tests are usually false positives
• when using sequence of tests, the post-test probability of the first test becomes the pre-test probability for next test

Question 8

dependent variable

Answer 8

outcome or response variable

Question 9

independent variable

Answer 9

variable associated with the outcome of interest that contributes information about the outcome in addition to the provided by other variables considered simultaneously

Question 10

hierarchy of evidence

Answer 10

1a: SR of RCTS or N of 1 trials
1b: RCT
2a: SR of cohort studies
2b: cohort studies
3a: SR of case control studies
3b case control study
4: case study, case series, cross sectional study
5: expert opinion, clinical experience

Question 11

central tendency

Answer 11

mean: best choice of numeric, symmetrically distributed data; not good for skewed data bc affected by extreme scores

median: best for non symmetrical data; not affected by extreme scores

mode: limited utility; nominal or ordinal data, common in surveys

Question 12

standard deviation

Answer 12

1: 68%
2: 95%
3: 99%